| 00:00.09 | spldart | And it's got to do with what? The Vid card? |
| 00:00.14 | JeffM | no |
| 00:00.32 | JeffM | basicly the client keeps track of what time it thinks it is on the server |
| 00:00.40 | JeffM | and every so often they try to sync |
| 00:00.45 | spldart | Oh... IC yeah |
| 00:00.51 | JeffM | and that tells me how far "off" the server was from the cleint |
| 00:01.11 | JeffM | it's so that when we send updates, we can stamp them with a server time, and keep it as accurate as possible |
| 00:01.15 | spldart | That was my first guess.. but the statement above confused me cuz I didn't pay attention to the context |
| 00:01.31 | JeffM | bast had his windows server bouncing around in the 3 to 6 second range, and that's not good |
| 00:01.32 | Epyon | is wondering of introducing BZFloat -_- |
| 00:01.39 | JeffM | Epyon, good lord why? |
| 00:01.46 | spldart | So does that .012 correspond to 120 ms lag time? |
| 00:01.47 | Epyon | Using GLFloat would be not wise |
| 00:01.58 | JeffM | spldart, no clock drift |
| 00:01.58 | Epyon | float also |
| 00:02.04 | JeffM | Epyon, use float |
| 00:02.11 | Epyon | k, if you say so |
| 00:02.16 | JeffM | don't make the code MORE complex |
| 00:02.18 | spldart | Sry.. didn't word that right |
| 00:02.28 | Epyon | We'll worry about it later if the need arises :P |
| 00:02.47 | Epyon | Nowhere are double vectors used? |
| 00:02.54 | JeffM | in general no |
| 00:02.54 | Epyon | In the code? |
| 00:03.00 | JeffM | we don't need that much accuracy |
| 00:03.05 | spldart | 120 ms drift of the clock sync? |
| 00:03.09 | Epyon | So no need for a templated vector class I guess. |
| 00:03.11 | JeffM | spldart, yes |
| 00:03.23 | JeffM | Epyon, no |
| 00:03.25 | JeffM | keep it simple |
| 00:03.32 | Epyon | KISS :P |
| 00:03.35 | JeffM | simple and clean |
| 00:04.07 | JeffM | and well named, everyone has to use them |
| 00:04.34 | Epyon | calling a class Vector3D is ok? :P |
| 00:04.43 | JeffM | it's descriptive ;) |
| 00:05.00 | Epyon | There shan't be any clashes. |
| 00:05.36 | jwir3 | You could take a page out of the CrystalSpace book & prefix it with something that verifies it'll be unique to the project: e.g. for CrystalSpace, we use csVector3 |
| 00:05.40 | JeffM | if there is, then that means we have a vector class you shoudl be using ;) |
| 00:05.47 | jwir3 | you could start a standard and do bzVector3D |
| 00:05.49 | JeffM | jwir3, we don't do that anywhere else |
| 00:05.50 | Epyon | jwir3, no |
| 00:06.05 | JeffM | if we have to later we will |
| 00:06.21 | spldart | Would that drift potentially exacerbate the existing lag? |
| 00:06.40 | jwir3 | oh ok |
| 00:06.43 | jwir3 | just a thought ;) |
| 00:06.43 | JeffM | spldart, it should not really affect lag at all |
| 00:06.55 | spldart | ok.. I'm done :) |
| 00:07.10 | JeffM | it is used in lag compensation |
| 00:08.37 | whodaman- | ~Constitution++ |
| 00:10.18 | *** join/#bzflag Marcello (n=837a3296@gateway/web/cgi-irc/zeebrothers.net/x-687638da75de4b98) |
| 00:10.18 | *** mode/#bzflag [+b *cell*!*@*] by ChanServ |
| 00:10.18 | *** part/#bzflag Marcello (n=837a3296@gateway/web/cgi-irc/zeebrothers.net/x-687638da75de4b98) |
| 00:24.35 | donny_baker | Erroneous: got the menu bar on the forums down to 766 px wide... got 9 px under your goal and added the wiki button :P |
| 00:24.38 | Epyon | wonders about BZF code conventions... |
| 00:24.46 | Epyon | Should "inline Vector3D() {}" |
| 00:24.49 | Epyon | realy be: |
| 00:24.52 | Epyon | inline Vector3D() |
| 00:24.53 | Epyon | { |
| 00:24.54 | Epyon | } |
| 00:24.56 | Epyon | ? O.o |
| 00:27.07 | *** join/#bzflag a_meteorite (n=a_meteor@unaffiliated/ameteorite/x-000000001) |
| 00:37.27 | JeffM | Epyon, you you never read devinfo did you? |
| 00:37.52 | *** join/#bzflag a_meteorite (n=a_meteor@unaffiliated/ameteorite/x-000000001) |
| 00:38.02 | Erroneous | that's actually an exception to devinfo |
| 00:38.12 | Erroneous | sometimes entire functions can go on one line |
| 00:38.15 | Erroneous | that's ok |
| 00:41.31 | Epyon | JeffM, I did |
| 00:41.42 | Epyon | That's why I'm asking |
| 00:42.10 | Epyon | JeffM, I need help :) |
| 00:42.36 | *** join/#bzflag KingofCamelot (n=davidsan@207-177-229-17.dsl.redshift.com) |
| 00:42.38 | blast007 | we can't provide that kind of help ;) |
| 00:42.54 | Winny | hehe |
| 00:43.03 | blast007 | oh, you mean with bz? |
| 00:43.09 | Epyon | JeffM, I added the new files, but svn stat reports a LOT of files changed O.o |
| 00:43.20 | Epyon | C/M |
| 00:43.35 | blast007 | what about svn diff |
| 00:43.35 | Epyon | All the .vcproj files actually |
| 00:43.54 | Epyon | Oh dammit |
| 00:44.00 | Epyon | I know why -_- |
| 00:44.34 | JeffM | upgraded them to VC9? |
| 00:44.58 | Epyon | The 71 ones to 8 :P by accindent |
| 00:46.15 | JeffM | that'd do it |
| 00:46.19 | JeffM | that reminds me |
| 00:46.54 | TheRedBaron | whodaman-: can you report your lagstats at the gu forum? |
| 00:47.44 | CIA-16 | BZFlag: 03JeffM2501 * r16795 10/trunk/bzflag/MSVC/VC9/: pull VC9 cus it doesn't work, and users can just upgrade the VC8 ones because they DO work with no mods. |
| 00:54.25 | CIA-16 | BZFlag: 03chaos-dev * r16796 10/trunk/bzflag/ (4 files in 3 dirs): Added a Vector3D class to common |
| 00:56.37 | JeffM | why did the .sln change Epyon ? |
| 00:57.21 | Epyon | JeffM, I added the files? |
| 00:57.29 | JeffM | no, those are in the vc proj |
| 00:57.33 | JeffM | not the .sln |
| 00:57.34 | Epyon | Oh darn |
| 00:57.49 | Epyon | hopes he didn't bust something |
| 00:57.57 | spldart | it built just now |
| 00:58.02 | spldart | vc8.. |
| 00:58.12 | Epyon | Aftre the commit? |
| 00:58.45 | spldart | I'm @ 796 |
| 00:58.54 | brlcad | Epyon: DEVINFO does speak to your convention question |
| 00:59.04 | spldart | just played for second though before I exited |
| 00:59.32 | brlcad | but that is true regarding exemptions for empty functions |
| 01:00.09 | JeffM | Epyon, set the line endings on new files to native |
| 01:00.33 | JeffM | or at least keep them consistent |
| 01:00.46 | Epyon | Hmm? |
| 01:00.59 | JeffM | the header had inconstent line endings |
| 01:01.15 | Epyon | Oh, because I copy/pasted it from another file :/ |
| 01:01.29 | JeffM | why woudl another file be difrent then your new file? |
| 01:01.30 | Epyon | Doesn't SVN convert them automagicaly? |
| 01:01.37 | JeffM | yes |
| 01:01.38 | Epyon | No idea |
| 01:02.14 | JeffM | good lord, the atributes on the thing are wicked wrong |
| 01:02.17 | Epyon | can't find the option in VC05 |
| 01:02.22 | JeffM | you set it to executable |
| 01:02.31 | Epyon | JeffM, I didn't :/ |
| 01:02.37 | Epyon | It did it itself |
| 01:02.39 | JeffM | your svn client did |
| 01:02.47 | JeffM | you using command line or tort? |
| 01:02.54 | Epyon | command line |
| 01:03.06 | JeffM | it is not configured correctly |
| 01:03.09 | Epyon | I'll check my config dile |
| 01:03.12 | Epyon | file* |
| 01:04.00 | CIA-16 | BZFlag: 03JeffM2501 * r16797 10/trunk/bzflag/include/Vector3D.h: consistent native line endings |
| 01:04.49 | brlcad | svn only does native if svn:eol-style=native on the file |
| 01:04.52 | CIA-16 | BZFlag: 03JeffM2501 * r16798 10/trunk/bzflag/MSVC/VC8/bzflag.sln: valid keys for the net project. |
| 01:04.59 | brlcad | if you're adding a new file, it has no props set |
| 01:04.59 | JeffM | brlcad, it was not set |
| 01:05.16 | JeffM | brlcad, unless you have your client set to set the props by mime type :) |
| 01:05.48 | JeffM | or extension in torts case |
| 01:05.53 | Epyon | auto-props is empty here :/ |
| 01:05.53 | brlcad | or a config file that auto-sets based on extension |
| 01:06.04 | JeffM | Epyon, should fix that |
| 01:06.13 | Epyon | I have no idea where that executable flag came from |
| 01:06.17 | Epyon | JeffM, to what? |
| 01:06.37 | Epyon | *.cxx = svn:eol-style=native? |
| 01:06.43 | Epyon | And h? |
| 01:06.53 | JeffM | http://pastebin.bzflag.bz/m5e796097 |
| 01:07.06 | Epyon | Thanks |
| 01:08.05 | Epyon | That still doesn't answer why the files were marked executable, does it? |
| 01:08.10 | brlcad | Epyon: curl http://bzflag.bz/~sean/subversion.config > ~/.subversion/config |
| 01:08.35 | brlcad | http://bzflag.bz/~sean/subversion.config is what I use, sets up a slew of the mime types, eol-style props, and autogen'd files to ignore |
| 01:08.57 | JeffM | Epyon, it's all detailed in our wiki, http://my.bzflag.org/w/SVN#Committing_Code_to_SVN |
| 01:11.11 | Epyon | brlcad, this ones wierd : *.com = svn:eol-style=native;svn:mime-type=text/plain -_- |
| 01:11.36 | JeffM | why is that odd? |
| 01:11.56 | Epyon | at least on windows com's were binary, weren't they? :) |
| 01:11.56 | jorgenpt | brlcad: Btw, shouldn't Makefile / Makefile.in _not_ be in the global ignores? OK for bzflag, but there you should've used svn:ignore on the dir instead, Makefile / Makefile.in are versioned on a few projects. |
| 01:16.42 | brlcad | jorgenpt: well yeah, if you deal with svn repos that manage those files directly |
| 01:16.48 | *** join/#bzflag Ratfink (n=434ce305@bz.bzflag.bz) |
| 01:16.55 | brlcad | kinda welcome to 1985 from my perspective |
| 01:17.38 | brlcad | automake, for all the warts of the autotools system is a vast improvement over hand coddling makefiles for all but tiny projects and projects *just* getting started |
| 01:18.20 | JeffM | fixes bzworldgen |
| 01:18.32 | CIA-16 | BZFlag: 03joevano * r16799 10/trunk/bzflag/README.WINDOWS: Spelling |
| 01:18.38 | Epyon | What was broken? |
| 01:18.43 | JeffM | all of it |
| 01:18.57 | Epyon | auto-props? :P |
| 01:19.01 | JeffM | ohh yeah |
| 01:19.07 | JeffM | line ending |
| 01:19.09 | JeffM | mime types |
| 01:19.13 | JeffM | executable flags |
| 01:19.49 | JeffM | inconsitent line endings |
| 01:20.39 | JeffM | rulesets are just text files right? |
| 01:21.09 | Epyon | yes |
| 01:21.20 | CIA-16 | BZFlag: 03JeffM2501 * r16800 10/trunk/bzflag/MSVC/VC8/bzflag.sln: unbork windows build, so that net links |
| 01:21.28 | JeffM | and yes you broke the sln ;) |
| 01:22.01 | Epyon | I will be slowly removing that vectors.h and substituting it with Vector4D, Vector3D and Vector2D if no one minds? |
| 01:22.19 | JeffM | make sure they are used first |
| 01:22.21 | JeffM | but yeah |
| 01:22.24 | JeffM | one is all we need |
| 01:22.55 | Epyon | It's not even in common O.o |
| 01:22.57 | *** join/#bzflag rob1n (n=rob1n@unaffiliated/rob1n) |
| 01:25.30 | Paul_PL | JeffM: I figured out what caused the 'World cache files disappeared...' massege |
| 01:25.38 | JeffM | cool |
| 01:25.40 | JeffM | what was it? |
| 01:25.47 | Paul_PL | I have a polish letter in my account name |
| 01:26.09 | Paul_PL | on the other accounts it runs smoothly :) |
| 01:26.21 | CIA-16 | BZFlag: 03JeffM2501 * r16801 10/trunk/bzwgen/MSVC/nsis/Thumbs.db: Thumbs.db does not belong in svn |
| 01:26.24 | JeffM | Paul_PL, ahh |
| 01:26.45 | Epyon | This one was NOT mine... |
| 01:26.51 | Epyon | i think -_- |
| 01:26.53 | Paul_PL | is this some kind of well known problem ? |
| 01:27.26 | CIA-16 | BZFlag: 03JeffM2501 * r16802 10/trunk/bzwgen/MSVC/ (15 files in 4 dirs): fix subversion properties, and line endings. |
| 01:27.29 | brlcad | don't think it's well known |
| 01:27.32 | brlcad | sounds like a bug |
| 01:27.33 | spldart | OMGLOL |
| 01:27.35 | JeffM | Paul_PL, well sota is, but it's not suprising |
| 01:27.37 | JeffM | yeah fix it ;) |
| 01:27.39 | brlcad | maybe you can fix that ;) |
| 01:27.47 | spldart | Thumbs got in trunk?! |
| 01:28.00 | spldart | gotta love windows |
| 01:28.14 | JeffM | spldart, thats the devs fault, not windos |
| 01:28.17 | CIA-16 | BZFlag: 03JeffM2501 * r16803 10/trunk/bzwgen/ (83 files in 6 dirs): fix subversion properties, and line endings. |
| 01:28.26 | JeffM | there are lots of files on all OSs that we want to not put in |
| 01:28.29 | JeffM | .obj, etc.. |
| 01:28.35 | JeffM | .o I mean |
| 01:28.39 | spldart | I know i know.. but it sure seems to find its way all over nstuff hehe |
| 01:29.11 | Paul_PL | thats what I was going to do :) don't have even to choose what to fix in addition to my proposal |
| 01:32.25 | *** join/#bzflag a_meteor (n=a_meteor@adsl-75-15-139-154.dsl.snlo01.sbcglobal.net) |
| 01:38.19 | *** join/#bzflag a_meteor_ (n=a_meteor@adsl-75-15-119-162.dsl.snlo01.sbcglobal.net) |
| 01:39.04 | *** join/#bzflag Lukstr (n=Lukstr@206-248-157-113.dsl.teksavvy.com) |
| 01:46.54 | CIA-16 | BZFlag: 03chaos-dev * r16804 10/trunk/bzflag/ (3 files in 3 dirs): Added a Vector2D class to common |
| 01:55.17 | CIA-16 | BZFlag: 03chaos-dev * r16805 10/trunk/bzflag/ (3 files in 3 dirs): Added a Vector4D class to common |
| 01:55.48 | *** join/#bzflag range-target (n=46f0bb0b@bz.bzflag.bz) |
| 01:57.40 | CIA-16 | BZFlag: 03brlcad * r16806 10/trunk/bzflag/ (18 files in 4 dirs): source files are not executable! |
| 01:57.49 | Epyon | AGAIN? |
| 01:58.06 | Epyon | I have no clue why that happens! |
| 01:58.14 | brlcad | run svn diff before you commit |
| 01:58.23 | brlcad | it'll tell you what is getting set |
| 01:58.40 | Epyon | Ok, I guess it will tell that executable is set... but why? |
| 01:59.08 | brlcad | that's up to your config and settings |
| 01:59.11 | Epyon | It does indeed |
| 01:59.26 | brlcad | if you took my config file and installed it properly, it shouldn't be happening |
| 02:01.28 | Epyon | brlcad, it's almost the same : http://pastebin.bzflag.bz/d26d7f6d0 |
| 02:01.30 | brlcad | is not happy with the vector math classes in common regardless of svn:executable |
| 02:01.49 | Epyon | brlcad, I asked JeffM about that |
| 02:02.45 | Epyon | I have no clue where that svn:executable flag gets set :( |
| 02:04.58 | brlcad | Epyon: more specifically, you asked *where* .. where is very much not lib3D .. either libCommon or libMath are where |
| 02:05.31 | brlcad | the answer to *if* it's needed/wanted/desirable is what I'd complain about |
| 02:06.13 | Epyon | Ah that one :) |
| 02:06.28 | brlcad | the benefit added is very petty imo, and can impact other factors (complexity, performance, maintenance) |
| 02:06.34 | Epyon | Well, vectors.h gets used and it's a mess |
| 02:07.08 | Epyon | All leading engines now implement Vectors as classes (OGRE, CS, Irrlicht) even for internal calculations |
| 02:07.28 | Epyon | Some of them (Irrlicht comes to mind) even template them. |
| 02:07.45 | Lukstr | templating vectors is nice |
| 02:07.47 | brlcad | never said vectors.h was a great idea either .. and if anything it means now there are two vector math codes |
| 02:07.52 | brlcad | so just added complexity |
| 02:08.13 | Epyon | brlcad, I'll be replacing vectors.h usage with Vector* classes |
| 02:08.43 | brlcad | and profiling before and after |
| 02:08.44 | Lukstr | sounds ambitious |
| 02:09.30 | Epyon | brlcad, I can profile cfvec3 versus Vector3D if you want, but Vector3D was built around the current industry standard :) |
| 02:09.52 | Epyon | (well, except ASMing some of the more complex operations as seen in OGRE and Irrlicht) |
| 02:09.53 | brlcad | or I swear I'll just rip them both out, that's the sort of encapsulation creep that is really counterproductive |
| 02:10.09 | brlcad | don't give me industry standard strawmen, do the profiles |
| 02:10.14 | Epyon | brlcad, then why does everyone do that? |
| 02:10.22 | Lukstr | brlcad: You're against a vector class? |
| 02:10.28 | Epyon | Scaary |
| 02:10.47 | brlcad | Lukstr: I'm against them being added carelessly and without their implementations being profiled |
| 02:10.52 | Epyon | is sad that JeffM isn't here to back him up. |
| 02:11.08 | brlcad | that's code that will not readily show up in a profile as it's spread out everywhere |
| 02:11.10 | Lukstr | hm |
| 02:11.20 | Lukstr | you could write a test suite |
| 02:11.25 | Epyon | Okay, I give up. |
| 02:11.25 | Lukstr | to compare the old and new |
| 02:11.41 | brlcad | that's what I'm saying is required |
| 02:11.52 | Lukstr | well that's not too much to ask |
| 02:12.13 | Lukstr | meanwhile, I'm getting ideas for my app :D |
| 02:13.27 | Epyon | Sooo, instead of writing and implementing code that (just for once) I know that is good, I should go and spend a month learning about profilers, and... ah whatever. |
| 02:13.30 | brlcad | Epyon: the point is not taking it on blind faith -- unless you pulled an implementation from someone else's project that has documented performance info, you're implementation is a complete unknown |
| 02:13.39 | brlcad | bs |
| 02:13.44 | brlcad | it doesn't take a month |
| 02:13.53 | brlcad | you're just being lazy, it's something you should have done already |
| 02:14.14 | Epyon | brlcad, this code is based on "3D Game Engine Design" by Dave Eberly. |
| 02:14.15 | brlcad | it takes two recompiles, and a web search for a tutorial on profiling |
| 02:14.27 | Epyon | Except for some reordering it's "by the book" |
| 02:14.31 | Lukstr | profiling is cake Epyon |
| 02:14.42 | brlcad | Epyon: and it has performance details? |
| 02:14.57 | brlcad | otherwise, being from a book is about as useful as being from a random website |
| 02:15.02 | Epyon | OGRE's Vector classes are a direct rip-off of WM3 (the engine from the book) |
| 02:15.27 | Epyon | And now tell me, that you doubt OGRE's performance and that we need higher quality. |
| 02:15.29 | brlcad | books are often written to be clear to understand with little regard for performance anyways |
| 02:15.43 | Epyon | *sigh* |
| 02:15.49 | brlcad | i'm saying you haven't done your own damn homework |
| 02:15.58 | brlcad | and you're reaching for any excus |
| 02:16.37 | brlcad | cite your references (with regards to performance) if you want it taken as-is |
| 02:17.10 | Epyon | No, it's just that I don't like to waste time on non-productive things like : a) arguing, b) profiling tested code |
| 02:17.26 | brlcad | i've done the performance tests, written papers on the topic -- plenty of people use utterly horrible OO'd math just for the hell of it (i.e. for the syntactic sugar) |
| 02:17.41 | brlcad | you haven't tested that code |
| 02:17.43 | Lukstr | profiling is an important part of developing |
| 02:17.47 | Lukstr | and it never hurts |
| 02:18.08 | brlcad | just because someone else uses a similar source doesn't mean it equates OR that they've tested it either |
| 02:18.15 | brlcad | and it's seriously not a big deal |
| 02:18.46 | brlcad | you should be good at it and doing it *anyways* |
| 02:20.12 | Epyon | I went through all (Irrlicht, WM, CS and OGRE) to find the best way to implement Vectors in an OO style. I know how they do it, and I know the reasoning behind their decisions. I COULD profile the Vector code thoroughly, but that would take time -- time that I'd rather spend implementing it. Is that a bad reasoning? |
| 02:20.31 | brlcad | uh, yes it is |
| 02:20.35 | Lukstr | Profiling takes not much time Epyon |
| 02:20.42 | Epyon | And the whole point of OO is that you can always change the underlying code |
| 02:20.51 | brlcad | vector classes don't change |
| 02:21.06 | brlcad | it's a level of abstraction that is very often not even needed or beneficial |
| 02:21.43 | Epyon | brlcad, then I ask again -- why do all the 3d engines do it anyway? |
| 02:21.46 | brlcad | new devs have to waste time consulting the classes to see what someone named a given operator |
| 02:22.44 | brlcad | Epyon: lots of people do things they shouldn't or actually do profile and find that for their context it's acceptable |
| 02:23.20 | brlcad | for your case, you've not done your part other than following suit to a similar "idea" |
| 02:23.28 | brlcad | profiling is not hard |
| 02:23.32 | brlcad | stop making it out to be |
| 02:24.14 | *** join/#bzflag JeffM (n=JeffM@adsl-76-243-140-190.dsl.lsan03.sbcglobal.net) |
| 02:25.12 | brlcad | there's also a huge discrepancy in the various ways you can encapsulate the math and how it impacts performance, *especially* for a non-templatized class structure |
| 02:25.34 | JeffM | did not realise how hard it was to find long sleve undershirts in march |
| 02:25.52 | Lukstr | do a search for gprof and you should realize it's fairly trivial to profile code |
| 02:26.06 | brlcad | seriously, this is a big deal to me -- if you don't profile, I will continue to bitch about it and/or will revert it |
| 02:26.21 | JeffM | oh he'll do it |
| 02:26.26 | JeffM | trust me |
| 02:27.03 | brlcad | just spent a half hour arguing why he shouldn't |
| 02:27.10 | brlcad | coulda had the profile done by then |
| 02:28.30 | *** join/#bzflag Lukstr (n=Lukstr@206-248-157-113.dsl.teksavvy.com) |
| 02:29.31 | jwir3 | night all |
| 02:29.55 | CIA-16 | BZFlag: 03chaos-dev * r16807 10/trunk/bzflag/ (6 files in 2 dirs): Removing vector classes -- not worth the time |
| 02:31.00 | JeffM | what was wrong with them? |
| 02:32.04 | Epyon | No reason to keep them if no one will use them. |
| 02:32.49 | JeffM | what was wrong with them that noone would use them and that brlcad wanted them profiled? |
| 02:33.21 | Epyon | "[03:01] * brlcad is not happy with the vector math classes" |
| 02:33.34 | JeffM | could you be anymore general then that? |
| 02:34.07 | spldart | You'll probably wanna read up on what transpired when you were gone jeff.. quite a bit of convo |
| 02:34.16 | Epyon | Yes : "vector classes are bad" |
| 02:34.24 | JeffM | spldart, I wanted a summary |
| 02:34.27 | brlcad | that's not what I said |
| 02:34.30 | JeffM | dosn't look like I'll get one |
| 02:34.45 | JeffM | someone is bitter |
| 02:34.49 | spldart | heh.. it was a bit, not heated, but warm |
| 02:35.03 | JeffM | can happen, the nature of discussions |
| 02:35.15 | JeffM | the goal no mater what is to move forward |
| 02:35.16 | *** join/#bzflag Erickj92 (n=erickj92@204.73.103.253) |
| 02:35.22 | Erickj92 | how do you jump while in mid air? |
| 02:35.34 | JeffM | Erickj92, a flag called wings |
| 02:35.36 | AHA | you need wings flag to do so |
| 02:35.42 | Erickj92 | ohh i see |
| 02:35.45 | Erickj92 | thanks |
| 02:35.45 | AHA | otherwise the person was cheating :) |
| 02:35.50 | Epyon | JeffM, I though so too. |
| 02:35.58 | Epyon | thought* |
| 02:36.09 | JeffM | Epyon, so how do you move forward |
| 02:36.36 | JeffM | somehow I don't think brlcad just blindly said "pull them" |
| 02:36.38 | brlcad | Epyon: my primary complaint is that there is zero impact analysis on what will otherwise be code that will spread throughout bz in a manner that is not easily quantified after the fact, so it deserves a profile; and you don't want to do the profile |
| 02:36.53 | JeffM | that makes sense |
| 02:37.17 | Epyon | I didn't say I didn't want to do the profile. I said that currently that wasn't my main concern. |
| 02:38.11 | brlcad | it's not your concern, so you're not going to do it basically .. otherwise, why complain about it and all excuses upon excuses as to why you think you don't need to |
| 02:38.17 | JeffM | apperanlty there is an I in team |
| 02:38.18 | brlcad | you know you wouldn't do it |
| 02:39.10 | Epyon | brlcad, ok, maybe. But that's more the reason to pull it off. |
| 02:39.33 | CBG | there are Is in life, time and priorities too. |
| 02:40.07 | brlcad | feels like trying to make you commit for the first time |
| 02:40.17 | brlcad | profiling seriously should be second nature task |
| 02:40.37 | Epyon | brlcad, this part of the conversation is what threw me off balance : |
| 02:40.38 | brlcad | it's like knowing how to use a debugger, one of those core developer requirements |
| 02:40.39 | Epyon | [03:09] <Epyon> brlcad, I can profile cfvec3 versus Vector3D if you want, but Vector3D was built around the currentindustry standard :) |
| 02:40.39 | Epyon | [03:09] <Epyon> (well, except ASMing some of the more complex operations as seen in OGRE and Irrlicht) |
| 02:40.39 | Epyon | [03:09] <@brlcad> or I swear I'll just rip them both out, that's the sort of encapsulation creep that is reallycounterproductive |
| 02:41.03 | brlcad | because you were already arguing strawman by that point |
| 02:41.04 | CBG | is this GSoC? |
| 02:41.23 | JeffM | CBG, indirectly |
| 02:41.24 | brlcad | saying "other people encapsulate, so why isn't mine okay" is bogus |
| 02:41.39 | JeffM | and the one trepan did, is as far as you know, not profiled |
| 02:41.43 | JeffM | I belive that is his only point |
| 02:41.53 | Epyon | Nope, you miss my point completely. |
| 02:42.25 | Epyon | I didn't argue with the fact that you wanted that code profiled. You were arguing about the idea of OO encapsulation itself. |
| 02:42.37 | spldart | There is something very reminiscent about this whole topic from a while back |
| 02:43.01 | Epyon | Or at least that's what I thought you were arguing about. |
| 02:43.14 | brlcad | that was *why* I was arguing that it needs to be profiled |
| 02:43.32 | JeffM | OO has cost |
| 02:43.35 | JeffM | that's for sure |
| 02:43.37 | brlcad | if a profile comes up clean, the hallelujah |
| 02:43.44 | JeffM | you have to be sure the cost is worth it |
| 02:43.53 | brlcad | exactly |
| 02:44.04 | Epyon | brlcad, it wont. OO code *will* be slower. |
| 02:44.23 | brlcad | not necessarily, and even still -- how *much* slower |
| 02:44.24 | JeffM | Epyon, but the question is how much |
| 02:44.45 | JeffM | a small slowdown is worth clean OO, a large slowdown is not |
| 02:44.51 | JeffM | that is why you profile |
| 02:44.54 | JeffM | so you know what to expect |
| 02:45.04 | Epyon | I can profile the functions themselves, but how can you profile the impact of the on BZF itself? |
| 02:45.17 | JeffM | Epyon, by profiling a sample implementation |
| 02:45.28 | JeffM | and making an educated INFORMED postulation |
| 02:45.31 | Epyon | JeffM, and where do I find that sample implementation? |
| 02:45.47 | JeffM | Epyon, you write it |
| 02:46.02 | JeffM | port part of BZ |
| 02:46.04 | JeffM | useit |
| 02:46.10 | JeffM | compre it to a profile with out it |
| 02:46.12 | JeffM | BINGO |
| 02:46.26 | Epyon | Yup. How much would that take? |
| 02:46.46 | JeffM | how much what? |
| 02:46.52 | JeffM | this is basic dev stuff man |
| 02:46.53 | Epyon | Time. |
| 02:46.54 | brlcad | and if anything, you profile before then hook it in then profile after -- check the profile results, check the framerate differences if any |
| 02:46.59 | Lukstr | profiling is not much more than compiling with different flags, for the most part |
| 02:47.26 | JeffM | it is taking a critical look at the actual execution of the code |
| 02:47.34 | Epyon | Lukstr, is wriing a port part of BZ compiling with different flags? :P |
| 02:48.05 | JeffM | Epyon, why can't you just ADMIT you don't want to do it |
| 02:48.07 | Lukstr | Not what I meant; to profile a chunk of code |
| 02:48.08 | JeffM | cus you never have |
| 02:48.10 | JeffM | it's simple |
| 02:48.14 | brlcad | wiring into every instance of cfvec2 and cfvec3 is pretty straightforward |
| 02:48.32 | JeffM | that'd get you at least a basic idea |
| 02:48.37 | Lukstr | you could just profile all of freaking BZ and then grep for the vector calls |
| 02:48.43 | CBG | Epyon: are you getting paid to do this? |
| 02:48.50 | JeffM | CBG, no |
| 02:48.51 | brlcad | and picking another piece of code that does the manually, convert that to the class |
| 02:48.54 | Epyon | CBG, no |
| 02:48.57 | brlcad | isolate it, and test it |
| 02:48.59 | JeffM | this is part of his application |
| 02:49.01 | CBG | then he doesn't have tp put up with the insults |
| 02:49.14 | Epyon | CBG, are you getting the same idea that I do? |
| 02:49.49 | CBG | that these guys would rather put their nuts in a blender than be vaguely nice to you? yes |
| 02:50.06 | Lukstr | erm... |
| 02:50.13 | a_meteorite | sounds painful |
| 02:51.13 | JeffM | wonders where the insuts are |
| 02:51.35 | JeffM | it just looks like a critical discussion of code, something that can and will happen |
| 02:51.36 | CBG | riiiight |
| 02:51.44 | JeffM | it is |
| 02:52.03 | JeffM | perhaps things were said before I joined tho |
| 02:52.05 | JeffM | so I don't kow |
| 02:52.33 | CBG | Insinuating that Epyon is not smart enough to do "basic dev stuff"... |
| 02:52.57 | JeffM | I never said that I was simply saything that this IS common dev stuff |
| 02:52.58 | spldart | smart? I thought it might be experience or desire that was in question |
| 02:53.10 | JeffM | he is very smart |
| 02:53.13 | JeffM | probably smarter then I |
| 02:53.22 | CBG | I won't argue that. |
| 02:53.25 | Epyon | CBG, it's okay, I already decided to pull my application for the core work. They'll find someone better that can iherently profile anyway. |
| 02:53.40 | JeffM | Epyon, no need to pull anything |
| 02:53.52 | JeffM | it's just a critical discussion, it happens |
| 02:54.18 | JeffM | I mean we had a similar discussion about the rules language did we not? |
| 02:54.55 | Epyon | JeffM, I can just envision the loads of arguments that I'll have to go through about each change that I wanted to make. And to be honest, I don't like such arguments more often that sometimes. |
| 02:55.21 | JeffM | Epyon, it is code that is closer to our hearts, so yes there will probalby be more |
| 02:55.38 | JeffM | that is why I wanted to impress upon you in the past days that the project you are looking at involves working with more people |
| 02:55.41 | *** join/#bzflag Lukstr (n=Lukstr@206-248-157-113.dsl.teksavvy.com) |
| 02:55.42 | JeffM | unlike last year |
| 02:55.52 | JeffM | so yes it will be different |
| 02:55.57 | Epyon | JeffM, and I can already envision each such argument reduced to personal attacks of the "you're to lazy to do that, admit it!", "you just want it your way!" kind. |
| 02:56.08 | JeffM | Epyon, I did not say you were lazy |
| 02:56.12 | Epyon | Which would be more or less true. |
| 02:56.25 | JeffM | I just see you fighting it so hard, that it seems like your covering something |
| 02:56.29 | JeffM | for prides sake |
| 02:56.40 | JeffM | and I just wanted to say, that it's ok to say you have not done something before |
| 02:56.57 | Epyon | I was fighting for the idea of OO vectors, not profiling. |
| 02:57.01 | JeffM | if that's not the case, then I'm sorry you took it as an insult, it was not intended as such |
| 02:57.14 | JeffM | Epyon, yes and I think that is where the misunderstanding happend |
| 02:57.26 | JeffM | I don't belive brlcad is saying no OO |
| 02:57.31 | JeffM | he was just saying make sure it's worth it |
| 02:57.33 | JeffM | that is all |
| 02:57.40 | Alesis-Novik | good night everyone |
| 02:57.43 | Epyon | If brlcad wrote "do the profiling then" after the quoted text I'd probably be browsing now through gprof manual. |
| 02:58.31 | JeffM | well he may have been frustrated at the length of conversation but I do not belive he wishes you any illwill |
| 02:58.41 | Epyon | But then suddenly after the OO discussion both of you attack me with "you're too lazy to do the profiling!" |
| 02:58.49 | JeffM | I did not say that |
| 02:59.08 | Epyon | [03:48] <JeffM> Epyon, why can't you just ADMIT you don't want to do it |
| 02:59.08 | Epyon | [03:48] <Lukstr> Not what I meant; to profile a chunk of code |
| 02:59.08 | Epyon | [03:48] <JeffM> cus you never have |
| 02:59.08 | Epyon | [03:48] <JeffM> it's simple |
| 02:59.20 | JeffM | that does not say that you are lasy |
| 02:59.27 | Epyon | It implies :) |
| 02:59.29 | JeffM | just that you are shying away from it due to inexperience |
| 02:59.32 | CBG | It's no compliment either. |
| 02:59.43 | JeffM | CBG, not all things have to be |
| 02:59.44 | spldart | Your not helping CBG |
| 02:59.52 | JeffM | there are levels inbtween complement and insult |
| 02:59.56 | brlcad | CBG: I never insinuated anything, I say what I mean |
| 03:00.16 | JeffM | Epyon, you've basicly said you don't know the profiler, t hat's fine |
| 03:00.31 | JeffM | heck, I odn't k now the linux one ether ;) |
| 03:00.43 | CBG | splI think Epwould disagree |
| 03:00.43 | JeffM | the point of contention was what seemed like reulctance |
| 03:01.35 | JeffM | if I incorretly attributted it to inexperince with the tools, I appologise |
| 03:01.49 | JeffM | just as you incoorectly thought I was calling you lazy |
| 03:02.05 | JeffM | I can assure you lazyness on your part never came into my mind |
| 03:03.03 | brlcad | Epyon: eehh.. you must have selective memory -- the *very* next statement after what you quoted was "do the profile" |
| 03:03.10 | brlcad | 22:10 <@brlcad> don't give me industry standard strawmen, do the profiles |
| 03:04.13 | Epyon | That was a very nice way to put it, y'know. |
| 03:05.30 | Epyon | I actually wanted to end that discussion as fast as possible, to get back to the next class I was writing, not to waste time. Hence my tries to wrap it up quickly. |
| 03:06.10 | brlcad | i.e. you would have been perfectly happy to end the discussion if you'd just have you way? |
| 03:06.34 | brlcad | exceptionally antisocial from my perspective because, as I mentioned, this *is* an issue worth doing the checks on for me |
| 03:06.43 | brlcad | it's not a waste of time |
| 03:06.50 | brlcad | arguing about it this long is a waste of time |
| 03:07.45 | Epyon | Yes -- however, as grep showed me, cfvec* are used widely in libobstacle, and I know how much stress goes on there, so I would think three times and consult you guys before touching it. I first planned to use the vector classes in the non-critical parts of code, then grep showed me the extent of them. |
| 03:08.13 | brlcad | if any of the other core devs have an honest complaint about anything I write, it's *my* responsibility to appease their concerns, do the extra work, justify the course of action, or change what I'm doing |
| 03:08.31 | Epyon | libobstacle wouldn't be something I'd touch without some performance measures. |
| 03:09.34 | brlcad | how sure are you about that? |
| 03:10.04 | Epyon | Absolutely. I'd be scared as hell to break something there that'd reduce the performance to a visible margin. |
| 03:10.17 | Epyon | by a* |
| 03:10.17 | brlcad | no, i mean that it's only in obstacle now |
| 03:10.33 | Epyon | Not only. |
| 03:10.35 | brlcad | if that's the case, you're right -- that is even more reason to profile |
| 03:10.38 | Epyon | It was a figure of speech. |
| 03:10.39 | brlcad | that's into the game sim calcs |
| 03:10.56 | Epyon | Darn it. |
| 03:10.58 | *** join/#bzflag a_meteor (n=a_meteor@adsl-75-15-113-85.dsl.snlo01.sbcglobal.net) |
| 03:11.08 | Epyon | I suggest creating an INGAME profiler. |
| 03:11.24 | Epyon | Apart from normal profiling that is. |
| 03:11.40 | Epyon | Read a nice article about ingame profilers once. |
| 03:11.52 | brlcad | don't think anyone would object, we've done some for that in the past |
| 03:12.10 | brlcad | it's pretty easy to inject a timer around a function too |
| 03:12.38 | Epyon | Using macros that can be turned off at release time. |
| 03:13.37 | Lukstr | hm, if you're going to do that (macros) what's the benefit over plain old profiling? |
| 03:13.58 | Epyon | Here we go again... |
| 03:14.07 | Epyon | Ah nvm. |
| 03:14.24 | Epyon | goes for a smoke. |
| 03:19.58 | Lukstr | night all. Expect application updates sometime tomorrow morning |
| 03:20.22 | *** join/#bzflag a_meteor_ (n=a_meteor@adsl-75-6-2-42.dsl.snlo01.sbcglobal.net) |
| 03:23.20 | Epyon | is back. |
| 03:23.51 | Winny | ff 2.0.0.13 is out! |
| 03:28.53 | Epyon | brlcad, so, if I were to do what I wanted, what would be the next step? (just curious) |
| 03:31.02 | brlcad | i'm not sure exactly what you wanted, but the next step if I were doing what you're doing would probably be to have your unmodified checkout and your modified checkout .. |
| 03:31.53 | brlcad | in the modified checkout, there's two things that need to be checked -- replacing vectors.h and replacing existing code where the math is done directly |
| 03:33.16 | brlcad | so replacing vectors.h is the easy one, you replace, compile unmodified and modified -- sanity check on the framerates, then recompile both with profiling (e.g. gprof, oprofile, valgrind, whatever) -- review/compare the profile results |
| 03:33.35 | *** join/#bzflag JeffM (n=Jeff@pool-71-109-218-124.lsanca.dsl-w.verizon.net) |
| 03:33.43 | Epyon | replacing vectors.h is not as easy as it sounds. |
| 03:34.16 | Epyon | Not if one doesn't want surprises later. |
| 03:34.38 | brlcad | for replacing a snippet, you can characterize that easily with a simple test.c or snippet from bz -- wrap it in a timer, recompile, look at before and after time results |
| 03:34.43 | *** join/#bzflag short_circuit (n=james@c-98-197-19-35.hsd1.tx.comcast.net) |
| 03:34.58 | brlcad | what is not easy about it? |
| 03:35.10 | Epyon | 10000 snippets :P |
| 03:35.13 | brlcad | grep shows every instance of it in use, you replace them all with your class |
| 03:35.39 | Epyon | Also it's not just a simple substitution. |
| 03:36.08 | Epyon | The vectors.h classes are very poor -- they don't have even the basic operators. |
| 03:36.18 | Epyon | There would be no gain in just substituting them. |
| 03:36.20 | brlcad | well it very well might be tedius (especially if it can't be regexpd, which would be surprising), but it's still straightforward |
| 03:37.13 | Epyon | There is no reason to substitute cfvec's for Vectors, if we don't let the code take advantage of why Vectors are nicer than cfvecs. |
| 03:38.00 | Epyon | That was my main intention here. Not substituting vectors.h with my classes, but cleaning up the code to be much more maintainable and easier to debug. |
| 03:38.45 | brlcad | there's also no reason to have two vector classes |
| 03:38.59 | brlcad | that would need to be one of the first things to consolidate regardless |
| 03:39.11 | brlcad | one of the two would need to go just for maintenance |
| 03:39.35 | Epyon | What do you mean? |
| 03:40.27 | brlcad | i mean you're duplicating functionality effectively, which includes additional functionality, but it still also duplicates |
| 03:40.37 | brlcad | there's no point to maintain two vector classes |
| 03:41.00 | Epyon | There is no. But you can't just substitute one for the other regardless. |
| 03:42.08 | brlcad | the same reason why you don't want to bother updating the 10000 snippets, we'd end up with 10000 snippets and 10000 new ones in different places in the code, entropy increased .. complexity increased, maintainability burden and it's just inconsistent |
| 03:42.46 | brlcad | syntacticly you maybe can't just substitute one for the other, but functionally they do the same thing |
| 03:42.47 | L4m3r | are we talking about a geometric vector, or a container? |
| 03:42.52 | Epyon | The point is I want to bother. |
| 03:42.58 | Epyon | L4m3r geo |
| 03:42.59 | JeffM | geometric L4m3r |
| 03:43.04 | L4m3r | just checking |
| 03:43.15 | brlcad | Epyon: that's good to hear |
| 03:43.25 | Epyon | At least I wanted :P |
| 03:43.26 | brlcad | sounds like you have 10000 snippets in front of you then ;) |
| 03:43.30 | Epyon | Now I'm not sure :P |
| 03:43.48 | brlcad | there really should be a way to regexp replace them |
| 03:44.00 | brlcad | at least the vast majority |
| 03:44.04 | Epyon | brlcad, that will lead to errors |
| 03:44.35 | Epyon | And the other reason against is that by doing it by hand you can at the same time make use of the added functionality. |
| 03:44.59 | Epyon | Luckily the vast majority of bugs that would appear that way will be immidately visible. |
| 03:45.15 | Epyon | At least as far where I saw the cfvecs used. |
| 03:46.25 | brlcad | you have to be careful with big replacements, but it can be done error free -- it's certainly easy enough to try and see if there are easy replacement patterns |
| 03:46.39 | Epyon | The problem with regexping them is the fact that the cfvec classes use an old paradigm of having separate functions do the vector ops, not the classes operators themself. |
| 03:47.19 | Epyon | If I would do it, I'd do it using automation but at a filescope. |
| 03:47.58 | Epyon | This way also the small commit rule would be preserved :P |
| 03:48.44 | JeffM | I whink think that it could be smartly automated |
| 03:49.08 | JeffM | how many calls to them are there? |
| 03:49.40 | brlcad | Epyon: also, to test performance, you don't want activity on head interfereing so you'd either not commit piecewise or make a branch |
| 03:50.38 | brlcad | commits are changesets, so they can/should be small in general or they can be 'big' if they are simple (e.g. *just* replacing one class with another) |
| 03:51.08 | brlcad | swapping over the old font system to the new font system touched hundreds of places in the code, but was one commit |
| 03:51.10 | Epyon | However as I stated before -- the idea is not to simply replace it, but to clean the code while doing so. |
| 03:51.17 | brlcad | the branch had hundreds of smaller steps |
| 03:52.06 | JeffM | there are less then 100 calls to the mathf unctions |
| 03:52.10 | brlcad | to do it incrementally though, "cleaning up" should be limited to _just_ the cfvec data |
| 03:52.48 | Epyon | That's the idea. |
| 03:53.13 | Epyon | However the problem is that the code now indiscriminately jumps between cfvec and float* |
| 03:53.54 | brlcad | dude |
| 03:54.10 | JeffM | so you make your new class have a float* autocast |
| 03:54.10 | brlcad | i just checked the code |
| 03:54.10 | JeffM | and convert as it goes |
| 03:54.10 | brlcad | there are only like 80 instances |
| 03:54.18 | brlcad | not 10000 |
| 03:54.18 | brlcad | you could be half-done by now |
| 03:54.24 | brlcad | that's baerly even worth automating |
| 03:54.26 | Epyon | count vec3add/sub etc |
| 03:54.35 | JeffM | Epyon: we did |
| 03:54.49 | JeffM | there are 2 instances of vec3add |
| 03:55.06 | JeffM | 32 instances of vec3Sub |
| 03:55.15 | Epyon | JeffM, ...and a lot of adding vectors by hand... |
| 03:55.16 | brlcad | 200 instances of all the vec3* |
| 03:55.24 | brlcad | that's still not a whole lot |
| 03:55.54 | *** join/#bzflag ChanServ (ChanServ@services.) |
| 03:55.54 | *** join/#bzflag FalseProphet (n=Gigi@unaffiliated/emonemo) [NETSPLIT VICTIM] |
| 03:55.54 | *** join/#bzflag pacman87 (n=Timothy@resnet-45-192.dorm.utexas.edu) [NETSPLIT VICTIM] |
| 03:55.54 | *** join/#bzflag blast007 (i=blast@pdpc/supporter/active/blast007) [NETSPLIT VICTIM] |
| 03:55.54 | *** join/#bzflag Constitution (n=const@unaffiliated/macsforme) [NETSPLIT VICTIM] |
| 03:55.54 | *** mode/#bzflag [+ov ChanServ blast007] by irc.freenode.net |
| 03:55.56 | Epyon | Don't tell me no one noticed that that 10000 was for marketing purposes D: |
| 03:56.08 | brlcad | no, i actually presumed you were being literal |
| 03:56.13 | brlcad | as that very well could be the case |
| 03:56.15 | Epyon | brlcad, uh, sorry |
| 03:56.20 | brlcad | there's more than 200k lines of code |
| 03:56.20 | JeffM | beeing a serious discussion and all |
| 03:56.49 | brlcad | exaggeration doesn't help regardless, especially two orders of magnitude |
| 03:57.13 | Epyon | *sigh* I admit |
| 03:58.47 | Epyon | What doesn't change a fact that there are countless places where the code readability would benefit from a Vector class where the calculations are done by hand. |
| 03:59.18 | *** join/#bzflag ViciousM (n=max@cpe-024-211-248-161.nc.res.rr.com) |
| 03:59.24 | JeffM | that would be part of the eventual refactoring |
| 04:00.09 | JeffM | once you are sure that the vector class is good |
| 04:03.05 | Epyon | One last question... |
| 04:03.45 | *** join/#bzflag think_tank (n=think_ta@c-71-194-147-60.hsd1.il.comcast.net) |
| 04:03.55 | Epyon | When trepan originally submitted this file around Sep 2 2004, did anyone ask him to do profiling of the vector ops? |
| 04:04.02 | Epyon | Rethorical question. |
| 04:04.12 | JeffM | he profiled most of his stuff |
| 04:04.40 | JeffM | his profiling is the reason we have an octree |
| 04:04.41 | Epyon | So... where are the results :) |
| 04:05.00 | JeffM | they were discussed, but I doubt kept |
| 04:05.38 | brlcad | profiling is rarely kept, you do it periodically when it matters and sanity check from time to time |
| 04:05.45 | JeffM | I recall him discussing much of his findings. |
| 04:05.54 | JeffM | others have as well |
| 04:05.58 | JeffM | I know blast has done some |
| 04:06.05 | JeffM | and DTRemenak as well IIRC |
| 04:06.08 | brlcad | we also profiled the entire opengl call stack at one point |
| 04:06.35 | Epyon | Well I'd like to use that part of BZF he used for profiling that's why I asked :) |
| 04:06.38 | brlcad | since the framerate dominates, you do have to isolate-time things like the math ops |
| 04:07.28 | Epyon | The one JeffM told me to write ^_^ |
| 04:07.32 | Erickj92 | how do i drop a flag? |
| 04:07.39 | JeffM | Erickj92: middle mouse |
| 04:07.40 | brlcad | usually it's just writing up an isolated piece of code with timers (a test app), or putting timers into the mainline code, or profiling -- depends on what's being measured |
| 04:07.44 | Erickj92 | ty |
| 04:07.54 | brlcad | or spacebar |
| 04:07.54 | JeffM | Epyon: I meant you'd write it by changing the existing calls |
| 04:07.57 | blast007 | Erickj92: also, by default, the spacebar |
| 04:07.59 | JeffM | not some new app |
| 04:08.05 | blast007 | brlcad: bah, too slow |
| 04:08.08 | Erickj92 | thank you |
| 04:08.12 | brlcad | :) |
| 04:08.29 | JeffM | you'd go and get timings before, then do your changes, and get timings after |
| 04:08.46 | JeffM | even if it's only in some of the cases ( the less then 100 ) that will get you a baseline |
| 04:10.09 | brlcad | it takes about ten lines of code to wrap something in a time |
| 04:10.15 | brlcad | *timer |
| 04:13.22 | Epyon | Sooo.... |
| 04:14.51 | *** join/#bzflag AAA_awright (n=AAA_awri@wsip-68-14-251-102.ph.ph.cox.net) |
| 04:14.54 | Epyon | Take a few snippets like this one : http://pastebin.bzflag.bz/d6a3099d : three versions, one original, one with cfvec3 and one with Vector3 and give the results for a 10000 passes, right? |
| 04:15.19 | Epyon | That will still be valid only for the platform/compiler it was compiled on. |
| 04:16.00 | JeffM | that dosn't realy profile BZs use of the code |
| 04:16.08 | JeffM | I belive we are all saying profile it IN bz |
| 04:17.06 | Epyon | That's the part I don't get. How can you profile the results of a task if you didn't do the task itself? |
| 04:17.25 | JeffM | you do parts of the task |
| 04:17.42 | Epyon | Which parts? |
| 04:17.51 | JeffM | the 100 or so bits that use it |
| 04:18.02 | JeffM | that's only part of the task |
| 04:18.12 | JeffM | since your final task is to replace all the 'manual' math right? |
| 04:18.16 | JeffM | or do like 20 of em |
| 04:18.17 | Epyon | Yes. |
| 04:18.41 | JeffM | if you do your profiling on the unmodified code first, you can see what calls the existinc doe most |
| 04:18.46 | JeffM | and just do those for now |
| 04:18.48 | JeffM | and get a baseline |
| 04:19.46 | brlcad | yeah, if you're scope/goal is that big -- there's several things that should be tested, can't base the whole change on a point-sample (not saying you were, just saying) |
| 04:19.48 | Epyon | To do that, I'd need to get a LOT better understanding of the code than that which I have now. |
| 04:20.10 | JeffM | Epyon: I would agree with you there |
| 04:20.20 | brlcad | you could still start with the cfvec3 before and after |
| 04:20.23 | Epyon | And to be honest "Replace vector math seems like a GSoC-size app" :P |
| 04:20.29 | JeffM | naw |
| 04:20.30 | brlcad | it's really not |
| 04:20.36 | JeffM | couple days |
| 04:20.42 | brlcad | maybe a couple *jinx* |
| 04:21.46 | *** join/#bzflag BulletCatcher (n=bc@unaffiliated/bulletcatcher) |
| 04:26.07 | JeffM | go go power rangers! |
| 04:26.46 | Epyon | The point is -- either I would have to somehow proove that Vector*D classes are fine, or profile every bit of code that I want them to be put in. |
| 04:27.38 | JeffM | or show that they don't affect the performance in a signifigant way, when used by BZ |
| 04:27.46 | brlcad | not every bit, just a good faith effort to make sure it's good |
| 04:28.25 | brlcad | a few profiles and testing some sections of code that might matter |
| 04:29.28 | Epyon | What I don't understand, is why a test of the Vector class itself isn't good enough. |
| 04:29.44 | Epyon | After all it we can't profile it for every usage |
| 04:30.16 | JeffM | because, lets say you find out that it is 10% slower |
| 04:30.21 | JeffM | just as an example |
| 04:30.30 | JeffM | 10% called once is ok |
| 04:30.37 | JeffM | 10% called 1000000 times is not |
| 04:31.08 | JeffM | so unless you do some form of larger scale testing you can't tell how the testing of the small bit, will affect the whole |
| 04:31.40 | JeffM | you've already admited that you KNOW it will be slower |
| 04:31.47 | Epyon | I think that I could prove that most of the time it's the same speed. |
| 04:31.59 | JeffM | but some of the times it WILL be slower |
| 04:32.06 | Epyon | It is only the things that I want to use extra that might be slower. |
| 04:32.07 | JeffM | cus I doubt it will be faster |
| 04:32.16 | Epyon | None of the things in the code as it would stand. |
| 04:32.24 | Epyon | inlining is inlining after all. |
| 04:32.43 | Epyon | The asm code generated is equivalent. |
| 04:32.44 | JeffM | it's not so much a "proof" this isn't achedemic |
| 04:33.24 | JeffM | it's a good faith research to be sure that it's not joing to impose undue slowdowns |
| 04:33.40 | Epyon | However, as far as I can show that the asm code is the same in case of the compiler I'm using, I cannot guarantee that it's the same on every compiler. |
| 04:33.47 | JeffM | if you are confident that it will be jsut as fast then doing a simple profile should not be too hard |
| 04:34.17 | Epyon | I know that MSVC and GCC handle proper inlining good, but I can't say that for the others :( |
| 04:34.36 | brlcad | and you can't just base it on something like fps, as we're not cpu bound -- increasing the sim time will mostly just cull away players that were playing perfectly fine |
| 04:35.10 | brlcad | Epyon: and knowing that, whether it's good or not matters -- that's the practical nature |
| 04:35.12 | Epyon | The more you talk, the less I understand what I'm supposed to do -_- |
| 04:35.32 | Epyon | At this particular moment I havn't got a clue left at all. |
| 04:35.35 | brlcad | it could be golden in msvc, but then if it killed performance on gcc, it'd still be a non-starter |
| 04:36.06 | JeffM | well then perhaps this task is not well suted to your current level of coding experince. |
| 04:36.23 | purple_cow | what's this task? |
| 04:36.34 | brlcad | do two things, replace vectors.h (only, no peripheral damage) and write a standalone test snippet |
| 04:36.41 | Epyon | BTW, the Vector*D and cfvec* classes are syntatically the same, so this is just a matter of testing the things that are to be refactored. |
| 04:36.44 | JeffM | profiling vector math implementations purple_cow |
| 04:37.23 | Epyon | Just checked it - a simple rename allowed "typedef Vector3D cfvec3" |
| 04:37.23 | brlcad | Epyon: they're functionally the same, the definitions aren't iirc |
| 04:37.51 | brlcad | if that works, then the regexp replace that you said wouldn't work will work :P |
| 04:38.04 | Epyon | Well it will. |
| 04:38.19 | purple_cow | Step 1: Add IPP |
| 04:38.22 | purple_cow | Step 2: uhh |
| 04:38.29 | Epyon | But that's missing the whole point of introducing extended classes :P |
| 04:38.31 | JeffM | step 3: profit! |
| 04:38.37 | Epyon | IPP? |
| 04:38.51 | purple_cow | http://www.intel.com/cd/software/products/asmo-na/eng/perflib/302910.htm |
| 04:39.03 | Epyon | Ah, there's a difference... but... |
| 04:39.07 | Epyon | He he he |
| 04:39.19 | brlcad | Epyon: the point of doing that isn't merits of the class, it's reduction of redundancy and sanity checking the replacement |
| 04:39.49 | purple_cow | are vector math ops really the slow point? |
| 04:39.56 | purple_cow | CPUs are pretty good at math |
| 04:40.25 | JeffM | purple_cow: not really, it was more of a "if you are going to replace a bunch of stuff everywhere with a thing, make sure the thing is gonna be fine first" |
| 04:40.44 | Epyon | Maybe it's the hour (5:39) but I'm completely lost now. |
| 04:40.54 | Epyon | Guess it's not my thing to do then. |
| 04:41.06 | JeffM | maybe it is |
| 04:41.06 | brlcad | just said the two things to do |
| 04:41.15 | JeffM | just depends on how much effort you want to put into it :) |
| 04:41.43 | Epyon | I'll stick to the BZWGenV2 proposal and TP after all. |
| 04:42.08 | JeffM | what patch you gonna sumbit? |
| 04:42.54 | Epyon | Probably none, as 80% of the applicants. I'm afraid of the main branch like of fire now. |
| 04:43.21 | JeffM | all applicants have to submit something |
| 04:43.36 | Epyon | Wont touch it without a full set of profilers and debuggers running. |
| 04:44.03 | JeffM | get some rest you'll feel better about it tomorow ;) |
| 04:44.05 | Epyon | Might decrease performance by one thousand of a percent during loading time, and that'd be bad. |
| 04:44.15 | brlcad | from exaggeration to dramatic hyperbole? :) |
| 04:44.58 | Epyon | hyperboles make wonders in illustrating a point sometimes ;] |
| 04:45.09 | brlcad | that you're a drama queen? |
| 04:45.15 | brlcad | point taken |
| 04:45.26 | JeffM | yeah I don't think the point you wanted to illustrate is what you illustrated |
| 04:45.29 | JeffM | get some rest man :) |
| 04:45.30 | Epyon | That wasn't the one I was making ;] |
| 04:45.49 | Epyon | Isn't everyone sometimes? :> |
| 04:45.51 | JeffM | yes it was the one you made, not what you wanted to make. |
| 04:46.00 | purple_cow | I'm never a drama queen |
| 04:46.18 | JeffM | purple_cow is always the prom king! |
| 04:46.33 | JeffM | or the cold hard fist of justice! |
| 04:46.42 | brlcad | or a drag queen :) |
| 04:46.42 | JeffM | obey the fist |
| 04:46.43 | Epyon | Guess it's the downside of being in a theatre class in highschool -_- |
| 04:48.03 | Epyon | And to think it all started because such code gives me the shivers : http://pastebin.bzflag.bz/m2f26d131 |
| 04:48.35 | Epyon | And I just wanted to make it "right" :/ |
| 04:48.55 | JeffM | depends on your definition of right |
| 04:49.10 | JeffM | that code is just what the compiler is going todo |
| 04:49.17 | JeffM | so it is very fast |
| 04:49.35 | brlcad | sounds like a personal problem, you can't say that code isn't obvious/clear, pretty easy to maintain, and has no performance issues |
| 04:50.03 | Epyon | tankColor[teamIndex] = tank; is just one line, and shows that "tank" is a bad name for that variable indeed xP |
| 04:50.04 | JeffM | anything you do to it will slow it down, you jsut have to justify it |
| 04:51.02 | brlcad | cept that it's no longer obvious (have to know what = operator is doing, what the types of lhs and rhs are), is a curve for new devs, and potentially has performance impact |
| 04:51.25 | JeffM | the benifits may outweight the drawbaks |
| 04:51.33 | brlcad | entirely depends on the context |
| 04:51.41 | brlcad | and requirements |
| 04:51.47 | Epyon | Bad example that one is just ugly. But in cases of more complex vector operations a good Vector class may be the line between seeing a bug and not seeing it. |
| 04:51.58 | brlcad | it's a great example |
| 04:52.33 | brlcad | it's at the core of what you're trying to change, and why we're saying that there are some sanity checking that needs to happen |
| 04:52.40 | Epyon | I can turn the tables on that one however, and I will gladly do that now. |
| 04:52.53 | Epyon | Do you want Vector classes? |
| 04:53.03 | brlcad | *sigh* |
| 04:53.11 | brlcad | tuning any given case still misses the point |
| 04:53.23 | Epyon | Just answer the question :) |
| 04:53.47 | brlcad | eh |
| 04:53.48 | JeffM | are you going to implment them? or just revert the code you pulled? |
| 04:53.54 | brlcad | "it depends"?? wth |
| 04:54.10 | Epyon | brlcad, so think about it :P |
| 04:54.14 | brlcad | it's not an absolute, the idea was good faith efforts to determine if it's worth it |
| 04:54.43 | brlcad | Epyon: if you're not going to do the work, then the question is moot, aiight |
| 04:55.19 | Epyon | Because otherwise it starts to look wierd -- I want to do them, and others neither want it, nor want to help with it. So why should I do it at all? |
| 04:55.24 | brlcad | so stop wasting time, if you don't want to do it and/or aren't going to do it or are going to keep complaining and second/third questioning every step of the way, then just forget it |
| 04:55.35 | brlcad | seriously, you've dragged this on for two hours now |
| 04:55.41 | Epyon | Not me, you :P |
| 04:55.48 | brlcad | no, you have |
| 04:55.57 | brlcad | when another dev raises a concern you response |
| 04:56.03 | brlcad | you respond |
| 04:56.04 | Epyon | I was working. You stoped me, and started the discussion. |
| 04:56.09 | brlcad | that's *your* responsibility |
| 04:56.13 | brlcad | not the person asking the questions |
| 04:56.22 | brlcad | that's how it works |
| 04:56.39 | brlcad | if you don't like that, then fine -- you certainly don't have to |
| 04:56.41 | Epyon | Now the only thing I want is a straight answer -- do you need it? do you want it? |
| 04:57.00 | JeffM | if you are just talking about those 2 files, we have them |
| 04:57.09 | Epyon | It would be stupid to (again!) make a need for something and the work on it. |
| 04:57.16 | brlcad | if it's going to be a discussion like this the whole way along to get you to profile and test it, then hell no |
| 04:57.39 | brlcad | if it turns out to be a great improvement after you've done the good faith effort testing, then sure |
| 04:58.27 | Epyon | And the point is that I'm trying 2 hours to understand how you want it tested. |
| 04:58.50 | brlcad | i've told you several times, you quickly diverge from how back to why |
| 04:59.06 | brlcad | if you're not going to be convinced of the why, I can't help you |
| 04:59.19 | Epyon | I am, you're not. |
| 04:59.35 | brlcad | you know, this has gone on long enough |
| 04:59.43 | Epyon | Right. |
| 04:59.47 | Epyon | G'night. |
| 04:59.50 | brlcad | read the log if you have to |
| 04:59.54 | brlcad | it's all there |
| 05:06.01 | mibocote | hey, where does the game send password info? |
| 05:06.12 | Constitution | to the list server |
| 05:06.14 | blast007 | mibocote: which password? |
| 05:06.32 | blast007 | global auth password, admin password, or local auth password? |
| 05:06.43 | blast007 | global auth is the one in the Join Server menu |
| 05:06.48 | JeffM | if it's in the password field under your name, it goes to the list |
| 05:06.52 | JeffM | admin goes to the server |
| 05:06.59 | JeffM | local auth goes to a password scammer |
| 05:07.07 | blast007 | hehe |
| 05:07.13 | JeffM | sad but true |
| 05:07.25 | JeffM | there is a reason we don't have it anymore ;) |
| 05:07.54 | Constitution | goes to grep his server logs |
| 05:08.05 | blast007 | if we make the list server easier to set up, there's no need for it |
| 05:08.07 | Constitution | ;) |
| 05:08.29 | JeffM | indeed |
| 05:08.48 | blast007 | we could even do a list server data class that uses SQLite, so little to no config then :) |
| 05:08.49 | L4m3r | I'd rather work on that "stab people in the face" thing |
| 05:08.53 | blast007 | hehe |
| 05:09.04 | JeffM | L4m3r: alt + s |
| 05:09.20 | Epyon | Dammit, read it again, and I couldn't even find an answer to a basic question. |
| 05:09.35 | JeffM | Epyon: get some rest |
| 05:09.50 | Epyon | JeffM, can't I'm too ... hmm... excited :/ |
| 05:10.05 | JeffM | that dosn't mean you still shoundn't try |
| 05:10.10 | JeffM | come back again when you are fresh |
| 05:10.16 | blast007 | Epyon: try counting lines of code, I mean, sheep. |
| 05:10.17 | JeffM | it helps wonders trust me |
| 05:10.27 | mibocote | so the password you enter when joining a game goes to the list server? |
| 05:10.32 | blast007 | mibocote: yup |
| 05:10.43 | JeffM | mibocote: yeah it basicly goes to the fourms |
| 05:11.14 | mibocote | then why is a global authentication daemon listed under the gsoc ideas? |
| 05:11.27 | JeffM | mibocote: cus we want one that isn't based aroudn the fourms |
| 05:11.30 | JeffM | and that has more features |
| 05:11.50 | JeffM | right now if the server the fourms are on, goes down, noone can auth |
| 05:12.00 | JeffM | so we'd like to have a network of servers that can use that |
| 05:12.04 | JeffM | and then tie our fourms into that |
| 05:12.19 | Epyon | JeffM, ok, but promise me you'll help me with that profiling. I have no clue how to do it, and that annoys the hell out of me. I don't care about the app anymore, I just want to know what the hell I was supposed to do. |
| 05:12.19 | JeffM | basicly giving us a network of servers that we own, instead of just one |
| 05:12.45 | JeffM | Epyon: rest and talk about it tomorow |
| 05:13.10 | JeffM | I think we've all had enough of the subject for one day |
| 05:13.38 | JeffM | mibocote: so basicly right now the list uses the fourms, we'd like to make it where the fourms use the list :) |
| 05:13.46 | JeffM | and the list can have redudancy |
| 05:14.18 | mibocote | alright, and the list server tells the game server if the name/pass is valid? |
| 05:14.20 | Epyon | goes on reading man gprof to sleep -_- |
| 05:15.10 | blast007 | mibocote: yeah |
| 05:15.23 | JeffM | mibocote: yes |
| 05:15.31 | JeffM | when you auth as a player, you are sent a token |
| 05:15.39 | JeffM | then you send that token to the game server |
| 05:15.48 | blast007 | the server could call the CHECKTOKEN action to do that. Current that is unused by bzfs itself, as they piggyback the token check onto the ADD message |
| 05:15.58 | JeffM | then the game server sends your name and token to the list, and the list says if you are cool or not |
| 05:15.59 | blast007 | but some web services use the CHECKTOKEN action |
| 05:16.17 | JeffM | yes some web serives work like the game server |
| 05:17.12 | mibocote | where are the messages for the auth defined, they don't seem to be in Protocol.h |
| 05:17.16 | JeffM | the system is designed to keep the password out of the server |
| 05:17.22 | JeffM | mibocote: they arn't |
| 05:17.24 | JeffM | it's all HTTP |
| 05:17.28 | JeffM | goes thru curl |
| 05:17.30 | JeffM | not the netcode |
| 05:17.34 | mibocote | ah, okay |
| 05:17.42 | JeffM | http://my.bzflag.org/db/ |
| 05:17.43 | *** join/#bzflag Epyon_ (n=Epyon@90-156-93-82.magma-net.pl) |
| 05:17.49 | JeffM | it's all calls to that interface |
| 05:18.20 | mibocote | would it be desirable to keep a similar setup, just with distribution/replication? |
| 05:18.36 | JeffM | well that'd be a basic idea |
| 05:18.49 | JeffM | and at some level we have to keep the existing interface for old clients |
| 05:19.02 | JeffM | the current methods discussed were to have the backend converted to LDAP |
| 05:19.11 | JeffM | and then have the lists be ldap nodes |
| 05:19.22 | JeffM | and the same for our other services |
| 05:19.31 | JeffM | like fourms, group manager, wiki, etc.. |
| 05:20.25 | JeffM | personaly I think HTTP makes it easy |
| 05:20.43 | JeffM | and the comunication to/from the list/game is not that complicated that it needs a very deep protocoll |
| 05:20.50 | JeffM | there isn't lots of interaction |
| 05:20.50 | mibocote | yeah |
| 05:20.53 | JeffM | just a request and responce |
| 05:20.57 | JeffM | the big thing is the backend |
| 05:21.09 | JeffM | we want to get away from using the forums as our user management |
| 05:21.16 | JeffM | cus it sucks :) |
| 05:21.26 | JeffM | we also want to have a seperate group manager |
| 05:21.39 | JeffM | and all those things would need an independnet user manager that they can connect to |
| 05:21.51 | JeffM | ldap seemed to be a good fit, as it is standard and well suported |
| 05:22.55 | Constitution | for the patch that applicants are required to submit, how significant of an issue should it address? |
| 05:23.12 | JeffM | Constitution: dosn't need to be anything much |
| 05:23.21 | mibocote | So what the idea entails is a replicated/distributed auth/list server to prevent outages that might use ldap as a backend for user storage and retains the current protocol, leaving the client, essentially, unchanged |
| 05:23.24 | JeffM | just smomething to show you've built and know our standards, etc.. |
| 05:23.33 | Constitution | okay |
| 05:23.39 | JeffM | and "get" how our patches work |
| 05:24.03 | JeffM | mibocote: that is one method, we'd be open to ideas in changing the client if there are advantages |
| 05:24.24 | JeffM | I mean if the client can do a direct ldap auth and that is better for wahtever reason we'd be willing to look into that |
| 05:24.44 | JeffM | there are probably also features we'd like to have that may require changes to the client |
| 05:24.53 | JeffM | like registering from inside the game |
| 05:25.43 | JeffM | or maybe using the list server to track usage data from our users |
| 05:25.54 | JeffM | since I personaly would like to know a number of things that they do :) |
| 05:27.01 | mibocote | what kind of things? |
| 05:27.39 | JeffM | video card, GL renderer, screen size, OS, anon usage data things |
| 05:27.52 | mibocote | ah, ic |
| 05:28.06 | blast007 | I don't think the client should do direct LDAP auth |
| 05:28.12 | JeffM | blast007: ok |
| 05:28.13 | mibocote | probably have to ask them before sending that |
| 05:28.20 | blast007 | just seems less secure to me |
| 05:28.21 | JeffM | I was just sayin we are open to ideas |
| 05:28.31 | blast007 | since LDAP would need to be open to the world then |
| 05:28.32 | JeffM | mibocote: yes have an option to provide anon data |
| 05:28.45 | mibocote | the level of indirection is required for the token option |
| 05:28.46 | JeffM | but I can't ask untill I have a way to collect it ;) |
| 05:28.49 | blast007 | I'd rather have the OpenLDAP server listening only on the local loopback address |
| 05:29.05 | blast007 | and then the services on each server would talk to their local LDAP server |
| 05:30.17 | blast007 | and having the LDAP directly in the client makes it harder to back out of that later, if need be. It also could make changing our implementation more difficult. |
| 05:30.44 | mibocote | blast007, I don't think anyone was seriously considering direct ldap auth |
| 05:31.42 | JeffM | mibocote: blast007 has done a fair amount of research on it so he'll know a lot more then I. in fact he would be ( I'd hope ) the mentor for that project in SoC |
| 05:31.44 | blast007 | mibocote: it was mainly in response to Jeff's comment |
| 05:32.08 | blast007 | brlcad has done more looking into OpenLDAP |
| 05:32.17 | blast007 | I've mainly just hit it with hammers ;) |
| 05:32.17 | JeffM | yeah I'm just talkin out my butt, just tryin to say that I dunno what we'd do, but I'm not dead set on any one method. :) |
| 05:32.27 | JeffM | blast007: yeah but he ain't takin a project |
| 05:32.32 | blast007 | k |
| 05:32.36 | L4m3r | iirc most of blast's work has been abstraction :) |
| 05:32.45 | JeffM | it's more then I know |
| 05:32.45 | blast007 | abstraction ftw :) |
| 05:32.52 | L4m3r | so the underpinnings don't matter a whole lot |
| 05:32.59 | JeffM | I just know how I want it to work form an admins point of view |
| 05:32.59 | L4m3r | not yet anyway |
| 05:33.18 | blast007 | mibocote: yeah, I actually started to rewrite the list server to be a bit more modular, and not be tied to the forum itself |
| 05:33.18 | AAA_awright | An authentication system needs a few different parts: password validation, token validation, server list, and player info. There would be an inner, protected API that BZFlag services would use to authenticate people (LDAP, whatever). |
| 05:33.36 | blast007 | I stalled when I started thinking it might be better in C/C++ than PHP |
| 05:33.37 | AAA_awright | In addition, there would be a frontend system that gives out tokens to clients and authenticate them, for the bzflag client and weblogin.php would both use it. |
| 05:34.35 | AAA_awright | The big question is what can LDAP/another solution do better then MySQL/custom solutions like what we have now. |
| 05:34.58 | blast007 | AAA_awright: the main reason for LDAP is that it is designed for our exact need :) |
| 05:35.24 | blast007 | we have a set of users, groups, and servers |
| 05:35.33 | AAA_awright | For users and password authentication, it is a trivial choice. But what else? |
| 05:35.39 | blast007 | sounds a lot like what LDAP has ;) |
| 05:36.09 | JeffM | does ldap not also provide distributed services? |
| 05:36.16 | blast007 | it does |
| 05:36.23 | AAA_awright | And is it faster? Lower overhead? I hear a custom C server being thrown about, but is it actually better? |
| 05:36.36 | JeffM | and standardised connection to exsting web apps? |
| 05:36.38 | blast007 | AAA_awright: have you worked with LDAP before, or even something like Windows Server's Active Directory (which is somewhat? compatible with LDAP too) |
| 05:37.03 | AAA_awright | I haven't worked with LDAP (I don't pretend to either), but I know a tiny bit |
| 05:37.35 | JeffM | AAA_awright: the problem I fear with custom stuff is we'll burn a lot of time doing custom integrations with other services ( fourums, wiki, etc) for only a marginal speedup |
| 05:37.50 | blast007 | JeffM: if *any* speedup |
| 05:37.57 | JeffM | I'm sure we could code up a system that is wicked fast |
| 05:37.59 | blast007 | the LDAP stuff has been around for a long time |
| 05:38.00 | JeffM | and super optimised |
| 05:38.16 | JeffM | blast007: yes, that means it's robust, not nessiarly fast ;) |
| 05:38.17 | blast007 | you can tune it for syncing over a 56k link if you needed to |
| 05:38.25 | JeffM | irc has been aroudn forever, but it ain't fast |
| 05:38.47 | blast007 | true ;) |
| 05:38.51 | JeffM | the lowered maint costs are a huge benifit to us over something we have to handhold. |
| 05:39.17 | AAA_awright | With a central API in place, it shouldn't be hard. The problem is, what authentication schemes do phpBB, Mediawiki, Drupal, etc support (including custom support)? We are bound to those. |
| 05:39.20 | JeffM | I think it's potential benifits are good enough at least for a proof of concept test to be done |
| 05:39.29 | blast007 | AAA_awright: all of those support LDAP ;) |
| 05:39.42 | JeffM | AAA_awright: we ain't writin our own fourms and wiki and mainpage ;) |
| 05:39.49 | JeffM | and they are going to need auth |
| 05:39.50 | AAA_awright | Because someone programmed a custom authentication handler. |
| 05:39.58 | blast007 | phpBB3 has it built-in, Mediawiki is dropping a single file in, and maybe 10 or 20 lines of a config, and Drupal has a plugin |
| 05:39.59 | JeffM | that's more code for us to maintain |
| 05:40.09 | AAA_awright | True. And a consideration. |
| 05:40.50 | JeffM | honestly, we kinda want to get out of the web app biz and get on with making games ;) |
| 05:41.06 | JeffM | so the more existing code we can leverage the better |
| 05:41.19 | JeffM | even if it means taking a small hit from being perfectly optimial |
| 05:41.34 | JeffM | just the fact that we can spread the load will make the service way better then it is now |
| 05:41.38 | JeffM | computers are cheap :) |
| 05:42.10 | mibocote | also, if you plan to add a global chat function that people can use out of the game, most jabber servers support ldap |
| 05:42.25 | blast007 | mibocote: ah, that's an interesting point there too |
| 05:42.26 | JeffM | a decent point |
| 05:42.40 | blast007 | phpBB3 has something about jabber in it as well |
| 05:43.03 | blast007 | didn't have time to play with that though |
| 05:43.28 | AAA_awright | No, no global chat. Bad idea. Very bad. We have IRC for that. |
| 05:43.43 | blast007 | AAA_awright: I'd like it myself |
| 05:43.53 | blast007 | make it more of a community in-game |
| 05:43.54 | JeffM | indeed |
| 05:44.06 | AAA_awright | There is a community already, look around you. |
| 05:44.08 | JeffM | guesss AAA_awright has never played an MMO |
| 05:44.23 | AAA_awright | Heh, you nailed that one |
| 05:44.32 | blast007 | so instead of just "Find Server" and "Join Game", you could chat with people right there, find out where they were playing, or ask some people to join you at a certain server |
| 05:44.37 | JeffM | if it is done well, it can be very fun |
| 05:44.39 | blast007 | and have a friends-list right there |
| 05:44.40 | JeffM | a lobby would be good |
| 05:44.45 | JeffM | with friends lists |
| 05:44.46 | JeffM | yeah |
| 05:44.48 | AAA_awright | *whimper* |
| 05:44.57 | blast007 | and the ability to invite people to a server, etc |
| 05:45.03 | JeffM | AAA_awright: just cus you don't like it dosn't mean 1) you have to use it, and 2) it's a bad idea |
| 05:45.13 | AAA_awright | We need to plan how to track players first then, I do like the idea of knowing where players are. |
| 05:45.28 | blast007 | AAA_awright: track them with the global chat system ;) |
| 05:45.32 | JeffM | that's where the ldap comes in ;) |
| 05:45.35 | blast007 | if they're on that, they're in the game client |
| 05:45.48 | blast007 | yeah, LDAP too :) |
| 05:46.04 | BulletCatcher | Don't forget GPS units on the tanks. :-) |
| 05:46.57 | BulletCatcher | On a completely different note: treads.png needs to be added to the bzflag_DATA list in data/Makefile.am so it will get installed. |
| 05:47.20 | JeffM | BulletCatcher: we are desuporting linux |
| 05:47.23 | AAA_awright | I'll play with this idea first though, then I will smash it. When would players be connected to the chat? |
| 05:47.53 | JeffM | AAA_awright: depends on the implementaiton, but I'd think if you were set to join the lobby channel, then on join |
| 05:47.58 | blast007 | AAA_awright: 24/7. We would be sponsored by the government. |
| 05:48.07 | AAA_awright | When not connected to a server? After you register? |
| 05:48.12 | AAA_awright | Who would moderate? |
| 05:48.19 | JeffM | your mom |
| 05:48.28 | L4m3r | AAA_awright: not you, so don't worry about it |
| 05:48.40 | blast007 | we could have CAN help out! |
| 05:49.04 | AAA_awright | heh. |
| 05:49.06 | AAA_awright | Actually.... hm... |
| 05:49.21 | CIA-16 | BZFlag: 03JeffM2501 * r16808 10/trunk/bzflag/data/Makefile.am: treads make tanks go vrooom! |
| 05:49.22 | L4m3r | Carefully Appointed Nannies? |
| 05:49.40 | JeffM | L4m3r: no seriously, his mom, she offered |
| 05:50.26 | BulletCatcher | JeffM: thanks |
| 05:50.27 | AAA_awright | jots down note mental note, change project names on takeover |
| 05:50.47 | JeffM | AAA_awright: you could have forked ages ago if you wanted ;) |
| 05:52.36 | AAA_awright | The way I see this proposed global chat is rather silly. People join a server by default, until you choose another server to visit. I don't like the idea of it, your client is doing something unexpected. |
| 05:52.58 | blast007 | expect the unexpected |
| 05:53.07 | blast007 | and really, lots of games have lobbies |
| 05:53.16 | JeffM | AAA_awright: and many have chat channels |
| 05:53.20 | L4m3r | I could see "Lobby |
| 05:53.21 | JeffM | local, global, clan |
| 05:53.26 | L4m3r | " as the first option on the main menu |
| 05:53.51 | blast007 | I'd also like to remove the "Team" option from the Join Game menu ;) |
| 05:53.51 | L4m3r | but that ties into a bunch of UI issues that need to be ironed out |
| 05:54.02 | blast007 | and have it be a dialog after the user joins the server |
| 05:54.03 | JeffM | blast007: yeh that needs to be done |
| 05:54.14 | AAA_awright | I don't think it fits the style of BZFlag. Maybe there could be a Lobby option on a BZFlag server that is also an IRC server (Slightly better, getting there, but very bad, still). |
| 05:54.32 | JeffM | IRC is icky |
| 05:54.48 | noyb | amen. |
| 05:55.03 | AAA_awright | It is fairly straightforward, and forkable. |
| 05:55.13 | *** join/#bzflag JBdiGriz (n=JBdiGriz@adsl-75-36-46-10.dsl.sndg02.sbcglobal.net) |
| 05:55.13 | *** mode/#bzflag [+o JBdiGriz] by ChanServ |
| 05:55.14 | JeffM | and wordy |
| 05:55.45 | L4m3r | I like the idea of a "bridge" system |
| 05:55.48 | AAA_awright | That doesn't mean it is awful. XML is also wordy. |
| 05:56.00 | JeffM | it also dosn't have an auth concept |
| 05:56.02 | JeffM | that's the big thing |
| 05:56.17 | AAA_awright | I am not saying implement IRC in the client, I say use the native chat in bzproto. |
| 05:56.17 | L4m3r | have a proprietary proto for the bzflag client and an IRC/Jabber interface |
| 05:56.30 | L4m3r | hell, theoretically, multiple chat protocols could be used. |
| 05:56.40 | JeffM | AAA_awright: there are many ways to implment it |
| 05:56.47 | JeffM | we could use pidgin connection plugins |
| 05:56.51 | JeffM | it' open source ;) |
| 05:57.06 | JeffM | AAA_awright: but you don't like the concept, so you won't use it |
| 05:57.10 | JeffM | that's fair enough |
| 05:57.12 | AAA_awright | A Jabber account for each BZFlag player is interesting... player@users.bzflag.org or whatever... |
| 05:57.14 | L4m3r | ideally any system built into the client should be dead simple anyway, so we don't need a full-fledged messaging system |
| 05:57.19 | AAA_awright | For outside use. |
| 05:57.29 | *** join/#bzflag coffee (n=whodaman@pdpc/supporter/student/whodaman-) |
| 05:57.42 | brlcad | yay, finally pinpointed the crash on head I think |
| 05:57.52 | JeffM | whoot |
| 05:57.59 | JeffM | AAA_awright: what are you going to call your fork? |
| 05:58.05 | AAA_awright | Of? |
| 05:58.10 | JeffM | bzflag |
| 05:58.16 | JeffM | you mentioned taking over |
| 05:58.21 | AAA_awright | >_> |
| 05:58.37 | AAA_awright | CAN I mean. |
| 05:58.48 | blast007 | you CAN do it! |
| 05:58.55 | L4m3r | lol |
| 05:59.17 | L4m3r | what, you're going to break a fork off of Meteorite's doomed implementation? |
| 05:59.21 | JeffM | ahh |
| 05:59.48 | AAA_awright | Meteorite is helping, I have done most of the direction though (even if he says otherwise) |
| 06:00.17 | L4m3r | Lol |
| 06:00.30 | L4m3r | frankly, you're wasting your time |
| 06:00.30 | AAA_awright | It could become a very capable system, I think, with enough brewing time. |
| 06:00.31 | blast007 | ah, that's neat. OpenFire (a Jabber server) has LDAP support ootb |
| 06:00.51 | JeffM | and a cool name |
| 06:01.04 | JeffM | makes sleep |
| 06:01.26 | AAA_awright | I took it from a marketing standpoint, why would people want to use CAN? I thought since we collect basic, necessary data, why not also offer server usage statistics. |
| 06:01.43 | L4m3r | The major servers have good admins in place, and the minor servers don't care anyway |
| 06:01.52 | L4m3r | So you want to trick people into it? |
| 06:02.03 | brlcad | slowly catches up with the thread |
| 06:02.08 | AAA_awright | That is not the goal |
| 06:02.10 | brlcad | yeah, the client shouldn't be talking directly to ldap |
| 06:02.21 | JBdiGriz | wonders why brlcad is still awake |
| 06:02.23 | brlcad | needs an intermediator |
| 06:02.25 | AAA_awright | ditto brlcad, I thought that was clear though |
| 06:02.46 | L4m3r | probably the only thing with a solid concensus :P |
| 06:02.54 | AAA_awright | :D |
| 06:02.56 | brlcad | ldap is just a backend implementation detail that is big-win because if fully integrates the web services with our auth |
| 06:03.11 | mibocote | is a ldap auth/list server big enough for a gsoc project, though? my understanding now is that it would be modifying the existing auth/list server to use ldap as a backend and configuring ldap to replicate across several server |
| 06:03.15 | brlcad | wonders why JBdiGriz really is wondering after all these years :) |
| 06:03.31 | blast007 | mibocote: or writing a new list system |
| 06:03.34 | AAA_awright | I support the use of a common authentication system, and if LDAP is the most capable thing available, by all means use it. |
| 06:03.38 | brlcad | mibocote: it's more than big enough |
| 06:03.44 | blast007 | the current 'list' is a PHP script |
| 06:03.45 | brlcad | borderline too big in the grand scheme |
| 06:03.59 | AAA_awright | and MySQL/phpBB database |
| 06:04.04 | blast007 | yup |
| 06:04.17 | brlcad | if someone finished it early, there's plenty that can be tossed in to beef it up task-wise .. stats tracking, karma services |
| 06:04.41 | AAA_awright | :/ Don't use the "k" word just yet, please |
| 06:04.50 | mibocote | haha |
| 06:05.00 | AAA_awright | We haven't even gotten how we serve data down, let alone *karma* |
| 06:05.11 | L4m3r | you only hate karma because you'd lose, AAA_awright :) |
| 06:05.27 | brlcad | but just finishing the implementation, integration with the client and servers, and a modestly secure validating auth mechanism is more than enough work |
| 06:05.54 | brlcad | blast007: you actually getting convinced to join the dark side?? :) |
| 06:06.10 | blast007 | brlcad: the dark side of the toast? |
| 06:06.12 | brlcad | (C++ bzauthd) :) |
| 06:06.17 | blast007 | sure |
| 06:06.21 | brlcad | hehe |
| 06:06.30 | brlcad | how hard would it be to translate what you have? |
| 06:06.32 | a_meteorite | L4m3r: I feel pretty offended calling it a "doomed implementation" |
| 06:06.53 | brlcad | or "how much" do you have |
| 06:06.54 | a_meteorite | Just because we don't talk about the things we do publicly, especially most of our plans, certainly doesn't mean it is doomed |
| 06:07.06 | AAA_awright | The Executive Chair speaks! |
| 06:07.12 | blast007 | brlcad: hard to say. I don't "write" C++ applications ;) |
| 06:07.23 | a_meteorite | The implementation in all actuality is nearly finished, it just needs the finishing interface to it |
| 06:07.29 | blast007 | I just copy code from other areas, and change stuff until it works |
| 06:07.50 | AAA_awright | Interface is getting there. Works pretty well, I am not happy just yet, but CAN is getting there. |
| 06:08.14 | mibocote | what is CAN ? |
| 06:08.25 | blast007 | Community Administration Network, iirc |
| 06:08.38 | Constitution | Commuter-Area-Network |
| 06:08.41 | brlcad | AAA_awright: we have data down -- the client and server already have a chatter protocol exchange going -- first step is to change that protocol to be with a different daemon .. the data itself doesnt' need to change |
| 06:08.51 | brlcad | blast007: at least not "yet" :) |
| 06:08.54 | AAA_awright | Agent now, may still change |
| 06:08.58 | blast007 | brlcad: heh... |
| 06:09.12 | blast007 | I did write a C++ app recently, or the start one one, at least |
| 06:09.29 | blast007 | I wrote the start of a system to query bzflag servers |
| 06:09.36 | AAA_awright | brlcad: The current implementation is stable, works, and is threaded, I don't see why a web server and HTTP transfer isn't the best thing there is. |
| 06:09.58 | AAA_awright | We didn't want to write these things from scratch, iirc? |
| 06:10.06 | brlcad | AAA_awright: now's *really* not the time to rehash the fundamental plans yet again :) |
| 06:10.40 | brlcad | it's been discussed to death, there are many reasons for moving off of http |
| 06:11.04 | AAA_awright | We have not decided on protocol yet, how can we decide on implementation? |
| 06:11.09 | a_meteorite | wonders what the conversation was about to begin with, only came in after seeing the ping |
| 06:11.10 | brlcad | it could still be http but there's very little point, and that's not one of the critical aspects |
| 06:11.54 | brlcad | the protocol doesn't drive the implementation, the *requirements* drive the implementation |
| 06:12.04 | brlcad | the requirements drive the protocol too |
| 06:12.10 | AAA_awright | HTTP has very little overhead, is widely supported, and supports many things that are used. If HTTP isn't the best transport protocol, then I don't think anything is. |
| 06:13.05 | brlcad | dead |
| 06:13.06 | brlcad | horse |
| 06:13.22 | AAA_awright | Fine. |
| 06:13.26 | brlcad | ty |
| 06:14.01 | brlcad | mibocote: are you considering proposing the topic? |
| 06:15.04 | AAA_awright | brlcad: Another thing to consider, how do we authenticate tokens? Maintain reverse compatibility with the existing server, or what? |
| 06:15.11 | mibocote | brlcad, strongly considering |
| 06:15.36 | AAA_awright | I am talking about web login mostly |
| 06:16.10 | a_meteorite | mibocote: What are you considering? |
| 06:16.12 | brlcad | AAA_awright: we don't need to maintain compatibility, this is intended as a major update so things can change as needed |
| 06:16.42 | brlcad | oh, you mean for the web services that presently hook to bzbb |
| 06:16.44 | AAA_awright | bzfs and clients are not the only things using the list server |
| 06:16.47 | AAA_awright | Yeah |
| 06:17.18 | AAA_awright | Sorry, I just realized about those implementations |
| 06:17.32 | mibocote | a_meteorite, the global authentication daemon |
| 06:17.47 | brlcad | web services can change as needed when needed, there's no service/operability guarantee there, though we would try to contact folks if needed |
| 06:17.48 | a_meteorite | Ah |
| 06:17.55 | brlcad | otherwise, I don't really see that changing at first stab |
| 06:18.17 | brlcad | bzbb would hook into the directory, auth the user and return the web token just the same |
| 06:18.19 | AAA_awright | mibocote: It seems trivial to me, you have a service that implements the backend BZFlag authentication API (LDAP or whatever) that can store and validate tokens. |
| 06:18.57 | AAA_awright | Anyone consider encryption as a possibility? |
| 06:19.13 | a_meteorite | HTTPS! |
| 06:19.19 | brlcad | for the web services or are you back to the client/server-bzauthd layer? |
| 06:19.21 | AAA_awright | Not to beat on your dead horse, brlcad |
| 06:19.22 | a_meteorite | Hehe :) |
| 06:19.27 | AAA_awright | but meteorite nailed it |
| 06:19.34 | L4m3r | laments the fact that the client strips out ANSI color codes |
| 06:19.41 | L4m3r | as opposed to the server |
| 06:19.59 | a_meteorite | Yeah, I had a plugin that made good use of ANSI color codes |
| 06:20.07 | a_meteorite | Had to mod the client |
| 06:20.10 | brlcad | AAA_awright: for the web services or are you back to the client/server-bzauthd layer? |
| 06:20.25 | AAA_awright | Oh, uh, I think? |
| 06:20.35 | Constitution | a_meteorite: I asked you for that patch ;) |
| 06:20.36 | brlcad | either way "yes", it's been talked about for both |
| 06:20.59 | a_meteorite | Constitution: I still haven't found it. I have probably 2 dozen branches all with their own little mods... |
| 06:21.09 | a_meteorite | I could easily recreate it tomorrow, |
| 06:21.16 | brlcad | web service could be https today if the certificates were installed on the server and the config files set up |
| 06:21.24 | a_meteorite | (I would tonight, but I'd like to get some sleep for a change) |
| 06:21.25 | Constitution | a_meteorite: spotlight ftw |
| 06:21.28 | Constitution | but whatever |
| 06:21.44 | a_meteorite | 5 hours of sleep sucks, trust me |
| 06:22.28 | brlcad | for bzauthd, I'd rather see public/private key + passwords getting managed |
| 06:22.41 | AAA_awright | The implementation seems very simple to me, we choose a common internal password authentication, and create a few services with it (bzauthd or http), including the forum, website, image uploader, and a token store. For 3rd party services (servers, web programs), you use the token store. |
| 06:22.55 | brlcad | otherwise the communcation pipe itself really doesn't matter, that's not a problem in the least right now |
| 06:23.57 | brlcad | AAA_awright: i'm glad you think it's simple, but what's your point in that statement? :) |
| 06:24.14 | AAA_awright | We don't need a Google project around it. |
| 06:24.16 | brlcad | it's never been said to be complicated |
| 06:24.43 | a_meteorite | More like someone Just Needs to Do It™ |
| 06:24.47 | brlcad | well, there is where you'd be missing the other things it can/should entail if there are time |
| 06:25.07 | brlcad | the client mods alone are worthy of several weeks effort |
| 06:25.53 | brlcad | one of the main drivers for having this whole system is so that the players can fully register and auth from inside the client |
| 06:26.58 | brlcad | that involves temp tokens with expiration criteria, account creation and account management UI work, usability testing |
| 06:27.15 | AAA_awright | I like web based regristration, we can change the registration requirements independent of the client (captcha, email, personal info) |
| 06:27.17 | brlcad | the protocol mods needed to support in-game reg |
| 06:27.31 | AAA_awright | But I see why client registration is appealing |
| 06:27.50 | brlcad | well, they should *also* be able to register via the web and/or have a web-based reg still work if it was done first |
| 06:27.55 | brlcad | but unifying the registration process |
| 06:28.03 | brlcad | it's absurd that you have to leave the game to play the game |
| 06:28.05 | AAA_awright | Makes it harder |
| 06:28.19 | AAA_awright | You don't have to at all. |
| 06:28.22 | brlcad | it's a fundamental usability flaw, a huge on imo |
| 06:28.50 | brlcad | you do for an entire class of functionality |
| 06:29.19 | AAA_awright | Who says you must register? The list server can be programmed to not show servers that require registration (a list server protocol design thought) |
| 06:29.23 | brlcad | and not really just a minor set of features, some are pretty major like servers rejecting without global auth |
| 06:29.27 | brlcad | that just hurts the community |
| 06:30.28 | brlcad | never said there weren't other ways to hack around the problem, you could hide stuff from them, but given it's tied into the permissions system, that's not realistic |
| 06:31.17 | brlcad | I can tie a person's ability to speak in the game to global auth .. so what, if the server ties *any* permission to global auth, you don't list them? that's kinda pointless and poorly designed |
| 06:31.18 | AAA_awright | When designing this protocol, keep future expandability in mind. I think I supported a newline separated list of servers, with urlencoded data, so we could communicate which servers require what, and not break reverse compatability. |
| 06:31.28 | brlcad | especially when in-game auth fixes the fundamental problem |
| 06:32.48 | AAA_awright | I don't see any big problems walking around with flashing neon signs saying 04FIX ME (sorry), but yes there are some UI issues |
| 06:33.41 | brlcad | and? |
| 06:33.57 | brlcad | just because you don't see it doesn't mean they aren't there |
| 06:34.01 | AAA_awright | They require proportional fixes. |
| 06:36.12 | AAA_awright | I very much like the concept of the list server, I would work some parts different ways, like the ability to request different data (request XML, urlencoded data, plain text), some data on each server, and I would make user authentication a totally seperate service (I would use the same login script as web login does) |
| 06:36.57 | brlcad | anyways, like I said it boils down to the problems that have been identified that we're trying to fix -- that includes auth consolidation, in-game auth, decoupling from phpbb, and more flexible permissions/group management (players, teams/groups, servers, etc) |
| 06:39.13 | AAA_awright | It doesn't seem too hard in implementation to me, minimal client changes, many easy backend changes to get a more constant structure. |
| 06:40.10 | brlcad | well, you know where the patches tracker is and should know where DEVINFO resides |
| 06:40.31 | AAA_awright | Mhm... |
| 06:41.28 | brlcad | it's still a perfectly viable and high-priority gsoc project that will take most of the summer to implement properly |
| 06:42.07 | brlcad | if someone puts forth a great proposal, it'd be a strong candidate to make the short list |
| 06:42.36 | AAA_awright | It sounds like a month of work for two people, to get 90% feature complete. |
| 06:42.50 | brlcad | heh |
| 06:43.14 | AAA_awright | (keep in mind it it the last 10% that takes 50% of the time) |
| 06:43.21 | AAA_awright | (or something stupid like that) |
| 06:43.27 | brlcad | well then that'd be easily three months for one person to get 100% then while communicating with all the other devs, collaborating on the design, communicating progress, testing, profiling, documenting |
| 06:44.28 | brlcad | regardless, there's no problem getting done earlier |
| 06:44.43 | brlcad | there are many other things that can be thrown in |
| 06:44.43 | AAA_awright | I have some clients, web programs, bzfs server clones, etc to prototype with, I'll think up something. (Didn't I say I was going to do that a while ago?) |
| 06:45.37 | AAA_awright | On a side note, does attaching a bzfs server instance to a callsign sound like a good idea? |
| 06:45.37 | brlcad | AAA_awright: i've said it to you before -- working in isolation really is useless to the dev team -- you have to work openly, visibily, interact productively, given and take on design issues, work with others |
| 06:46.20 | brlcad | it just results in one person that knows their code and N others that know nothing about it, and potentially outright disagree with the approach taken |
| 06:46.23 | AAA_awright | brlcad: I think I tried to get input on some documents with little success |
| 06:46.25 | brlcad | there's no getting around that |
| 06:46.49 | brlcad | then that's a fundamental 'interact productively" problem |
| 06:46.54 | AAA_awright | I'll try again, and get inout on features. I think I covered everything I would like to see in a list server frontend. |
| 06:47.41 | AAA_awright | Wish to open a Wiki page on this, and use the discussion page for comments and input? |
| 06:47.47 | brlcad | if I recall, the problems were mostly because you just become defensive on your view, and wouldn't take other (fundamental) design requests into account |
| 06:48.15 | brlcad | or you restated the obvious as a champion for coming up with an idea that had been long discussed and agreed upon, seemed to be those two extremes |
| 06:48.48 | AAA_awright | I blame myself for that |
| 06:48.49 | brlcad | which is just yet another indicator that you're not working with others, just throwing stuff at others and telling them to take it |
| 06:49.59 | AAA_awright | Well, how about the Wiki page idea? I seriously do not believe a great control system requires massive changes to anything. |
| 06:50.10 | brlcad | your involvement would be greatly appreciated, that goes for any dev that can be productive and work well/effectively with the group |
| 06:50.38 | brlcad | it's not massive changes, but it's also not just a couple days of effort |
| 06:51.06 | AAA_awright | There are some larger issues to tackle together, like how do we transfer Wiki edits, etc, which can be done collaboratively. |
| 06:51.08 | brlcad | and the time to educate, discuss, and iterate/justify the design and implementation IS important, sometimes more important than the implementation itself |
| 06:51.53 | AAA_awright | brlcad: Wiki page to prototype a structure collaboratively, go for it? |
| 06:52.18 | brlcad | there is one already started actually |
| 06:52.45 | brlcad | it's not been reviewed in a long time, needs lots of updates |
| 06:52.49 | AAA_awright | Must be an orphan then? I know there is a page on the current system... |
| 06:52.51 | brlcad | http://my.bzflag.org/w/BZAuthd |
| 06:53.26 | brlcad | again, it's out of date |
| 06:53.50 | brlcad | but that would be the *location* to hash out a design |
| 06:54.21 | AAA_awright | Hm. I would like to start from scratch. |
| 06:54.36 | brlcad | that's part of my point :) |
| 06:55.33 | brlcad | working with non-optimal mess is important, iterate on the design, otherwise you lose the effort and mindset that has gone into it (and are much more likely to come up with something that doesn't fit well as it's not discussed) |
| 06:56.20 | brlcad | it has terminology that are more understood, for example, and attempts to define terms for consistency |
| 06:56.58 | brlcad | nibm is a deadly disease |
| 06:57.03 | AAA_awright | Karma is a joke, you can't do it on the list server, that is one thing a 3rd party will have to do, and to make it easy, you must provide the necessary data to let them compile karma stats on players (CAN for example) |
| 06:57.33 | brlcad | sure you could |
| 06:58.09 | brlcad | list service actually becomes just a secondary service -- it's an entity management/authentication service |
| 06:58.20 | brlcad | listing servers is just a class of entities it knows about |
| 06:58.27 | AAA_awright | A network of trust could be done, but that is complex, and again, best left to a 3rd part |
| 06:58.39 | brlcad | listing players and tracking attributes for any entity are other roles |
| 06:58.55 | AAA_awright | Where we trust servers to provide accurate data, and we trust clients to trust servers. |
| 06:59.31 | brlcad | either way, karma is still just one piece and by far one of the more important pieces |
| 06:59.48 | AAA_awright | If we can trust a server, what prevents a player from manipulating their karma? Anyways, not the issue at hand, I would like to see someone elses complete design |
| 07:00.30 | AAA_awright | CAN implements a simple karma system, and a simple network of trust. That is what I mean by 3rd party. |
| 07:00.38 | brlcad | and for what it's worth, you're also apparently not reading that page in full either |
| 07:00.46 | brlcad | it actually specs karma as a separate service |
| 07:00.53 | brlcad | and otherwise just defines it |
| 07:01.13 | mibocote | Thanks for your help guys, early morning lecture for me. Hope to see you again tomorrow. |
| 07:01.19 | brlcad | mibocote: cya |
| 07:01.21 | AAA_awright | Cya mibocote |
| 07:01.28 | AAA_awright | I was actually looking at http://my.bzflag.org/w/Karma_System |
| 07:02.26 | brlcad | well i'm not going to get into the design and depoyment of karma itself, that's a long drawn out bikeshed discussion |
| 07:03.02 | AAA_awright | Yes, an entirely different beast |
| 07:03.17 | brlcad | i'm just referring to the high-level concepts, of which "karma" is no different than a basic "honor"/merit metric you might see in other games that are attached to players as a system attribute |
| 07:03.54 | brlcad | and as a player attribute, bzauthd would be a broker for that data as it's a broker for all player data |
| 07:04.19 | brlcad | s/player/entity/ |
| 07:04.48 | brlcad | same applies to servers and groups as well |
| 07:05.25 | AAA_awright | I just saw http://my.bzflag.org/w/Match_Management_System , I know I am (unsuccessfully) working on Masterbot, which is designed as an advanced platform for running bzfs servers. |
| 07:08.03 | brlcad | running bzfs servers is only loosely connected to match management |
| 07:09.17 | AAA_awright | I was hoping to seperate plugins and such from the match management. You tell masterbot what you want done, it does it. It can schedule servers and allocate teams to certain players, for example (ideally) |
| 07:10.03 | brlcad | you could still tack that on top |
| 07:10.17 | brlcad | that page mostly addresses the actually execution of the match |
| 07:10.25 | brlcad | starting/stopping it right then and there |
| 07:10.36 | brlcad | pausing, swapping players |
| 07:10.43 | Manu | yawns |
| 07:10.44 | AAA_awright | Masterbot could run that too (ideally). |
| 07:11.10 | brlcad | doesn't sound like something a bot really needs to be involved with |
| 07:11.19 | AAA_awright | Right now I am stuck at how do you let people start servers (I can start servers, I am working out permissions) |
| 07:11.43 | brlcad | the server is presently the "umpire" if you will and running matches are an integrated feature, hence integrated commands for controlling them |
| 07:11.50 | AAA_awright | And SQLite. That is giving me more problems then I can believe. |
| 07:12.02 | brlcad | starting servers and the sort is a level higher (and could be useful) |
| 07:12.18 | AAA_awright | Yeah |
| 07:14.07 | brlcad | yawns back at Manu |
| 07:14.21 | AAA_awright | This is a very high-low level thing (at the same time?) |
| 07:14.33 | brlcad | fg |
| 07:15.26 | AAA_awright | It will start a server when you ask it to, no command line stuff needed, but you can program custom callbacks to server events too. |
| 07:15.33 | AAA_awright | fg? |
| 07:15.39 | brlcad | ww |
| 07:20.23 | *** join/#bzflag QuantumBeep (i=QB@72.190.51.216) |
| 07:25.14 | L4m3r | should we (bzbb admins) deal with impostor accounts as we see them? |
| 07:27.40 | AAA_awright | brlcad: Does http://my.bzflag.org/wiki/index.php?title=BZFlag_Services_Proposal work? I don't want to make an orphan page, though (even though it would be in the category list) |
| 07:29.27 | AAA_awright | Hm, I'll do this tomorrow then. |
| 07:29.30 | AAA_awright | Later. |
| 07:29.43 | Manu | contributes to spam the mentors mailing list |
| 07:29.47 | Manu | :) |
| 07:29.48 | *** join/#bzflag bz39 (n=7dec87a6@bz.bzflag.bz) |
| 07:30.47 | brlcad | L4m3r: sure |
| 07:31.00 | *** join/#bzflag bz26 (n=7dec87a6@bz.bzflag.bz) |
| 07:31.02 | brlcad | hibernates |
| 07:31.13 | bz26 | hi |
| 07:31.33 | Constitution | hi |
| 07:31.42 | bz26 | wow this irc is working |
| 07:31.53 | brlcad | you expected it not to? |
| 07:31.59 | bz26 | i've tried before |
| 07:32.02 | Manu | remembers brlcad it's spring. No time to hibernate ;) |
| 07:32.05 | bz26 | had errors with my name |
| 07:32.12 | bz26 | cos it had space in it |
| 07:32.28 | bz26 | any devs here? |
| 07:32.34 | bz26 | [active] |
| 07:32.55 | Manu | ~ask |
| 07:32.56 | ibot | Questions in the channel should be specific, informative, complete, concise, and on-topic. Don't ask if you can ask a question first. Don't ask if a person is there, just ask what you intended to ask them. Better questions more frequently yield better answers. We are all here voluntarily or against our will. |
| 07:34.09 | bz26 | Where can I learn about making bzflag bots. Apart from studying the code. Any other way? |
| 07:34.54 | brlcad | http://my.bzflag.org/w/BZRobots |
| 07:35.06 | brlcad | src/bzrobots/ |
| 07:35.39 | bz26 | read that already, hasn't given me any help |
| 07:36.13 | brlcad | are you a user or a dev? |
| 07:36.19 | bz26 | user |
| 07:36.28 | brlcad | bots aren't quite ready for the public to create them just yet |
| 07:36.32 | brlcad | else you'd see documentation |
| 07:36.37 | bz26 | ah. |
| 07:36.42 | brlcad | still intended for devs, there's a little more polish they need |
| 07:36.48 | brlcad | just a little |
| 07:37.04 | bz26 | if i was determined, that wouldn't be bad, right? |
| 07:37.05 | brlcad | there also would have been public announcements |
| 07:37.11 | brlcad | and probably a competition |
| 07:37.15 | brlcad | with prizes and such |
| 07:37.27 | Manu | bz26: right |
| 07:37.32 | brlcad | if you're determined, you can make anything happy |
| 07:37.38 | brlcad | s/happy/happen/ |
| 07:38.07 | bz26 | if it's going to be polished, will it be easier then? |
| 07:38.19 | bz26 | or will it be the same |
| 07:38.51 | Manu | probably it will be easier :) |
| 07:39.09 | bz26 | :/ unsure whether to do anything now then |
| 07:39.10 | Manu | that's the idea AFAIK |
| 07:39.42 | Manu | bz26: for now you can read the code to see how it works actually |
| 07:40.13 | bz26 | Well, I'm not reading the code yet, I'm speedLearning C++ lol |
| 07:40.29 | Manu | heh |
| 07:41.31 | bz26 | If I was able to understand some of the code, and i wanted to try something simple like make bot keep rotating... |
| 07:41.46 | bz26 | is it true i must download like a gigabyte of stuff? |
| 07:41.48 | bz26 | to run it |
| 07:42.32 | Manu | probably the easier way to understand it is to download the bzflag source code (not 1 Gb) and see how autopilot works |
| 07:42.57 | bz26 | that's what i'm doing, i'm wondering if i can try some of my own stuff |
| 07:43.02 | Manu | then you can try to modify it adding the AI |
| 07:43.15 | bz26 | if i modify some stuff, how do i get it to run? |
| 07:43.26 | Manu | first step is learn to compile it |
| 07:43.44 | Manu | autogen.sh && configure && make |
| 07:44.36 | Manu | bz26: http://my.bzflag.org/w/Category:Compiling |
| 07:44.56 | Manu | and http://my.bzflag.org/w/Main_Page ;) |
| 07:45.14 | bz26 | The -solo bot is more relevant to me. Can't find it |
| 07:45.46 | Manu | first is the step #1 (learn to compile it) |
| 07:46.03 | Manu | second try to change something and get it working ;) |
| 07:47.10 | bz26 | I'd be happy even if i could just figure out how to make the bot move forward :) |
| 07:47.26 | bz26 | The -solo bot is more relevant to me. Can't find it |
| 07:47.27 | bz26 | The -solo bot is more relevant to me. Can't find it |
| 07:47.37 | bz26 | oops, tried pasting |
| 07:48.47 | Manu | the -solo #n is only a parameter |
| 07:49.41 | bz26 | the -solo bots are different from auto bots |
| 07:49.55 | bz26 | are their codes at different places ? |
| 07:50.19 | Manu | autopilot manages your tank |
| 07:50.30 | Manu | -solo creates virtual tanks :) |
| 07:50.41 | Manu | or bots |
| 07:50.54 | bz26 | i mean they play differently |
| 07:51.03 | Manu | true |
| 07:51.08 | bz26 | autopilot uses superflags, the other doesn't |
| 07:51.19 | bz26 | and a whole lot of other things |
| 07:51.24 | Manu | bots have limitations |
| 07:51.43 | bz26 | i'm not too interested in bot and flag interaction right now, just basics |
| 07:52.58 | Manu | try to look for RobotPlayer.h ;) |
| 07:53.15 | Manu | and playing.cxx |
| 07:53.20 | CIA-16 | BZFlag: 03brlcad * r16809 10/trunk/bzflag/src/bzflag/BackgroundRenderer.cxx: (log message trimmed) |
| 07:53.20 | CIA-16 | BZFlag: finally finally finally pinpointed what was causing the crash if there wasn't a |
| 07:53.20 | CIA-16 | BZFlag: config file present. the problem was related to the test for the various |
| 07:53.20 | CIA-16 | BZFlag: quality modes (which only occurs when there isn't a config file) where the |
| 07:53.21 | CIA-16 | BZFlag: moonGState was apparently bogus. looks like the state was getting incorrectly |
| 07:53.23 | CIA-16 | BZFlag: stashed/set causing much opengl unhappiness in the app. looks like the |
| 07:53.25 | CIA-16 | BZFlag: moonTexture was half disabled, but the state wasn't updated to match. ripping |
| 07:53.44 | Manu | bz26: you must be able to find it by sourself if you try to modify it :) |
| 07:53.48 | brlcad | son of a ..... that was hell to pinpoint |
| 07:54.00 | Manu | heh |
| 07:54.24 | bz26 | lol, ty for all your help |
| 07:54.25 | brlcad | welp, the mac client finally works again here, next to fix the crash-on-exit |
| 07:54.35 | Manu | bz26: bbl :/ |
| 07:54.52 | Manu | brlcad: crash on exit is not so important ;) |
| 07:55.50 | brlcad | yeah, but it's really annoying :) |
| 07:56.17 | brlcad | i have to wait 20 sec every time I quit the client while it autogenerates a crash report |
| 07:56.24 | brlcad | hard to iterate on dev quickly |
| 07:56.37 | brlcad | plus the crash on exit is related to the new font system, which I hooked in |
| 07:57.06 | brlcad | though *I* left it working, /me blames jeff |
| 07:57.58 | brlcad | (in all fairness, he fixed a crash on windows that broke linux/mac, which I refixed differently but still rebroke windows, which he changed again, which rebroke linux/mac and brings us current |
| 08:12.33 | *** part/#bzflag bz26 (n=7dec87a6@bz.bzflag.bz) |
| 08:28.29 | Manu | brlcad: we can fix it and break the windows one ;) |
| 08:29.01 | Manu | like ping pong :) |
| 09:02.49 | *** join/#bzflag utrrrongeeb (n=mhciipec@ties-pool5-10.isn.net) |
| 09:23.39 | *** join/#bzflag Daveman (n=dave@pool-141-153-244-48.mad.east.verizon.net) |
| 09:28.58 | *** join/#bzflag LongDon (n=LongDon@134.106.59.202) |
| 09:29.24 | *** part/#bzflag LongDon (n=LongDon@134.106.59.202) |
| 09:39.34 | *** join/#bzflag ruskie (i=ruskie@sourcemage/mage/ruskie) |
| 09:43.32 | *** join/#bzflag ChanServ (ChanServ@services.) |
| 09:43.32 | *** mode/#bzflag [+o ChanServ] by irc.freenode.net |
| 10:01.23 | *** join/#bzflag ViciousM (n=max@cpe-024-211-248-161.nc.res.rr.com) |
| 10:17.43 | *** join/#bzflag Suspect (n=Suspect@68-113-163-117.dhcp.plt.ny.charter.com) |
| 10:26.42 | *** join/#bzflag PrezzKennedy (i=Matthew@74.86.45.130) |
| 10:51.39 | *** join/#bzflag PrezKennedy (i=Matthew@74.86.45.130) |
| 11:08.35 | *** join/#bzflag Upsetter (n=Ups@i577B4B77.versanet.de) |
| 11:11.37 | *** join/#bzflag KingofCamelot (n=davidsan@207-177-229-17.dsl.redshift.com) |
| 11:12.07 | KingofCamelot | Anyone around? |
| 11:12.57 | whodaman | mmhm |
| 11:13.42 | KingofCamelot | Think you could answer a question about building form the SVN? :) |
| 11:14.32 | [dmp] | ask they question and see :) |
| 11:14.37 | [dmp] | the* |
| 11:14.42 | whodaman | yay [dmp] to help me out |
| 11:15.07 | whodaman | KingofCamelot, I'm a complete nooby but maybe :D |
| 11:15.38 | KingofCamelot | When I start the build with MSVC8, when it tries to build common it has the following error message: "fatal error C1083: Cannot open source file: '..\..\src\common\Vector4D.cxx': No such file or directory" |
| 11:15.50 | KingofCamelot | and then repeats that message for Vector3D, and Vector2D |
| 11:15.52 | whodaman | looks at [dmp] |
| 11:16.22 | Winny | KingofCamelot: I believe Epyon was working on some of those files |
| 11:16.24 | KingofCamelot | Everything else seems to build fine, except it can't link to common.lib since common.lib never builds :P |
| 11:16.31 | Winny | not sure if the project is updated |
| 11:16.38 | KingofCamelot | D'oh |
| 11:16.41 | [dmp] | i dont think it is |
| 11:16.50 | KingofCamelot | So a solution would be? |
| 11:17.03 | Winny | no idea |
| 11:17.08 | [dmp] | i think i saw a commit-mail saying that vector was deleted |
| 11:17.08 | KingofCamelot | Heh |
| 11:17.26 | KingofCamelot | This is with the latest code pulled down from SVN |
| 11:17.40 | [dmp] | "Removing vector classes -- not worth the time" |
| 11:17.52 | [dmp] | I think you can safely remove the files from your project :) |
| 11:18.22 | KingofCamelot | Will try :) |
| 11:19.44 | KingofCamelot | Appears that it built successfully now :) |
| 11:20.37 | [dmp] | goodie :) |
| 11:20.54 | [dmp] | i knew there was a good reason for me reading the commit-mails :) |
| 11:21.19 | KingofCamelot | hehe |
| 11:22.53 | KingofCamelot | I should really go to bed but I want this damn thing to build :P |
| 11:23.40 | [dmp] | is in the "good" timezone. Where you got most of the day to work :) |
| 11:24.09 | KingofCamelot | wonders which one is the "good" timezone. :P |
| 11:25.06 | *** join/#bzflag tupone (n=atupone@pdpc/supporter/active/Tupone) |
| 11:25.06 | *** mode/#bzflag [+v tupone] by ChanServ |
| 11:25.24 | [dmp] | Currently its CET, but i doubt it will stay good :) |
| 11:26.03 | *** join/#bzflag Saturos (n=Saturos@p57A0DF8D.dip.t-dialin.net) |
| 11:26.11 | KingofCamelot | Nothing good lasts forever eh? :) |
| 11:26.36 | [dmp] | nope :( |
| 11:27.28 | KingofCamelot | March 31st is getting close. :| |
| 11:28.46 | [dmp] | Which is a good thing: Pay-day and I start at new company on tuesday. But also makes it impossible to work on gsoc officially. |
| 11:28.58 | [dmp] | but.. i'll manage :) |
| 11:29.37 | KingofCamelot | I just really learned about gsoc about, oh, two days ago, hehe. The deadline seems awfully close. :P |
| 11:31.03 | [dmp] | it is :) |
| 11:31.38 | KingofCamelot | Thanks for the moral support. ;) |
| 11:35.27 | Constitution | ~brlcad++ |
| 11:35.51 | Constitution | in fact |
| 11:35.57 | Constitution | ~brlcad+++++++++++++++++++++++++++++ |
| 11:36.27 | whodaman | lawl |
| 11:39.12 | KingofCamelot | hmm...what is this nonsense about Autohunt.obj, lol |
| 11:41.50 | *** join/#bzflag Dessy (n=Dessy@88-134-204-43-dynip.superkabel.de) |
| 11:43.05 | *** join/#bzflag KTL (n=KTL@213.219.144.106.adsl.dyn.edpnet.net) |
| 11:46.34 | *** join/#bzflag SunnyCode (n=chatzill@cEC01BF51.dhcp.bluecom.no) |
| 11:50.31 | SunnyCode | Hei |
| 11:50.53 | Constitution | hi |
| 11:51.22 | whodaman | ello |
| 11:51.31 | KTL | hoi |
| 11:54.03 | KingofCamelot | Anyone have any idea why the Autohunt object might not be building? |
| 11:56.53 | Constitution | what is an "Autohunt object"? |
| 11:57.53 | KingofCamelot | When building from the SVN source code, it should create Autohunt.obj when building bzflag...atleast bzflag thinks so, lol |
| 11:58.53 | Constitution | you mean AutoHunt.o? |
| 11:59.03 | Constitution | the object file? |
| 12:00.10 | KingofCamelot | The compiler calls it AutoHunt.obj, but sure, lol |
| 12:00.13 | Constitution | as opposed to a model file in the .obj format |
| 12:00.19 | Constitution | what compiler? |
| 12:00.23 | KingofCamelot | Yes, the object file |
| 12:00.27 | KingofCamelot | MSVC 8 |
| 12:00.34 | Constitution | ah |
| 12:00.39 | KingofCamelot | fatal error LNK1104: cannot open file '.\bzflag_debug\AutoHunt.obj' |
| 12:01.26 | Constitution | and the AutoHunt.cxx and AutoHunt.h files are included in the project? |
| 12:01.31 | KingofCamelot | Yep |
| 12:01.40 | KingofCamelot | just doesn't seem to be making the obj, is making all the other ones |
| 12:01.51 | Constitution | hmm |
| 12:02.09 | Constitution | maybe something got messed up in yesterday's changes to the project files |
| 12:02.57 | KingofCamelot | I know I had to remove the Vector files from common earlier to get that one to build correctly |
| 12:03.03 | KingofCamelot | so ya, its a possibility |
| 12:05.38 | Constitution | you might try reverting to r16800 and see if it works |
| 12:06.02 | *** join/#bzflag Paul_PL (n=pawelbl6@chello087206115169.chello.pl) |
| 12:07.11 | Constitution | and of course, patches are always welcome ;) |
| 12:07.37 | KingofCamelot | Would have to figure out what the problem is first :P |
| 12:14.01 | *** join/#bzflag Pimpi (n=frank@gondolin.pimpi.org) |
| 12:16.23 | *** join/#bzflag ruskie (i=ruskie@sourcemage/mage/ruskie) |
| 12:18.55 | KingofCamelot | Apparently my bzflag project somehow got screwed up...maybe during trying to fix common, dunno. But updated it to the latest SVN version to overwrite it, and now AutoHunt.obj appears :) |
| 12:22.07 | Constitution | good :) |
| 12:25.22 | *** join/#bzflag Saturos (n=Saturos@p57A0DF8D.dip.t-dialin.net) |
| 12:27.56 | KingofCamelot | And wow, no offense to anyone, but the current server listing is kinda awkward to use, heh |
| 12:29.46 | [dmp] | KingofCamelot: http://my.bzflag.org/w/Google_Summer_of_Code/2008#Enhanced_server_listing :-) |
| 12:30.11 | KingofCamelot | I know ;) Thats why I was taking a look at it |
| 12:30.15 | KingofCamelot | I'm going to apply for that :P |
| 12:30.34 | [dmp] | ah |
| 12:30.48 | [dmp] | well, lots of room for improvements :) |
| 12:30.55 | KingofCamelot | Indeed, heh |
| 12:31.17 | KingofCamelot | Any specific reason that you select a server, then go back a screen to hit connect? |
| 12:32.03 | [dmp] | I dont know |
| 12:32.18 | KingofCamelot | Just seems kinda ood |
| 12:32.20 | KingofCamelot | *odd |
| 12:36.25 | KingofCamelot | Anyways, now that everything is compiling nicely, I'm going to sleep :P |
| 12:40.49 | [dmp] | good night :) |
| 12:41.07 | Constitution | KingofCamelot: pressing enter on the "Find Server" screen selects the highlighted server and returns you to the "Join Game" screen... pretty much an essential step |
| 12:41.10 | whodaman | night, [dmp] |
| 12:41.55 | Constitution | you might find it interesting and/or helpful to set up a list server on your local machine, and make a client and server talk to it :) |
| 12:42.01 | Constitution | anyway, night |
| 12:42.41 | KingofCamelot | Well, I get that it selects the server, then takes you back so that you can click connect...but why not just let you connect from the Find Server menu? :P |
| 12:43.23 | Constitution | you might choose a different team or select a different option of some kind after choosing a server to connect to :) |
| 12:43.45 | Constitution | for instance, I might see a game I want to observe, and another that I'd want to play in |
| 12:44.29 | KingofCamelot | Ah, k, makes slightly more sense now :) |
| 12:44.41 | KingofCamelot | I'll play around with it more tomorrow...its almost 6 AM...sleep calls, hehe |
| 12:44.56 | Constitution | see ya :) |
| 13:02.29 | *** join/#bzflag ruskie (i=ruskie@sourcemage/mage/ruskie) |
| 13:03.55 | jorgenpt | brlcad: I'll probably get a reply from Google about my summer internship soon-ish, then I'll know if and how I can participate in BZFlag's GSoC project(s). :-) |
| 13:11.07 | *** join/#bzflag pacman_87 (n=timothy@nat-202-129.arlut.utexas.edu) |
| 13:14.49 | *** join/#bzflag O-Neil (n=Miranda@dslb-088-070-246-167.pools.arcor-ip.net) |
| 13:17.43 | *** join/#bzflag Dessy (n=Dessy@88-134-204-43-dynip.superkabel.de) |
| 14:08.36 | *** join/#bzflag spldart (n=short_ci@adsl-70-240-251-10.dsl.hstntx.swbell.net) |
| 14:13.06 | TheRedBaron | perhaps a pinch OT - but this paper - http://www.sigcomm.org/sigcomm2004/workshop_papers/net610-aggarwal.pdf the people who wrote it - have they ever been on irc? |
| 14:33.19 | *** join/#bzflag BulletCatcher (n=bc@unaffiliated/bulletcatcher) |
| 14:40.23 | *** join/#bzflag whodaman (n=whodaman@pdpc/supporter/student/whodaman-) |
| 14:42.28 | *** join/#bzflag whodaman (n=whodaman@pdpc/supporter/student/whodaman-) |
| 14:42.59 | Manu | TheRedBaron: isn't that document a bit old? ;) |
| 14:43.50 | TheRedBaron | 2005 ? |
| 14:44.30 | Manu | 2004 |
| 14:44.41 | Manu | things changed in bzflag since then |
| 14:48.28 | *** join/#bzflag whodaman (n=whodaman@pdpc/supporter/student/whodaman-) |
| 14:48.53 | TheRedBaron | has the DR code since then? |
| 14:51.15 | Manu | I guess yes |
| 14:52.30 | Manu | IIRC there was a discussion about it here, years ago. |
| 14:52.55 | TheRedBaron | its still listed on the wiki of ideas as needing improvement. |
| 14:53.02 | Manu | a big change was made from 1.X to 2.X |
| 14:53.14 | Manu | 2.0.0 was released in 2005 |
| 14:53.28 | TheRedBaron | but my primary reason for asking was - the paper mentions my college. |
| 14:53.44 | Manu | :) |
| 14:55.47 | Manu | TheRedBaron: but really DR needs to ve improved ;) |
| 14:56.20 | *** join/#bzflag Gilly (n=btw@83.146.211.140) |
| 15:01.35 | TheRedBaron | Manu: oh ok |
| 15:03.17 | Manu | TheRedBaron: are you interested in programming bzflag? :) |
| 15:04.01 | TheRedBaron | Manu: if I knew how ;) |
| 15:04.27 | TheRedBaron | but more than likely - my incompetence would be staggering :) |
| 15:04.48 | Manu | TheRedBaron: http://www.dummies.com/WileyCDA/DummiesTitle/C-For-Dummies-5th-Edition.productCd-0764568523.html |
| 15:04.51 | Manu | :) |
| 15:05.03 | spldart | hehe |
| 15:05.12 | TheRedBaron | no good - I need the "utter idiot" edition :P |
| 15:05.34 | Manu | TheRedBaron: ok http://www.developerfusion.co.uk/show/3847/ |
| 15:05.45 | Manu | ;) |
| 15:06.02 | TheRedBaron | now you're talking :P |
| 15:06.59 | Gilly | "Would you rather hire this guy, having read The complete idiot's guide to C, or this other guy, who ..." :) |
| 15:08.32 | Manu | :) |
| 15:08.41 | TheRedBaron | "... who only had to read C++ for dummies" |
| 15:25.55 | Manu | TheRedBaron: how's your new life? :) |
| 15:27.28 | *** join/#bzflag JeffM (n=JeffM@unaffiliated/jeffm) |
| 15:27.57 | Constitution | morning JeffM |
| 15:28.02 | JeffM | I could leave but I'll just stay.. all my stuff's here anyway |
| 15:28.06 | JeffM | word up |
| 15:29.29 | TheRedBaron | Manu: I'm enjoying it :) |
| 15:30.01 | Manu | TheRedBaron: my best wishes to you :) |
| 15:30.27 | CBG | This can't be right... Nobody enjoys married life... |
| 15:30.36 | JeffM | many do |
| 15:30.39 | JeffM | I did for a while |
| 15:30.41 | Manu | CBG: nobody? |
| 15:30.45 | CBG | Has she brainwashed you already, TheRedBaron? |
| 15:30.49 | CBG | Manu: so I heard :P |
| 15:31.03 | Manu | CBG: after 20 years I have no complains ;) |
| 15:31.06 | Constitution | ~CBG-- |
| 15:31.23 | Constitution | er wait |
| 15:31.38 | JeffM | give him a couple years and he'll probably change his tune :) |
| 15:31.39 | CBG | Manu's brainwashed too! :( |
| 15:31.41 | CBG | lol Constitution |
| 15:31.45 | CBG | Whatever you do, I won't know what you intended. |
| 15:32.02 | TheRedBaron | :P |
| 15:32.15 | CBG | Manu: I always thought of you as about 25 years old.. :o |
| 15:32.21 | *** join/#bzflag Lukstr (n=lukstr@131.104.8.47) |
| 15:32.24 | Manu | lol |
| 15:32.44 | TheRedBaron | CBG: he is - he's in an arranged marriage :P |
| 15:32.52 | CBG | lol :D |
| 15:32.56 | Lukstr | howdy |
| 15:33.01 | CBG | ahoyhoy |
| 15:33.01 | spldart | heh |
| 15:33.27 | CBG | Are you married or in a long-term relationship, Lukstr? |
| 15:34.22 | Lukstr | define ltr |
| 15:34.36 | Lukstr | and how come those are the only options? |
| 15:34.42 | CBG | Gah. Another one lost... |
| 15:34.53 | CBG | Lukstr: You could always say "no". |
| 15:35.24 | Lukstr | well, define a long-term relationship? |
| 15:35.37 | CBG | Say... More than an hour...? |
| 15:35.50 | Lukstr | then yes |
| 15:36.09 | *** join/#bzflag hobbsc (n=hobbsc@ip70-178-55-100.ks.ks.cox.net) |
| 15:36.13 | CBG | I don't think there are many of us left.. :( |
| 15:36.24 | Lukstr | who, single people? |
| 15:37.39 | CBG | Yeah.. But we like to be called Normies as in 'Normal People'. |
| 15:37.50 | Lukstr | ....k |
| 15:37.58 | Lukstr | things CBG is !normal... |
| 15:38.04 | Lukstr | s/things/thinks/ |
| 15:38.07 | Lukstr | :D |
| 15:38.39 | Manu | thinks CBG is a lonley wolf |
| 15:38.49 | CBG | howls into the wind |
| 15:38.52 | Manu | s/wolf/chiwawa/g |
| 15:39.01 | Manu | ;) |
| 15:39.04 | TheRedBaron | He's a peach! |
| 15:39.09 | TheRedBaron | Round and fuzzy |
| 15:39.23 | CBG | yaps into the wind |
| 15:39.54 | Constitution | CBG: actually, in a crowd like this, I think there must be a good, hearty group of us ;) |
| 15:40.19 | Lukstr | it's usually something like 50% at the places I've worked |
| 15:40.22 | CBG | We just lost 3, Constitution! |
| 15:40.44 | *** join/#bzflag ruskie (i=ruskie@sourcemage/mage/ruskie) |
| 15:46.30 | *** join/#bzflag MyLittleChainsaw (n=cd9020fa@gateway/web/cgi-irc/zeebrothers.net/x-d276c8619285f244) |
| 15:47.06 | MyLittleChainsaw | can anyone here unban me from bzchat |
| 15:47.36 | JeffM | why'd you get banned in the first place |
| 15:47.46 | brad | don't use the zeebrother's cgi irc and you will be able to get in |
| 15:47.55 | JeffM | ahh yeah zee |
| 15:48.03 | JeffM | autoban on that one :) |
| 15:48.15 | MyLittleChainsaw | only one i can get to |
| 15:48.24 | MyLittleChainsaw | skool blocks all others |
| 15:48.40 | JeffM | you can't get to the one on the fourms? |
| 15:49.32 | MyLittleChainsaw | nope the site is filtered through the system |
| 15:49.38 | Lukstr | MyLittleChainsaw: I'm guessing you don't happen to have a server somewhere running sshd on port 23 |
| 15:49.38 | JeffM | then sorry, can't help you |
| 15:51.39 | CBG | ROFL Lukstr |
| 15:52.16 | Lukstr | CBG: What? That's what I did in high school |
| 15:52.46 | CBG | Lukstr: you != MyLittleChainsaw |
| 15:53.11 | Lukstr | CBG: I never said I was? |
| 15:53.41 | CBG | I found it funny that you thought maybe he would do the same as you |
| 15:54.04 | MyLittleChainsaw | srry i nno respond cant get caught |
| 15:54.12 | MyLittleChainsaw | umm |
| 15:54.17 | Lukstr | CBG: Well it's a very sane solution |
| 15:54.17 | JeffM | MyLittleChainsaw, then maybe you shouldn't be doing this |
| 15:54.19 | MyLittleChainsaw | <PROTECTED> |
| 15:54.44 | Lukstr | MyLittleChainsaw: If it's that big a deal at your school, maybe you shouldn't be doing this, like JeffM says |
| 15:54.44 | MyLittleChainsaw | heh lol |
| 15:54.52 | JeffM | I'm not going to unblock the zeebrothers CGI IRC, when it's just so you can shirk schoolwork |
| 15:55.02 | JeffM | we have enough CGI IRC problems as is |
| 15:55.26 | *** join/#bzflag praetorian (n=praetori@124-171-25-117.dyn.iinet.net.au) |
| 15:56.26 | *** join/#bzflag ruskie (i=ruskie@sourcemage/mage/ruskie) |
| 15:56.44 | *** join/#bzflag QuantumBeep (i=QB@72.190.51.216) |
| 15:56.53 | MyLittleChainsaw | fine but i be wokin and its not that big of a deal |
| 15:56.56 | MyLittleChainsaw | lol |
| 15:57.08 | *** join/#bzflag Wyk3d (n=Wyk3d@lns-bzn-47f-81-56-254-111.adsl.proxad.net) |
| 15:57.18 | MyLittleChainsaw | bbl |
| 15:57.45 | CBG | oook |
| 16:00.49 | Constitution | no idea why his school would have an exemption or whatever for zeebrothers... it isn't exactly a mainstream IRC gateway site |
| 16:01.16 | Gilly | :) |
| 16:01.50 | Constitution | as far as non-bz uses, anyway |
| 16:01.55 | blast007 | Constitution: they probably operate on a black-list, not a white-list |
| 16:01.57 | CBG | A school I used to go to had a whitelist policy, a teacher had to ask for any sites to be unblocked as everything was blocked by default... |
| 16:02.06 | CBG | Except miniclip.com. That was never blocked.. |
| 16:02.17 | blast007 | and the zeebrothers site site probably isn't marked as a site to block |
| 16:02.31 | Constitution | blast007: and m.b.o is? |
| 16:02.33 | blast007 | the *.bzflag.org is obvioulsy a game site, so they block that |
| 16:02.40 | Constitution | hm |
| 16:03.07 | ruskie | solution at my old school was simple... no internet access during classes |
| 16:03.14 | blast007 | heh |
| 16:03.37 | ruskie | they simply took down the relevant interface on a linux box that was used for the class net access |
| 16:03.37 | Lukstr | ruskie, we used to play scorched earth all class :D |
| 16:03.48 | ruskie | Lukstr, q2 here :) |
| 16:03.55 | CBG | That would mess up a LOT of lessons these days. |
| 16:03.56 | Lukstr | I win :D |
| 16:04.11 | ruskie | but then we also played liero :) |
| 16:04.27 | ruskie | and tetrinet :) |
| 16:04.34 | Lukstr | netris! |
| 16:04.35 | CBG | And had to cook your food over a fire? :P |
| 16:04.58 | Gilly | i did my very last exam in high school today so i guess i can talk about "my old school" too :) we didn't have computers in classes, only the special computer class and we didn't get in there much |
| 16:04.59 | ruskie | gas stove @home... so yeah.. it's still over a fire :) |
| 16:05.18 | TheRedBaron | Gas Stove > electic |
| 16:05.28 | Gilly | now i'm just wondering which uni to pick :P |
| 16:05.29 | ruskie | well considering this was a computer related course we had quite a few weekly classes there |
| 16:05.37 | blast007 | writes down "liero" for later |
| 16:05.44 | jorgenpt | Liero is cool |
| 16:05.47 | jorgenpt | Was, atleast. |
| 16:05.47 | Gilly | liero is a finnish game :) |
| 16:05.48 | blast007 | for when I get the Pentium 200MMX from my work ;) |
| 16:05.55 | ruskie | blast007, OpenLieroX, nil and similar :) |
| 16:05.57 | blast007 | and maybe a K5 and a 486 |
| 16:05.58 | Gilly | liero is a funny word for a worm |
| 16:06.35 | ruskie | though frankly nothing beats the original liero with instant reload and nukes as the current weapon... |
| 16:06.41 | ruskie | literaly spray and pray :) |
| 16:06.56 | ruskie | though I usually showered my oponents with caltrops... that was nasty... |
| 16:07.03 | TheRedBaron | blast007: if you want old hardware - I can give you all mine |
| 16:07.20 | ruskie | <-- has some old hw as well... |
| 16:07.33 | ruskie | a p133 128mb mem that is used as a music playing box runing xmms2 :) |
| 16:07.59 | ruskie | and a few other such bits and pieces doing nothing |
| 16:09.52 | TheRedBaron | has an old 386 cpu and mobo on the shelf |
| 16:10.07 | TheRedBaron | complete with ram |
| 16:10.43 | ruskie | I have some 3 or 4 of those... |
| 16:10.51 | ruskie | am386 dx40 |
| 16:11.20 | ruskie | and it's an extremly small form factor at that... |
| 16:13.41 | TheRedBaron | nice. Mine is fairly large form factor |
| 16:16.07 | BulletCatcher | The src/other/curl/lib/ca-bundle.h file is always generated from scratch, and so ought to be removed from the Subversion repository. |
| 16:18.35 | *** join/#bzflag javawizard2539 (n=chatzill@c-24-10-227-207.hsd1.co.comcast.net) |
| 16:18.46 | ruskie | TheRedBaron, basicaly mine is in the range of a mini-itx or some such... maybe a bit bigger... but not by much :) |
| 16:20.56 | spldart | Heh.. I have some old hardware I'm about to give away too |
| 16:21.09 | spldart | But my 200mmx server stays @ work :) |
| 16:21.09 | ruskie | I ussualy keep it... |
| 16:21.41 | ruskie | if nothing else I slowly canibalise it for various bits and pieces... |
| 16:22.01 | ruskie | when I put hw away it trully is in no-working condition... |
| 16:22.15 | spldart | I keep enough to have 9 pc's running in the house @ any 1 time and extra enough to bring 2 more into operation @ a moments notice |
| 16:22.22 | spldart | The rest gets handed out :) |
| 16:22.26 | ruskie | that many? |
| 16:22.34 | spldart | Brady Bunch |
| 16:22.46 | Lukstr | two to three for me, boxes of spare parts for the rest |
| 16:22.48 | spldart | The kids need computers |
| 16:23.00 | ruskie | I have 4 desktops that need to be available... 1 fw, 1 server, 1 music box, 1 media center |
| 16:23.28 | ruskie | though from what I currently have around I could probably setup atleast 3 more... |
| 16:23.52 | ruskie | hmm 4 actually |
| 16:23.54 | spldart | Xp Pro corp comes in handy for all those installs ;) |
| 16:24.23 | spldart | that and debian |
| 16:24.39 | Lukstr | spldart, did you make an unattended XP disc? |
| 16:24.41 | ruskie | all I need is a source mage gnu/linux, xubuntu, openbsd and a geexbox live around with me :) |
| 16:25.01 | ruskie | xubuntu for my mom's craptop |
| 16:25.05 | ruskie | openbsd for firewall |
| 16:25.09 | ruskie | geexbox for media center |
| 16:25.10 | TheRedBaron | whoever yanked MSCV/VC9 - broke autogen.sh: MSVC/Makefile.am:1: required directory MSVC/VC9 does not exist. configure.ac:1248: required file `MSVC/VC9/Makefile.in' not found |
| 16:25.10 | TheRedBaron | <PROTECTED> |
| 16:25.13 | spldart | No.. It's an official corp disk that was designed to not need activation. Just enter the serial and go |
| 16:25.16 | ruskie | and source mage everywhere else |
| 16:25.40 | TheRedBaron | thats from svn |
| 16:25.41 | Lukstr | spldart, right, well you can make an unattended install with it |
| 16:25.45 | spldart | All these are dualboots |
| 16:25.55 | ruskie | no dualboot here :) |
| 16:26.00 | spldart | hehe |
| 16:26.10 | Lukstr | why would I ever want to shutdown a PC :P |
| 16:26.12 | ruskie | <-- is proud nobody around here actually has any need for a winbox... |
| 16:26.18 | spldart | oop.. I think Jeff pulled vc9 |
| 16:26.42 | spldart | I'm still forced to use windows for my video editing/encoding software |
| 16:27.13 | TheRedBaron | I remember seeing the commit - but couldn't remember who pulled it |
| 16:27.40 | Lukstr | ruskie, I have a winbox, if you're referring to this channel |
| 16:27.43 | BulletCatcher | It looks like JeffM removed MSVC/VC9 in rev 16795. |
| 16:27.48 | ruskie | Lukstr, no |
| 16:27.53 | ruskie | Lukstr, this household :) |
| 16:27.56 | Lukstr | ah |
| 16:27.56 | TheRedBaron | i'm using rev 16809 |
| 16:27.56 | Lukstr | hehe |
| 16:28.14 | Lukstr | Soon as wine starts working with my dad's accounting software he'll be un-windows |
| 16:28.14 | spldart | 16795 was the pull |
| 16:29.13 | *** join/#bzflag SpazzyMcGee (n=Spazzy_M@unaffiliated/spazzymcgee) |
| 16:29.25 | spldart | "pull VC9 cus it doesn't work, and users can just upgrade the VC8 ones because they DO work with no mods." |
| 16:30.10 | blast007 | meaning, open the VC8 .sln file ;) |
| 16:30.14 | blast007 | VC9 will upgrade it |
| 16:31.03 | TheRedBaron | blast007: are you referring to me? |
| 16:31.13 | blast007 | no |
| 16:31.20 | TheRedBaron | ok, sorry. |
| 16:31.29 | blast007 | referring to what spldart just said the line above me |
| 16:31.30 | *** join/#bzflag Saturos (n=Saturos@p57A0DF8D.dip.t-dialin.net) |
| 16:31.42 | spldart | I said nothing.. I quoted jeff |
| 16:31.53 | blast007 | well, right |
| 16:31.55 | spldart | o_o |
| 16:31.57 | spldart | hehe |
| 16:32.08 | JeffM | steals DTRemenak's fanuc manual |
| 16:38.32 | *** join/#bzflag JBdiGriz (n=JBdiGriz@adsl-75-36-46-10.dsl.sndg02.sbcglobal.net) |
| 16:38.32 | *** mode/#bzflag [+o JBdiGriz] by ChanServ |
| 16:41.07 | BulletCatcher | TheRedBaron: autogen.sh should work again after you delete the line containing "MSVC/VC9/Makefile" from the configure.ac file and the VC9 line from MSVC/Makefile.am . Don't forget to remove the backslash from the VC8 line above the latter. |
| 16:42.52 | JeffM | bahhh |
| 16:43.27 | Saturos | JeffM: got any news regarding the shirts? |
| 16:44.14 | JeffM | Saturos, they will ship tomorow |
| 16:44.15 | CIA-16 | BZFlag: 03JeffM2501 * r16810 10/trunk/bzflag/ (MSVC/Makefile.am configure.ac): silly non automatic makefiles. |
| 16:44.44 | *** join/#bzflag Think_Differentl (n=shedelj@unaffiliated/thinkdifferentl/x-838543) |
| 16:44.59 | Saturos | JeffM: ah, ok.. cool |
| 16:45.11 | Saturos | all went seamlessly? |
| 16:45.11 | JeffM | TheRedBaron, that should fix ya |
| 16:45.15 | JeffM | Saturos, no |
| 16:45.20 | JeffM | but good enough |
| 16:45.48 | Saturos | ew.. what went wrong? |
| 16:46.34 | JeffM | just some billing issues |
| 16:46.48 | JeffM | the shirts should have shipped monday |
| 16:47.00 | *** join/#bzflag TimRiker (n=timr@99.203.77.217) |
| 16:47.00 | *** mode/#bzflag [+o TimRiker] by ChanServ |
| 16:47.04 | JeffM | but they were delayed because I added stuff on to the order, and didn't have a credit card on file |
| 16:47.24 | JeffM | so there was time back and forth on getting ahold of the correct person to get that info to |
| 16:47.52 | Saturos | ok, I see |
| 16:47.54 | JeffM | the sales lady was nice enough to put them into production before she got the billing info, but it would have been faster if I had not added stuff after |
| 16:48.09 | Saturos | more hoodies? |
| 16:48.18 | JeffM | no |
| 16:48.21 | JeffM | 2 more shirts |
| 16:49.13 | Saturos | Got my navy in S? |
| 16:49.38 | JeffM | yes |
| 16:49.57 | JeffM | what is your gmail? |
| 16:50.21 | Saturos | my gmail? I never check my gmail. |
| 16:50.39 | JeffM | I just need the name, so I can send you an invite |
| 16:50.42 | JeffM | for a document |
| 16:50.53 | Saturos | saturoks@googlemail.com I believe |
| 16:51.32 | BulletCatcher | JeffM: autogen.sh still fails due to the backslash at the end of MSVC/Makefile.am line 3. |
| 16:51.38 | JeffM | bahh |
| 16:52.17 | CIA-16 | BZFlag: 03JeffM2501 * r16811 10/trunk/bzflag/MSVC/Makefile.am: picky picky picky |
| 16:52.30 | BulletCatcher | While I have your attention, the src/other/curl/lib/ca-bundle.h file is always generated from scratch, and so ought to be removed from the Subversion repository. |
| 16:52.50 | JeffM | BulletCatcher, even on windows? |
| 16:53.10 | BulletCatcher | I suspect so. The Makefile always just overwrites it. |
| 16:53.13 | *** join/#bzflag jwir3 (n=IceChat7@cs-wlc-163.cs.umn.edu) |
| 16:53.21 | JeffM | windows dons't use a makefile |
| 16:53.28 | JeffM | nor a configure |
| 16:53.34 | JeffM | it's probably there cus it is the windows one |
| 16:53.46 | JeffM | our project files don't generate code |
| 16:53.59 | *** join/#bzflag Gnurdux (n=gnurdux@c-69-251-255-231.hsd1.md.comcast.net) |
| 16:54.40 | QuantumTim | Is there any way to switch between hardware and software rendering? |
| 16:54.43 | JeffM | I'll test it tho |
| 16:54.54 | JeffM | QuantumTim, no that's up to your video driver |
| 16:54.56 | QuantumTim | it's just my fps is really low, and I can only think I've got it stuck in software mode |
| 16:54.58 | JeffM | are you in software now? |
| 16:55.06 | BulletCatcher | autogen.sh works now, btw |
| 16:55.07 | JeffM | what does it say for the GL string when bzflag starts? |
| 16:55.14 | QuantumTim | well, even on medium graphics, with lighting and shadowing off, it's really low |
| 16:55.18 | QuantumTim | umm, I'll check |
| 16:56.23 | QuantumTim | OpenGL Driver: GeForce FX 5200/AGP/SSE2 |
| 16:56.30 | JeffM | you ain't using software then |
| 16:56.38 | JeffM | you on windows or nix? |
| 16:56.38 | QuantumTim | yeah... |
| 16:56.41 | QuantumTim | doze |
| 16:56.48 | JeffM | you recently install windows again? |
| 16:56.51 | QuantumTim | nope |
| 16:56.59 | JeffM | odd |
| 16:57.02 | QuantumTim | yeah |
| 16:58.04 | JeffM | BulletCatcher, that file is needed by windows. |
| 16:58.08 | QuantumTim | I've only just started using my pc again though (been at uni with lappy), but I don't remember it being bad before |
| 16:58.09 | blast007 | try turning on everything except lighting, shadows, and energy saver. if you have some stuff disabled, it might actually run slower on newer hardware. |
| 16:58.26 | JeffM | QuantumTim, you can also install current drivers |
| 16:58.28 | QuantumTim | hmm, okay, I'll give it a go |
| 16:58.32 | BulletCatcher | JeffM: Okay, then it must obviously stay. :-) |
| 16:58.35 | QuantumTim | yeah, I was going to try that next |
| 16:59.15 | JeffM | BulletCatcher, it's just like we have a pregenerated config.h for windows too |
| 16:59.31 | JeffM | cus windows dosn't have a config step, because it's all prety standardised |
| 16:59.40 | jwir3 | hey brlcad, you there? |
| 17:01.17 | *** join/#bzflag DTRemenak (i=DTRemena@about/essy/CrazyCoder/DTRemenak) |
| 17:01.18 | *** mode/#bzflag [+v DTRemenak] by ChanServ |
| 17:04.19 | BulletCatcher | denigrates Windows for not having a complete set of development tools. |
| 17:04.32 | Lukstr | currently BZW formats are still backwards compatible? |
| 17:05.17 | JeffM | Lukstr, what do you mean? |
| 17:05.22 | QuantumTim | JeffM: It seems to be, if I turn off the radar and message pane, it goes up to about 70 (should be at 75), with them on it goes down to 35... |
| 17:05.39 | JeffM | 1.10 can not read the new objects from 2.0 |
| 17:05.55 | JeffM | QuantumTim, yeah fonts are slow ;) |
| 17:06.01 | Lukstr | right but BZFS can read old maps? |
| 17:06.15 | QuantumTim | aah, okay |
| 17:06.20 | QuantumTim | I'll try new drivers then |
| 17:06.56 | QuantumTim | hmm, but radar JeffM? |
| 17:07.10 | spldart | yeah... |
| 17:07.16 | spldart | I can run my 1.7 map on 2.0 |
| 17:07.29 | spldart | but my 2.0 map wont on 1.7 |
| 17:07.33 | Lukstr | okay |
| 17:07.42 | Lukstr | that's all I'm wondering |
| 17:08.05 | spldart | He's probably getting busy @ work like I am |
| 17:08.18 | spldart | Boss and Manager are yelling at each other again |
| 17:08.28 | QuantumTim | aah, and if I put the radar opacity up to max, it goes up to 70 again |
| 17:08.33 | QuantumTim | even with the message pane |
| 17:08.51 | BulletCatcher | spldart: That's better than having them yell at you. :-) |
| 17:09.01 | spldart | Clarification.. we don't work together hehe |
| 17:09.05 | spldart | Yes BC |
| 17:09.15 | spldart | I just hate yelling |
| 17:09.56 | BulletCatcher | Is wearing headphones an option? |
| 17:10.32 | spldart | I got my own office with doors and everything... Music through the speakers and movies are an option |
| 17:10.47 | spldart | I have more liberty than most :) |
| 17:11.14 | spldart | heck.. I have a key and the alarm codes |
| 17:11.50 | jwir3 | spldart: just do what I do when I get into a boring class - sleep |
| 17:11.50 | *** join/#bzflag Keeemo (n=Epyon@90-156-93-82.magma-net.pl) |
| 17:12.05 | spldart | Wouldn't be the first time :-0 |
| 17:12.18 | jwir3 | like right now, my iinstructor is babbliing about sporadic and periodic real time something...blah... equation... whatever |
| 17:12.32 | Epyon | mornin all |
| 17:12.43 | jwir3 | So i'm reading GSoC applications ;) |
| 17:12.46 | blast007 | QuantumTim: with the radar opacity to max, iirc, it only re-renders the message panel when it changes |
| 17:12.46 | spldart | mornin |
| 17:12.53 | TheRedBaron | I used to have a job like that. own office and walls. During lunch time i'd crawl under me desk and sleep |
| 17:12.54 | jwir3 | morning Epyon |
| 17:12.55 | blast007 | so goes back to fonts again ;) |
| 17:13.14 | QuantumTim | blast007: but even with just the radar up, and lower opacity, it has low fps |
| 17:13.23 | blast007 | hmm |
| 17:13.28 | TheRedBaron | blast007: I think with opacity set to max - it doesn't render the world behind the radar and chat box |
| 17:13.28 | QuantumTim | which suggests it isn't the fonts that are the problem |
| 17:13.40 | blast007 | TheRedBaron: that too |
| 17:13.46 | QuantumTim | also, if I have the menu up, the fps drops |
| 17:13.48 | blast007 | QuantumTim: what resolution are you running at? |
| 17:13.50 | jwir3 | QuantumTim: Are you sure that you don't have like Mesa installed, and it's defaulting to software rendering? |
| 17:14.04 | blast007 | and you have blending turned on? |
| 17:14.10 | QuantumTim | which suggests the transparency is a problem |
| 17:14.16 | blast007 | and what quality level are you using? |
| 17:14.17 | QuantumTim | yep, I've tried with it on and off |
| 17:14.19 | spldart | client reports opengl I believe was said earlier |
| 17:14.27 | QuantumTim | Medium, but High is a bit worse |
| 17:14.34 | blast007 | hrm |
| 17:14.38 | QuantumTim | I'm at 1280x1024 atm I think |
| 17:14.52 | QuantumTim | no wait |
| 17:14.53 | jwir3 | spldart: Oh yeah... missed that sorry |
| 17:14.58 | QuantumTim | I put it down to 1024x768 |
| 17:14.58 | blast007 | try experimental, with full textures and other settings (minus lighting and shadows) |
| 17:15.27 | spldart | I run one of my rigs, with 5200 fx, @ 1024x768 and it does well... hrmm |
| 17:15.50 | *** join/#bzflag Wits (n=Warinthe@1.168.188.72.cfl.res.rr.com) |
| 17:16.38 | Lukstr | JeffM, I attempted to answer your question. If it's still vague, let me know. |
| 17:17.35 | JeffM | ohhhkay |
| 17:17.51 | Lukstr | er, in regards to the app :P |
| 17:17.59 | JeffM | I kinow |
| 17:18.05 | Lukstr | yay |
| 17:18.18 | JeffM | QuantumTim, radar can be slow |
| 17:18.25 | JeffM | there are like 3 modes for radar |
| 17:18.29 | Lukstr | class over; bye all |
| 17:18.35 | QuantumTim | nope, experimental doesn't change it much |
| 17:18.36 | spldart | cyah |
| 17:18.48 | QuantumTim | JeffM: but radar with full opacity works at 70 |
| 17:19.17 | JeffM | your card may not do well |
| 17:21.03 | spldart | I'm gonna verify my 5200 when I get home... model, current driver and everything.. I think I have the same card in one of my machines |
| 17:21.08 | QuantumTim | also, there doesn't seem to be a big difference between radar, messages, and both |
| 17:21.12 | blast007 | worked well on my 5200, but I haven't tried lately ;) the fan died |
| 17:21.25 | QuantumTim | I'm getting the latest drivers now though |
| 17:21.27 | spldart | Fan had died on mine too |
| 17:21.33 | spldart | Hadda fix that heh |
| 17:21.39 | blast007 | spldart: I replaced mine once |
| 17:22.05 | blast007 | but now the new one is really hanging on... the thermal paste is like glue now :/ |
| 17:22.28 | blast007 | need to get it off to throw on the replacement |
| 17:23.05 | *** join/#bzflag cjhopman (n=cjhopman@24-183-44-113.dhcp.mdsn.wi.charter.com) |
| 17:23.05 | jwir3 | QuantumTim: Does it exhibit problems with other software? |
| 17:24.05 | BulletCatcher | FYI, I just did a full build (starting with autogen.sh) using rev 16811 and produced a working client and server. Developers should again be free to focus on the game itself. |
| 17:24.25 | QuantumTim | jwir3: not that I've noticed, but I don't think I've used many other opengl/directx software lately |
| 17:24.31 | jwir3 | hmm |
| 17:24.35 | QuantumTim | I've only been back on my pc for about a week |
| 17:24.39 | jwir3 | i see |
| 17:24.41 | JeffM | BulletCatcher, haha, there are many more bugs to fix |
| 17:24.55 | jwir3 | bbl |
| 17:25.59 | BulletCatcher | JeffM: I understand, but successful compilation is an important baseline. |
| 17:26.26 | JeffM | it was only busted for like 8 hours ;) |
| 17:27.37 | BulletCatcher | That's, like, forever in internet time. :-) |
| 17:27.41 | JeffM | you'll live |
| 17:27.50 | *** join/#bzflag tupone (n=atupone@pdpc/supporter/active/Tupone) |
| 17:27.50 | *** mode/#bzflag [+v tupone] by ChanServ |
| 17:27.54 | brlcad | jwir3: yes |
| 17:28.15 | spldart | Yay! Boss is smiling again. |
| 17:29.23 | spldart | Heh.. 8 hours... that's not even time enough to break a sweat |
| 17:29.25 | javawizard2539 | n/motd |
| 17:29.32 | blast007 | ~n |
| 17:29.33 | ibot | hmm... n is the disease bzflag players get when they play too much. There is no cure. |
| 17:29.34 | javawizard2539 | Ack, sry, didn't mean to send that |
| 17:32.04 | *** join/#bzflag zee149 (n=c35d1584@gateway/web/cgi-irc/zeebrothers.net/x-ec5e89f190c0dd5e) |
| 17:34.21 | Epyon | brlcad, JeffM, when will you have time? I'd like to discuss yesterdays "issue" more "properly" ;) |
| 17:35.00 | Epyon | finished reading through the last night logs. |
| 17:35.31 | JeffM | I'm at work now, but ok, whatever |
| 17:36.32 | brlcad | similar has very limited time for probably the next 6 hours but can inject comments |
| 17:36.56 | JeffM | hands brlcad the comment railgun |
| 17:37.03 | JeffM | that way you can do it at range |
| 17:37.04 | Epyon | I can leave that for the next 6h. |
| 17:37.23 | Epyon | feels like on a execution ground O.o |
| 17:37.31 | brlcad | i think your first step is to just get familiar with gprof and/or oprofile and/or valgrind |
| 17:37.45 | JeffM | or the intel thing that was linked last night |
| 17:38.01 | Epyon | brlcad, I've been wondering how thise things work with MSVCEE |
| 17:38.30 | Epyon | My linux box currently is a server set up in the closet so I can't count on it for bzflag. |
| 17:39.18 | Epyon | I know the full MSVC has it's own profiler, but I'd like to avoid... hmm... streching. |
| 17:42.52 | brlcad | JeffM: that was the cow being somewhat tounge-in-cheeky, that's a intel-specific utility lib -- mostly for things like high-performance lapack routines |
| 17:43.05 | QuantumTim | okay, so maybe this problem is somewhat more fundamental |
| 17:43.07 | brlcad | (at lesat with regards to math) |
| 17:43.20 | JeffM | brlcad, I was thinking of vtune |
| 17:43.31 | QuantumTim | I rebooted after installed in the gfx drivers (first time since I turned this on a few days ago) |
| 17:43.32 | brlcad | ah, yes, vtune would do something useful |
| 17:43.38 | QuantumTim | and it wouldn't boot |
| 17:43.46 | QuantumTim | got stuck on the XP loading screen |
| 17:43.52 | QuantumTim | last known good didn't work either |
| 17:43.55 | JeffM | QuantumTim, that is not good, the card may be pooping |
| 17:43.58 | QuantumTim | yeah |
| 17:43.59 | brlcad | i'd still say give gprof a try first, it's really probably the easiest to get going |
| 17:44.08 | QuantumTim | maybe I'll try reinstalling XP |
| 17:44.13 | Epyon | What about valgrind? |
| 17:44.32 | Epyon | I heard very good reports on it. |
| 17:44.34 | QuantumTim | if that doesn't work, looks like I might need a new card :s |
| 17:44.50 | JeffM | brlcad, I don't belive gprof can work with VC |
| 17:44.56 | QuantumTim | JeffM: don't suppos you know of any tools that could tell me if there is in fact a problem with the card? |
| 17:45.04 | Epyon | JeffM, yep, I googled on that |
| 17:45.29 | JeffM | QuantumTim, pull it out and replace it with another see if that helps |
| 17:45.37 | *** join/#bzflag Erroneous (n=dtremena@67-131-219-2.dia.static.qwest.net) |
| 17:45.39 | QuantumTim | heh |
| 17:45.40 | JeffM | there is little software that can diagnose hardware |
| 17:45.52 | JeffM | since the software has to run on the hardware |
| 17:46.39 | QuantumTim | this is true, but you can surely tell the hardware to do some stuff, and see if it breaks or not |
| 17:46.55 | JeffM | Epyon, if I was you I'd see what software you can get from your school |
| 17:46.57 | QuantumTim | *something* is breaking when I boot anyway |
| 17:47.00 | Epyon | Ok, neither does valgrind |
| 17:47.13 | JeffM | QuantumTim, that tells me to replace the card right there |
| 17:47.15 | QuantumTim | but it does seem to be working for the most part, other than that |
| 17:47.17 | JeffM | 5200's are old |
| 17:47.27 | JeffM | if it can't take new drivers something is bad |
| 17:47.51 | QuantumTim | This is true |
| 17:48.36 | Epyon | JeffM, I can't anymore -- at our uni once you graduate you don't have access to MSDN |
| 17:48.53 | JeffM | Epyon, so you are no longer a student? |
| 17:49.04 | Epyon | JeffM, postgraduate |
| 17:49.12 | JeffM | and they do not have a program for them? |
| 17:49.16 | Epyon | Nope :( |
| 17:49.29 | JeffM | well then looks like you have to decide what you want to do |
| 17:49.32 | Epyon | That's what I was told when I wanted a Vista for testing purposes :( |
| 17:49.37 | brlcad | JeffM: oh .. didn't realize he's on windows |
| 17:49.41 | JeffM | yeah he is |
| 17:49.57 | JeffM | he's basicly got numega and vtune |
| 17:50.04 | JeffM | or full VC |
| 17:50.08 | Epyon | brlcad, if I were on Linux, I wouldn't ask at all about profiling, I'd just read the man's. |
| 17:50.30 | JeffM | install linux in a VM ;) |
| 17:50.31 | Epyon | brlcad, all the programs you suggested are linux only |
| 17:50.34 | brlcad | so vtune or msvc |
| 17:50.41 | JeffM | or numega |
| 17:50.52 | brlcad | of course they are -- I thought you were on linux |
| 17:51.00 | JeffM | devpartner is fun :) |
| 17:51.08 | Epyon | vtune is commercial? |
| 17:51.10 | brlcad | were you on mac, i'd make even a diff set of suggestions |
| 17:51.21 | JeffM | Epyon, they all are |
| 17:51.52 | Erroneous | do not profile in a VM |
| 17:52.00 | Epyon | That's what I did suppose :/ |
| 17:52.01 | JeffM | heh |
| 17:52.02 | Erroneous | it makes it lie :) |
| 17:52.17 | JeffM | tho devpartner has a community eddtion |
| 17:52.21 | JeffM | Epyon, google around a bit |
| 17:52.42 | Erroneous | setting up a minimal linux install and using oprofile is probably easiest |
| 17:52.47 | Epyon | Funny, in the heat of the "argument" yesterday, it didn't come out that I'm on windows. |
| 17:53.03 | Erroneous | and then you can do valgrind too, if you so desire |
| 17:53.05 | JeffM | http://www.compuware.com/products/devpartner/profiler/default.asp?cid=701000000004ej0AAA&focus=&productfocus=&source=Web+-+Evaluation+Request&offering=DevPartner&productfamily=DevPartner&desc=The+DevPartner+Community+Profiler+Edition+is+a+product+download+that+offers+customers+and+prospects+a+high-level+view+of+the+performance+profiling+capabilities+of+DevPartner+Studio+Professional+Edition.&trk=200601-74 |
| 17:53.08 | Erroneous | both are very cool tools |
| 17:53.08 | JeffM | wow |
| 17:53.11 | JeffM | ok that was too big |
| 17:53.12 | Erroneous | holy cow that is a huge url |
| 17:53.20 | JeffM | sorry |
| 17:53.24 | JeffM | didn't realise |
| 17:53.39 | JeffM | "Compuware DevPartner Performance Analysis Community Edition accelerates software development and improves the quality of your code by pinpointing performance bottlenecks in native Windows C/C++ and .NET applications. Whether you're creating new applications using the latest .NET technologies, or are working both with .NET and legacy components such as COM, COM+ or ASP, DevPartner Performance Analysis Community Edition is the profiler of choice for deta |
| 17:53.39 | JeffM | iled and accurate timing results." |
| 17:53.57 | JeffM | 45 day license |
| 17:54.21 | Erroneous | doesn't devpartner require manual code instrumentation? |
| 17:54.24 | Epyon | JeffM, I think it's full VC only anyway |
| 17:54.29 | JeffM | Erroneous, dunno |
| 17:54.33 | JeffM | Epyon, probably |
| 17:54.47 | Epyon | So I'm back at base #1 |
| 17:54.49 | JeffM | Epyon, that is why I said you have to decide what to do |
| 17:54.53 | JeffM | you have a number of options |
| 17:55.10 | Epyon | 1) steal, 2) blow up my PC :P |
| 17:55.14 | JeffM | god no |
| 17:55.32 | JeffM | spend money, install nix, or write some basic profiling code |
| 17:55.35 | *** part/#bzflag pacman_87 (n=timothy@nat-202-129.arlut.utexas.edu) |
| 17:55.47 | *** part/#bzflag BulletCatcher (n=bc@unaffiliated/bulletcatcher) |
| 17:55.49 | JeffM | all are viable options |
| 17:55.57 | Erroneous | or a usb thumbdrive and a linux livecd |
| 17:56.07 | Erroneous | then you don't have to worry about messing anything up on your pc |
| 17:56.11 | JeffM | that would fall under a variant of install linux ;) |
| 17:56.14 | Epyon | Translation : 1) "aquire" one of the needed win apps, 2) find a way to break up space on one of my harddrives and instal linux |
| 17:56.26 | JeffM | Epyon, no you simplify too much |
| 17:56.33 | JeffM | there are many ways to do each of them |
| 17:56.41 | JeffM | step back and take some time to evaluate them |
| 17:56.49 | JeffM | there is no rush |
| 17:56.55 | Erroneous | seriously, a usb thumbdrive and a linux livecd is not a bad option |
| 17:56.55 | Epyon | There is. |
| 17:57.08 | JeffM | Epyon, why is there a rush? |
| 17:57.19 | blast007 | Erroneous: I did that before when I was saiting for the hard drives for my server to arrive ;) |
| 17:57.31 | Epyon | I need to decide on wheter I want to go with the app I thought till the end of the app time and there's also the patch need :P |
| 17:57.37 | blast007 | though I actually set up an old K6/2 as a NAS, and did it over the network instead of USB |
| 17:57.42 | JeffM | Epyon, you can do another patch |
| 17:57.44 | Erroneous | yeah, I've done it on PCs where I don't have permission to repartition too |
| 17:58.04 | JeffM | and the choice for the project is simple, you know what you have to do, so you can decide if you want to do it. |
| 17:58.07 | Erroneous | mount /home onto the usb drive and it's all good, if a bit slow |
| 17:59.08 | JeffM | the patch you do does not have to be in the section of code you want to work in |
| 17:59.12 | spldart | mmMMmm live cd's for the win.. sometimes :) |
| 17:59.16 | JeffM | it's just to show you can follow our standards |
| 17:59.31 | Epyon | In the long run I think I would swap my PC for the server for profiling work. But that would need some extra work to set it all up. |
| 17:59.32 | JeffM | Epyon, you could fix a small bug, or even make a small change, write a plugin |
| 17:59.52 | *** join/#bzflag moriah (n=d8ada652@gateway/web/cgi-irc/zeebrothers.net/x-301bdad863f7fac3) |
| 18:00.12 | Erroneous | thought plugins were out of scope for yon patch requirement |
| 18:00.17 | JeffM | they are? |
| 18:00.28 | Epyon | JeffM, just like external apps are :P |
| 18:00.30 | JeffM | it's not mentioned |
| 18:00.37 | Erroneous | the idea is to get the applicant comfortable with preexisting internal code, right? |
| 18:00.43 | *** part/#bzflag moriah (n=d8ada652@gateway/web/cgi-irc/zeebrothers.net/x-301bdad863f7fac3) |
| 18:00.58 | JeffM | Erroneous, I thought it was to just show that they can follow our standards, and have built and tested, etc.. |
| 18:01.03 | JeffM | probalby should mention that on the page then |
| 18:01.09 | Erroneous | don't have to build to do a plugin |
| 18:01.18 | JeffM | plugin in trunk ;) |
| 18:01.21 | Erroneous | heh |
| 18:01.38 | Erroneous | good point |
| 18:01.48 | Epyon | JeffM, I still need however to decide if I want to apply for the "cleanup" app. Yesterdays story made me real scared of touching the mainline, and the idea of rewriting huge parts of it scares the hell out of me now :P |
| 18:02.55 | Erroneous | all you have to do is evaluate your own work. and we will understand if you spend time making changes, find out that they don't perform like you expected them to, and back them out. that's part of the normal process too. |
| 18:03.16 | QuantumTim | meh, this computer's quite old anyway (the gfx card was an upgrade, so newer than most of it!) |
| 18:03.17 | Erroneous | even if it is depressing sometimes :) |
| 18:03.33 | QuantumTim | I might as well just get a new computer all together =) |
| 18:04.45 | Erroneous | it's pretty much impossible to break trunk permanently. we've only managed that once. |
| 18:05.42 | Erroneous | so when you make changes to trunk, you just say, hey, here's the change, here's the tangible benefit, here's the impact as compared to current, and decide whether it's worth it or not. |
| 18:08.38 | JeffM | Erroneous, I'm cool with saying no plugins, I think it should just be clarified on the wiki :) |
| 18:08.48 | *** join/#bzflag KTL (n=KTL@213.219.144.106.adsl.dyn.edpnet.net) |
| 18:09.02 | JeffM | oh I may be on the wrong page |
| 18:09.10 | JeffM | nope I was |
| 18:09.32 | JeffM | all the "submit a patch" item does is link to the tracker |
| 18:10.14 | *** join/#bzflag KTL2 (n=KTL@213.219.144.106.adsl.dyn.edpnet.net) |
| 18:23.03 | *** join/#bzflag KingofCamelot (n=davidsan@207-177-229-17.dsl.redshift.com) |
| 18:24.52 | JBdiGriz | JeffM: Software which can diagnose hardware issues is often called "diagnostics" |
| 18:25.08 | JBdiGriz | is just catching up on the conversation |
| 18:25.41 | Epyon | Where's the most "uptodate" buglist? I doubt that the answer is sf tracker :P |
| 18:31.16 | JBdiGriz | It's certainly one of the most "uptodate" buglists. People tend to report bugs in many places, but they don't always make it into the buglists. The BUGS file is also a good place to go looking. |
| 18:32.58 | javawizard2539 | Does anyone think it would be useful to add an option to the options menu to choose whether or not you automatically pause when your window is minimized? |
| 18:35.37 | KTL | mentions that in some cases the game does not execute its main loop anymore, i thought when the game was minimized, -> lagkick. |
| 18:35.48 | javawizard2539 | This seems particularly useful for developing, so you could start 2 clients and try out different flags on one from the other |
| 18:37.59 | *** join/#bzflag moriah (n=d8ada652@gateway/web/cgi-irc/zeebrothers.net/x-bfb4cd37a00284e5) |
| 18:40.30 | Epyon | JeffM, how many of the applicants have already sent in a patch? |
| 18:41.36 | *** join/#bzflag jwir3 (i=joh03135@pepper.itlabs.umn.edu) |
| 18:44.46 | *** join/#bzflag moriah (n=d8ada652@gateway/web/cgi-irc/zeebrothers.net/x-457196231c1bea4a) |
| 18:44.55 | *** join/#bzflag Upsetter (n=scratch@e178091178.adsl.alicedsl.de) |
| 18:47.04 | Manu | JeffM: Let the introductions end ;) |
| 18:47.26 | jwir3 | Lol |
| 18:47.38 | jwir3 | I agree with that sentiment |
| 18:47.52 | *** part/#bzflag moriah (n=d8ada652@gateway/web/cgi-irc/zeebrothers.net/x-457196231c1bea4a) |
| 18:48.05 | Manu | probably sounds bad out of context jwir3 ;) |
| 18:48.09 | javawizard2539 | If I have a suggestion on a small change to bzflag, where should I post it? |
| 18:48.24 | Epyon | sf tracker? |
| 18:48.30 | javawizard2539 | It's way old |
| 18:48.42 | Manu | javawizard2539: what about here? :) |
| 18:48.47 | javawizard2539 | Ok |
| 18:48.52 | jwir3 | Manu: You mean the gsoc mentor list introductions? |
| 18:48.56 | javawizard2539 | My suggestiong was to change the description for geno |
| 18:49.09 | javawizard2539 | to Killing one tank kills that tank's whole team. You get one point for each member of the team you kill. |
| 18:49.14 | Manu | jwir3: yes, a introduction message |
| 18:49.20 | Manu | an* |
| 18:49.29 | jwir3 | Manu: :) I knew what you were talking about haha |
| 18:49.35 | javawizard2539 | This makes it clear that you get one point for each member, not just one point for the tank you actually killed |
| 18:50.22 | Manu | javawizard2539: wasn't enough clear earlier? |
| 18:50.42 | javawizard2539 | It said Killing one tank kills that tank's whole team |
| 18:50.51 | Manu | yes |
| 18:50.55 | javawizard2539 | Manu, which doesn't make it clear whether you get just one point or multiple points |
| 18:51.00 | Manu | and one kill is one point except for teamkilling |
| 18:51.16 | QuantumTim | javawizard2539: as soon as you use it you'll find out how it gives you points =) |
| 18:51.21 | javawizard2539 | I have |
| 18:51.25 | QuantumTim | unless you use it on a person with no teammates ;) |
| 18:51.38 | javawizard2539 | I know how it works, I'm just saying that it would reduce initial confusion |
| 18:51.41 | Manu | javawizard2539: I find your description a bit ofuscated :/ |
| 18:51.48 | javawizard2539 | Why? |
| 18:51.57 | QuantumTim | and somewhat verbose ;) |
| 18:52.25 | Manu | javawizard2539: my point of view.. |
| 18:52.33 | Manu | yes |
| 18:53.01 | javawizard2539 | So it's a good idea? |
| 18:53.17 | Manu | is like to say Laser, kills al the tank in its trajectory. And plus a point for every tank killed |
| 18:53.38 | javawizard2539 | Hmm |
| 18:53.42 | Manu | the point for every kill is no needed I guess |
| 18:54.03 | Manu | because it's implicit 1 kill one point |
| 18:54.08 | javawizard2539 | I still think there should be something in the description that says you get one point for each tank killed total, not just for the one you actually shoot |
| 18:54.46 | Manu | then we have to change every description :) |
| 18:54.55 | Manu | same for Laser, SW.. |
| 18:55.00 | javawizard2539 | Whast do you mean, every descriptiuon? |
| 18:55.18 | Manu | SW, also kills all the tanks in its action range |
| 18:55.27 | javawizard2539 | It says that in the description |
| 18:55.30 | javawizard2539 | "all tanks nearby" |
| 18:55.36 | QuantumTim | but it doesn't mention points :p |
| 18:55.39 | Manu | so following you, we must add and plus one point for every killed tank |
| 18:55.54 | javawizard2539 | and since you're physically impacting those tanks, it's impied you get points for each tank killed |
| 18:56.11 | Manu | a bit redundant because the points are implicit to the kills |
| 18:56.19 | QuantumTim | I think you'd only be confused if you equated points with "physical hits" |
| 18:56.27 | QuantumTim | but I doubt many people do |
| 18:56.36 | javawizard2539 | When you capture a team's flag, all of the tanks die but you don't get a point for each |
| 18:56.48 | javawizard2539 | So it's natural to assume similar for geno |
| 18:57.20 | Manu | javawizard2539: no die, just restarted |
| 18:57.50 | javawizard2539 | They die |
| 18:58.01 | Manu | it's different to be killed |
| 18:58.05 | javawizard2539 | sry, not in the literal bzflag code sense |
| 18:58.32 | javawizard2539 | A non-developer sees no difference between the two except that no points are awarded and the tank starts back on base |
| 18:58.36 | Manu | javawizard2539: but can be good to add points every time you capture ;) |
| 18:59.33 | Manu | javawizard2539: if you do capture, you don't kill so no points for the players |
| 18:59.40 | Manu | that's the difference |
| 18:59.47 | Manu | server weapons also kills |
| 18:59.52 | QuantumTim | javawizard2539: if you're confused, you'll find these things out as soon as you try them |
| 18:59.54 | Manu | but points are for nobody |
| 19:00.11 | javawizard2539 | Fine, I'm just trying to make it easier for newcomers to the game to figure out how it works |
| 19:00.49 | Manu | you mean the decription in the bzflag client javawizard2539 ? |
| 19:00.54 | QuantumTim | but it's better to give them the basics, rather than flood them with things that aren't really relevant, surely? |
| 19:00.54 | javawizard2539 | Yes |
| 19:01.06 | javawizard2539 | What shows up under the mousebox |
| 19:01.58 | Manu | from my point of view, descriptions in the client must be short and concises |
| 19:02.10 | *** join/#bzflag moriah (n=d8ada652@gateway/web/cgi-irc/zeebrothers.net/x-cb17a77c46d4c12f) |
| 19:02.17 | javawizard2539 | That is |
| 19:02.22 | Manu | but other devs can talk about it |
| 19:02.23 | javawizard2539 | It's one heck of a lot shorter than PZ |
| 19:02.28 | javawizard2539 | Ok |
| 19:02.48 | Manu | I agree with an extended info in the wiki, for example |
| 19:03.06 | javawizard2539 | Ok |
| 19:04.24 | javawizard2539 | Another question. If you have geno, and you shoot, then die, then someone gets hit with your shot, it doesn't count as a geno shot. Is there a reason it was done this way? |
| 19:04.46 | Manu | it doesn't I don't thin so ;) |
| 19:05.14 | Manu | you can even be kicked for using geno on your own team and players still die after you have been kicked |
| 19:06.05 | javawizard2539 | It would be trivial to change it so that geno has it's own ShotType, so that each shot carries whether it's a geno shot or not, instead of the tank itself carrying this information |
| 19:06.14 | javawizard2539 | JeffM, what would your opinion be on this? |
| 19:06.30 | QuantumTim | it already works like that, doesn't it? |
| 19:06.36 | javawizard2539 | No |
| 19:06.48 | blast007 | even in trunk? |
| 19:06.59 | QuantumTim | it must, because some servers have a 1 shot limit |
| 19:06.59 | javawizard2539 | I'm looking at trunk right now |
| 19:07.07 | QuantumTim | and you drop the flag as soon as you've shot |
| 19:07.25 | javawizard2539 | But if you die before the shot hits, it doesn't count as geno |
| 19:07.39 | QuantumTim | why does dying make a difference to the shot? |
| 19:07.52 | blast007 | QuantumTim: probably lame logic |
| 19:08.10 | javawizard2539 | Geno doesn't have it's own ShotType |
| 19:08.14 | javawizard2539 | It seems like it should |
| 19:08.40 | QuantumTim | blast007: sounds like it, but it's definitely not to do with what flag the tank has |
| 19:09.20 | javawizard2539 | It seems like the best way to go about doing geno would have been to create a new ShotType (such as GenoShot) for it |
| 19:09.28 | QuantumTim | I can't say I remember having shot geno and then dying, but it would be some really strange logic if it did break like you're suggesting |
| 19:09.30 | Manu | javawizard2539: so if you have geno, shoot your last bullet and drop the geno flag.. the bullet doesn't kill the full team? :) |
| 19:09.41 | javawizard2539 | I don't know about that |
| 19:10.04 | Manu | javawizard2539: AFAIK geno works even after you get killed |
| 19:10.05 | javawizard2539 | I think so, but I can't remember. What I do remember is that if you shoot, die, and your shot hits someone, it doesn't kill their entire team |
| 19:10.42 | Manu | at last when you are kicked it works |
| 19:10.42 | javawizard2539 | Hmm |
| 19:10.43 | QuantumTim | It definitely works after you drop the flag |
| 19:10.44 | Manu | yes, even when you are kicked |
| 19:10.44 | QuantumTim | javawizard2539: how extensively have you tested this? |
| 19:10.44 | javawizard2539 | So it's probably something to do with dying specifically |
| 19:10.45 | javawizard2539 | A lot |
| 19:10.53 | Manu | then the server message is: killed by [Unknown] |
| 19:10.54 | javawizard2539 | Hold on |
| 19:11.02 | QuantumTim | but if it is to do with dying, that's just messed up |
| 19:11.25 | Manu | javawizard2539: actually you can tk yourself with genocide :) |
| 19:11.35 | Manu | then you are dead and still kills your full team |
| 19:11.37 | javawizard2539 | So now I find out that I've been testing against 2.0.10 |
| 19:11.48 | javawizard2539 | That could be the issue |
| 19:11.50 | QuantumTim | if shot.isgeno and tank.dead then shot.isgeno = false ;) |
| 19:11.58 | brlcad | Epyon: if getting a profiler in your environment is a problem, the other "task" involved just writing a test case with a simple timer (using gettimeofday() on unix or time() if you make it granular enough (like a 2+min) or even using bz's TimeKeeper class) and directly sample cvec, Vector, and plain |
| 19:12.11 | Manu | QuantumTim: heh |
| 19:12.16 | javawizard2539 | That's in the code? |
| 19:12.23 | brlcad | that test app shouldn't take more than a half hour to whip out |
| 19:12.32 | brlcad | hola manu, que tal? |
| 19:12.35 | QuantumTim | well, it sounds like something similar from what you're saying |
| 19:12.44 | Manu | brlcad: bien, y tu? :) |
| 19:12.58 | javawizard2539 | consults the code in trunk |
| 19:13.10 | brlcad | bien, que hora son por alli? |
| 19:13.14 | Manu | QuantumTim: shot.isgeno = false is an assignation not a comparation ;) |
| 19:13.27 | Manu | brlcad: las 20:13 :) |
| 19:13.49 | QuantumTim | Manu: I meant it as an assignment |
| 19:13.50 | brlcad | mm.. that's right .. UMT+0? |
| 19:13.57 | Manu | +1 |
| 19:14.12 | brlcad | ah, verdad.. stupid daylight savings |
| 19:14.20 | QuantumTim | but in any case, what language I'm using isn't clear (note the use of and rather than &&), so I guess it could be either ;) |
| 19:14.23 | Manu | UTC is 19:14 |
| 19:14.45 | Manu | QuantumTim: in objectpascal = is for comparation :) |
| 19:14.57 | QuantumTim | true, true |
| 19:15.05 | QuantumTim | if shot.isgeno and tank.dead then shot.isgeno := false ;) |
| 19:15.19 | javawizard2539 | QuantumTim, it appears that it's been fixed in 2.99 |
| 19:15.33 | javawizard2539 | and that my problem was that I was testing on 2.0.10 |
| 19:16.05 | Manu | brlcad: no habras leido todos los mensajes de la lista de mentores no? :) |
| 19:16.22 | brlcad | no todos |
| 19:16.29 | brlcad | pero casi todos.. |
| 19:16.39 | Manu | y continuan entrando... |
| 19:16.40 | *** join/#bzflag mibocote (n=matt@x-134-84-99-225.reshalls.umn.edu) |
| 19:16.48 | brlcad | por lo menos los de la mañana |
| 19:16.57 | brlcad | yeah, something like 300+ now |
| 19:16.59 | Manu | yo lei los primeros.. |
| 19:17.19 | Manu | cuando veo algun nombre conocido, lo leo :) |
| 19:18.14 | brlcad | pero los mensajes son cortos, y mucho más interesante que los mensajes en el -discuss list |
| 19:18.37 | Manu | si, es una manera de conocer gente |
| 19:19.04 | brlcad | algunos dan interes |
| 19:22.53 | *** join/#bzflag moriah (n=d8ada652@gateway/web/cgi-irc/zeebrothers.net/x-170cd15fea030f0a) |
| 19:24.42 | blast007 | hello again mibocote |
| 19:25.12 | JeffM | huh wqhat? |
| 19:25.15 | JeffM | I was out having lunch |
| 19:25.55 | JeffM | javawizard2539, what am I commenting on? |
| 19:26.57 | JeffM | geno is a flag type |
| 19:26.57 | javawizard2539 | JeffM, nvm |
| 19:27.04 | JeffM | and you can apply the flag type to any shot |
| 19:27.15 | javawizard2539 | I know, I found that out just a second ago |
| 19:27.18 | JeffM | ahh |
| 19:27.40 | JeffM | yeah eventualy even the flag effects will be pulled out seperate from the flags and then you can mix and match as needed |
| 19:27.50 | javawizard2539 | Ok |
| 19:28.35 | JeffM | you can have Genolazer with the weapon arena plug-in :) |
| 19:28.36 | *** join/#bzflag tupone (n=atupone@pdpc/supporter/active/Tupone) |
| 19:28.36 | *** mode/#bzflag [+v tupone] by ChanServ |
| 19:28.36 | JeffM | it's mean |
| 19:28.58 | Manu | :) |
| 19:33.13 | javawizard2539 | Another question about a description change |
| 19:33.32 | javawizard2539 | It seems like the description for seer should include that you can also see invisible bullets as normal |
| 19:33.37 | javawizard2539 | Does anyone agree with this? |
| 19:33.47 | JeffM | if it's true sure |
| 19:34.39 | *** join/#bzflag noyb (n=noyb@pool-71-119-173-104.lsanca.dsl-w.verizon.net) |
| 19:34.53 | javawizard2539 | if (shot && (shot->getShotType() != InvisibleShot || iSeeAll)) { |
| 19:35.07 | javawizard2539 | It is true |
| 19:35.17 | JeffM | then there ya go |
| 19:35.50 | javawizard2539 | Someone else would need to change this because I don't have commit access and it seems like a bit too small to submit as a patch |
| 19:38.44 | CIA-16 | BZFlag: 03dtremenak * r16812 10/trunk/bzflag/src/common/Flag.cxx: seer sees invisible bullets |
| 19:40.05 | CIA-16 | BZFlag: 03dtremenak * r16813 10/trunk/bzflag/src/common/Flag.cxx: LG doesn't just make you jump higher. |
| 19:40.27 | brlcad | JeffM: so I did finally pinpoint that bug, all the other leads I was following about backing out were basically false leads .. |
| 19:40.56 | brlcad | i think there is highly likely still some static initialization corruption, because I can't get a clean debug trace, but it does work now for me at least again |
| 19:42.57 | *** join/#bzflag torghul (n=chatzill@88.251.66.16) |
| 19:44.37 | JeffM | brlcad, did you find out what it could be? |
| 19:44.45 | *** join/#bzflag moriah (n=d8ada652@gateway/web/cgi-irc/zeebrothers.net/x-aa0a57c5278ad5f8) |
| 19:44.48 | JeffM | javawizard2539, what is your sourceforge name? |
| 19:45.02 | javawizard2539 | javawizard2539 |
| 19:45.08 | Erroneous | heh |
| 19:45.10 | javawizard2539 | My real name is Alexander Boyd |
| 19:45.40 | brlcad | JeffM: for the crash yes, it was in the commit message -- had to do with the moon rendering at high-quality .. bad state |
| 19:46.04 | brlcad | someone took out the moon texture, but didn't change the state |
| 19:46.22 | JeffM | javawizard2539, don't break anything ;) |
| 19:46.49 | JeffM | brlcad, hah |
| 19:46.55 | JeffM | misses the textured moon |
| 19:47.11 | TheRedBaron | moons JeffM |
| 19:47.17 | brlcad | for the static, no, don't know wth that might be going on -- I need to go old-school printf's I think in the singletons for starters or something |
| 19:47.21 | Manu | javawizard2539: if you break something go for the glue ;) |
| 19:47.40 | JeffM | not the singletons.... |
| 19:47.45 | JeffM | loves the singletons |
| 19:47.47 | JeffM | and managers |
| 19:47.52 | JeffM | ohh the mangers |
| 19:48.09 | javawizard2539 | So I get commit access? |
| 19:48.12 | JeffM | not get |
| 19:48.13 | JeffM | have |
| 19:48.15 | Erroneous | yup, congrats |
| 19:48.16 | JeffM | what is your bzbb name |
| 19:48.17 | javawizard2539 | Cool! |
| 19:48.21 | javawizard2539 | same |
| 19:48.24 | javawizard2539 | javawizard2539 |
| 19:48.36 | javawizard2539 | I use the same name just about everywhere |
| 19:48.48 | Erroneous | wonders aloud about the origins of the numbers |
| 19:49.20 | Erroneous | 'tis prime |
| 19:49.34 | brlcad | javawizard2539: you should commit something now to make sure you're good to go |
| 19:49.52 | JeffM | javawizard2539, just follow the guidelines, try not to break stuff, and never be afraid to ask :) |
| 19:50.05 | JeffM | yeah make sure I did it right |
| 19:50.18 | Erroneous | wonder if there's something significant about the 380th prime that I don't know though :) |
| 19:51.09 | *** join/#bzflag Suspect (n=Suspect@68-113-163-117.dhcp.plt.ny.charter.com) |
| 19:52.33 | *** join/#bzflag NyadJ55 (n=jason@dsl-242-81-95.telkomadsl.co.za) |
| 19:53.13 | javawizard2539 | brb in about 5 minutes, I'll commit something then |
| 19:54.00 | javawizard2539 | The numbers are my name on a phone |
| 19:54.05 | javawizard2539 | 2-5-3-9, A-l-e-x |
| 19:54.33 | Manu | your password too? ;) |
| 19:54.48 | NyadJ55 | how much bandwidth does a single game of bzflag use? |
| 19:54.58 | NyadJ55 | well like mb, cap |
| 19:54.58 | brlcad | NyadJ55: depends how many players |
| 19:55.00 | Manu | NyadJ55: it depends |
| 19:55.01 | Erroneous | ahhh |
| 19:55.01 | brlcad | and the map type |
| 19:55.12 | NyadJ55 | average for a large game? |
| 19:55.19 | L4m3r | number of shots too |
| 19:55.29 | Manu | brlcad: and the players bandwidth, right? :) |
| 19:55.30 | brlcad | wonders where the performance graphs from way back when are at |
| 19:55.47 | Erroneous | NyadJ55: typical rule-of-thumb is 4kilobytes per player-second |
| 19:56.14 | Manu | thinks 4k is too optimist |
| 19:56.26 | brlcad | moves bb_down and bb_redir into DeleteMe |
| 19:56.32 | NyadJ55 | woah! |
| 19:56.47 | Erroneous | Manu: that's as-measured, averaging 16 players and 10 shots |
| 19:57.13 | Erroneous | it doesn't scale linearly; I had a regression equation for it somewhere |
| 19:57.18 | javawizard2539 | Manu, what about it? |
| 19:57.27 | Manu | lemme see |
| 19:57.38 | Erroneous | but if you figure 4kb/p/s you'll be close for most servers |
| 19:57.47 | javawizard2539 | My password is a bunch of characters not even related to that |
| 19:58.18 | Erroneous | it's higher if you have lots of shots or a high _wingsJumpCount |
| 19:58.35 | Manu | Erroneous: you are right.. |
| 19:58.52 | Manu | in viper 10 players, 2.7K max |
| 19:59.03 | NyadJ55 | so that means 4000bytes*30players*2hours*60minPerHour*60secPerMin = 864MB!!!! |
| 19:59.38 | ruskie | 4096bytes to be precise |
| 19:59.41 | Manu | Erroneous: you talk about 4096 bytes? |
| 19:59.49 | brlcad | that's disconcerting .. can't find the network utilization graphs |
| 19:59.53 | Erroneous | Manu: yeah |
| 19:59.57 | brlcad | anyone remember that link? |
| 20:00.00 | Erroneous | outbound, per player, per second |
| 20:00.22 | Manu | yes |
| 20:01.37 | javawizard2539 | JeffM, Is my new flag (Cloaked Shot) approved? |
| 20:01.42 | Erroneous | NyadJ55: yeah, and it's actually higher than that when you get to 30 players. remember that it scales roughly as n^2. |
| 20:01.46 | javawizard2539 | In otherwords, could I commit it? |
| 20:02.04 | Manu | javawizard2539: cloaked shot? |
| 20:02.06 | brlcad | yeah, it's non-linear .. |
| 20:02.16 | NyadJ55 | javawizard2539: wow that's scary. 1GB has to last me the whole month! |
| 20:02.23 | ruskie | Manu, yeah... like cloak :) but for a bullet |
| 20:02.28 | NyadJ55 | from 1 to 30 |
| 20:02.48 | javawizard2539 | Stealth : Cloaking as Invisible Bullet |
| 20:02.51 | javawizard2539 | : CloakedShot |
| 20:02.59 | Manu | I don't like the idea, but because flags are optionals in the maps, there is no problem for me :) |
| 20:03.05 | ruskie | :) |
| 20:03.18 | ruskie | well it brings an edge to radar-only players :) |
| 20:03.30 | Manu | yes, that's a good point |
| 20:03.34 | Erroneous | javawizard2539: IIRC there were some devinfo violations. if those have been fixed, yeah, it's been approved. |
| 20:03.42 | Manu | almost only uses the radar while playing |
| 20:03.44 | *** part/#bzflag NyadJ55 (n=jason@dsl-242-81-95.telkomadsl.co.za) |
| 20:04.01 | javawizard2539 | IIRC, they were mostly with indentations and placement of newlines, correct? |
| 20:04.03 | ruskie | hmm would be fun to have a map with only ib and cs flags... |
| 20:04.04 | Erroneous | yeah |
| 20:04.06 | javawizard2539 | I fixed those |
| 20:04.09 | Erroneous | cool |
| 20:04.31 | Erroneous | go for it then :) |
| 20:04.31 | javawizard2539 | I'll check through it one last time and then commit it. |
| 20:04.46 | javawizard2539 | Oh wait |
| 20:04.51 | Erroneous | and it did include manpage changes also, right? |
| 20:04.54 | javawizard2539 | Is there anywhere I need to include documentation on the flag? |
| 20:04.57 | javawizard2539 | Ok |
| 20:05.10 | javawizard2539 | No, completely forgot |
| 20:05.12 | Manu | look at the man pages javawizard2539 :) |
| 20:05.14 | javawizard2539 | I'll work on that |
| 20:05.42 | javawizard2539 | What folder should I be looking under? |
| 20:05.46 | Erroneous | man |
| 20:05.53 | Manu | :) |
| 20:06.09 | Erroneous | bzflag.6.in, and maybe bzfs.6.in. |
| 20:06.45 | Erroneous | if you haven't edited manpages before, they're just plain text files. the formatting codes are wierd but you can just copy-n-paste from somewhere similar :) |
| 20:06.56 | javawizard2539 | Ok |
| 20:07.16 | javawizard2539 | Does it matter what type of newlines (<CR><LF> or <LF>) I use? |
| 20:07.23 | javawizard2539 | I'm on a windows box |
| 20:07.25 | Manu | bzflag.6.in in the brief description section |
| 20:07.28 | brlcad | cries, but keeps searching for the graph prettyness |
| 20:08.39 | Erroneous | javawizard2539: they're checked in as plain text files, so the conversion should be automatic. subversion rocks. |
| 20:08.39 | Manu | javawizard2539: IIRC there is no flag info in bzfs.6.in |
| 20:09.12 | Erroneous | javawizard2539: if you haven't already, please make sure your subversion autoprops are set correctly before you check in new files |
| 20:09.42 | Erroneous | http://my.bzflag.org/w/Auto-props |
| 20:11.10 | brlcad | neat |
| 20:11.23 | Manu | found it brlcad ? |
| 20:12.22 | javawizard2539 | I'm searching around, but I'm not finding any sort of config file |
| 20:12.33 | javawizard2539 | I'm using the regular svn command line tools on windows vista |
| 20:13.00 | Erroneous | interesting. I have no idea where it stores config :) |
| 20:13.13 | Manu | hasta la vista baby.. :) |
| 20:13.37 | Erroneous | you should be fine for editing existing files, but if you check in new ones from windows without setting those properly it'll screw the *nix people up :) |
| 20:14.14 | brlcad | yay |
| 20:14.20 | brlcad | manu: yes, I just did |
| 20:14.26 | brlcad | it was on the old db server |
| 20:14.29 | Manu | I see :) |
| 20:14.30 | Erroneous | javawizard2539: try C:\Users\*yourusernamehere*\AppData\Roaming\Subversion |
| 20:15.01 | brlcad | db hasn't been on-line in couple years |
| 20:15.10 | brlcad | i should see what else is orphaned there |
| 20:16.54 | Manu | brlcad, in bzflag.org site, go down and press the link: old news |
| 20:16.55 | Manu | :) |
| 20:17.30 | Manu | 404 |
| 20:18.01 | javawizard2539 | Erroneous, Thanks. I added that to the wiki page on autoprops as well. |
| 20:18.06 | brlcad | thinks manu should fix that |
| 20:18.17 | Erroneous | cool |
| 20:18.17 | JeffM | didn't Constitution fix and apply that patch for the CS? |
| 20:19.18 | Erroneous | the patch is not marked as closed |
| 20:19.25 | JeffM | m is the code in there? |
| 20:19.25 | Erroneous | so if he applied it he didn't finish the job |
| 20:19.39 | *** join/#bzflag Think_Differentl (n=shedelj@unaffiliated/thinkdifferentl/x-838543) |
| 20:19.43 | JeffM | but yeah, I'd say it's approved |
| 20:19.51 | JeffM | however it needs to get in there |
| 20:19.52 | Erroneous | nope, flags.cxx has no cloaked shot |
| 20:19.55 | JeffM | k |
| 20:20.07 | javawizard2539 | I'll apply it |
| 20:20.08 | JeffM | javawizard2539, then yeah go and do it but make sure you follow the standards |
| 20:20.13 | javawizard2539 | Ok |
| 20:20.45 | *** part/#bzflag pacman87 (n=Timothy@resnet-45-192.dorm.utexas.edu) |
| 20:21.52 | JeffM | back to tool offset markers :/ |
| 20:21.59 | Erroneous | got the wire run? |
| 20:23.00 | javawizard2539 | Just to check, here's how my addition to the manpage turned out: |
| 20:23.01 | javawizard2539 | .B Invisible Bullet (IB) |
| 20:23.03 | javawizard2539 | Shots are invisible on radar (except your own). They are visible |
| 20:23.04 | javawizard2539 | out\-the\-window. Sort of stealth for shots. |
| 20:23.06 | javawizard2539 | .TP |
| 20:23.08 | javawizard2539 | .B Cloaked Shot (CS) |
| 20:23.10 | javawizard2539 | Shots are invisible out\-the\-window. They are still visible |
| 20:23.11 | javawizard2539 | on radar. Sort of cloaking for shots. |
| 20:23.13 | javawizard2539 | .TP |
| 20:23.15 | javawizard2539 | .B Tiny (T) |
| 20:23.17 | javawizard2539 | Tank becomes much smaller and harder to hit. |
| 20:23.17 | JeffM | pastebin large things please |
| 20:23.18 | javawizard2539 | .TP |
| 20:23.20 | JeffM | dude |
| 20:23.20 | javawizard2539 | Is that correct? |
| 20:23.32 | javawizard2539 | What do you mean? |
| 20:23.42 | JeffM | http://pastebin.bzflag.bz |
| 20:23.52 | JeffM | insatead of a huge blob in channel |
| 20:24.18 | JeffM | the last sentance dosn't sound like you are sure of yourself |
| 20:24.28 | JeffM | and you just basicly say the name of the shot a different way |
| 20:25.07 | JeffM | the description should be the same as the one for cloaking |
| 20:25.12 | JeffM | but with the word shot |
| 20:25.14 | JeffM | instead of tank |
| 20:25.29 | *** join/#bzflag FalseProphet (n=Gigi@unaffiliated/emonemo) |
| 20:25.48 | Erroneous | meh, it's the same pattern as IB |
| 20:25.51 | Erroneous | works for me :) |
| 20:26.00 | JeffM | ahh |
| 20:26.06 | JeffM | is that the origonal text for IB? |
| 20:26.08 | javawizard2539 | I copied IB and used that |
| 20:26.10 | Erroneous | yeah |
| 20:26.23 | JeffM | cus it ether is, or isn't not 'sort of' ;) |
| 20:26.43 | Erroneous | it's better ingame, "Your shots don't appear on other radars. Can still see them out window." |
| 20:27.01 | JeffM | any rason the man page should not be the same as in game? |
| 20:27.04 | javawizard2539 | I could copy the in-game description and use that instead |
| 20:27.08 | Erroneous | not that I can think of |
| 20:27.19 | JeffM | I think they should be the same |
| 20:27.20 | Erroneous | we have more room in the manpage, so we can use complete sentences |
| 20:27.37 | JeffM | cloaking is |
| 20:27.38 | JeffM | Tank becomes invisible out\-the\-window but is still visible on radar. |
| 20:27.43 | JeffM | so I can see it saying, |
| 20:27.53 | Erroneous | so yeah, you're welcome to fix IB's manpage description too |
| 20:27.57 | JeffM | Your shots become invisible out\-the\-window but are still visible on radar. |
| 20:28.47 | javawizard2539 | Ok |
| 20:28.55 | javawizard2539 | I should probably mention that the glow is still visible |
| 20:29.00 | JeffM | naw |
| 20:29.03 | javawizard2539 | Actually, should the glow still be visible? |
| 20:29.09 | JeffM | we don't mention that treads are visiable ;) |
| 20:29.15 | javawizard2539 | I initially thought yes to balance power |
| 20:29.17 | JeffM | yeah glow should be |
| 20:29.21 | Erroneous | I'd say yes |
| 20:29.22 | javawizard2539 | Ok |
| 20:29.27 | Erroneous | it's like cloaked tanks leaving treadmarks |
| 20:29.29 | JeffM | we doa similar thing with treadmarks and CL tanks |
| 20:29.30 | JeffM | yeah |
| 20:29.31 | Erroneous | subtle, but present |
| 20:29.38 | javawizard2539 | Ok |
| 20:29.58 | javawizard2539 | Also, I modified roger so that it ignores cloaked shots in the same way it ignores invisible bullets |
| 20:30.10 | JeffM | wonders how long it will be before someone abuses CS world weapon shots as world lights ;) |
| 20:30.25 | Erroneous | hah, I give it a month after release |
| 20:30.53 | JeffM | it'll be horibly slow |
| 20:31.04 | JeffM | and getting the shot to hold still would be a pain |
| 20:31.22 | javawizard2539 | How would you get a shot to hold still? |
| 20:31.28 | Erroneous | just set the velocity to zero in an update |
| 20:31.32 | JeffM | yeah |
| 20:31.36 | Erroneous | we rebuild shots on updates now right? |
| 20:31.40 | JeffM | no |
| 20:31.46 | JeffM | but hopefully before 3.0 |
| 20:31.50 | Erroneous | ah, ok |
| 20:31.51 | JeffM | that's what I was working on |
| 20:32.00 | Erroneous | sometimes gets confused between what's been checked in and what hasn't :) |
| 20:32.05 | L4m3r | real world lighting would be nice |
| 20:32.13 | JeffM | L4m3r, lots of thigns would be nice ;) |
| 20:32.24 | JeffM | our current lighting system is not fast for lots of lights |
| 20:32.25 | L4m3r | but on the other hand, I've found that disabling lighting makes it much easier to see on some maps |
| 20:32.29 | L4m3r | particularly in the evening |
| 20:32.30 | JeffM | since it basicly segments geometry |
| 20:33.50 | javawizard2539 | Are there any other documentation files I need to modify besides the bzflag man page? |
| 20:34.04 | Erroneous | not that I recall |
| 20:34.12 | JeffM | I mean you can do a wiki page for it if you want |
| 20:34.18 | JeffM | but that's just being nice :) |
| 20:34.39 | Erroneous | someone should go through our sample config in /misc and make sure LG and CS make it in there though |
| 20:34.52 | Erroneous | and the config builder |
| 20:37.17 | JeffM | indeed |
| 20:37.22 | JeffM | votes for spldart |
| 20:37.25 | JeffM | ;) |
| 20:39.05 | *** join/#bzflag bier (n=bier@p54A540A5.dip.t-dialin.net) |
| 20:40.23 | *** join/#bzflag noyb_ (n=noyb@nat/sun/x-344cf90706fbfcf5) |
| 20:42.10 | javawizard2539 | Is there a pastebin for images? |
| 20:42.22 | Epyon | imageshack xP |
| 20:42.38 | Winny | or tinypic |
| 20:42.40 | Winny | or something |
| 20:42.42 | Winny | bayimg |
| 20:46.39 | javawizard2539 | I uploaded a screenshot of a CS in front of the tank, it's at http://www.trivergia.com/bzfi0006.png |
| 20:46.40 | AAA_awright | http://my.bzflag.org/w/Radar_Marker I like |
| 20:47.26 | JeffM | hmm if AAA_awright likes it, then we can't do it ;) |
| 20:47.57 | AAA_awright | ;) |
| 20:48.09 | javawizard2539 | Any suggestions before I finish up and commit? |
| 20:48.30 | JeffM | javawizard2539, do it |
| 20:48.39 | javawizard2539 | Ok |
| 20:48.49 | JeffM | the joy of SVN is that we can fix and change at any time |
| 20:49.02 | JeffM | more little commits are better then one big one |
| 20:50.25 | AAA_awright | What distro does the list server use? |
| 20:50.28 | Epyon | javawizard2539, what is that? |
| 20:50.40 | javawizard2539 | What is what? |
| 20:50.49 | javawizard2539 | The screenshot is a cloaked shot just in front of the tank |
| 20:50.55 | Epyon | Ah |
| 20:51.06 | JeffM | AAA_awright, it is not running linux |
| 20:51.14 | JeffM | the current list server us running BSD |
| 20:51.25 | JeffM | it has run on debian in the past |
| 20:52.02 | AAA_awright | Hm |
| 20:52.10 | Winny | What, not running on windows server 2003? |
| 20:52.12 | Winny | ducks |
| 20:52.24 | JeffM | Winny, in a VM on the BSD system |
| 20:53.23 | AAA_awright | Which particular BSD? |
| 20:53.45 | Erroneous | free |
| 20:58.28 | TheRedBaron | hrm |
| 20:59.05 | TheRedBaron | in trunk - common doesn't build. fatal errors on \src\common\Vector4d.cxx Vector 3d.cxx & vector2d.cxx |
| 20:59.34 | TheRedBaron | using latest SVN checkout. |
| 20:59.43 | TheRedBaron | this is on VC9 |
| 20:59.47 | *** join/#bzflag Lukstr (n=Lukstr@206-248-157-113.dsl.teksavvy.com) |
| 21:00.18 | javawizard2539 | Unversioned files won't be automatically added, right? |
| 21:00.23 | Erroneous | correct |
| 21:00.27 | javawizard2539 | Ok |
| 21:00.29 | Erroneous | you'd need to "svn add" them first |
| 21:00.31 | blast007 | Winny: the list server is just PHP. I could run it off my Nokia if I wanted :P |
| 21:00.39 | javawizard2539 | No, I don't want them to commit |
| 21:00.45 | Winny | blast007: ok? |
| 21:00.47 | Erroneous | ah, ok :) |
| 21:01.19 | blast007 | Winny: re: the Windows Server comment |
| 21:01.25 | blast007 | it would probably run just fine on that too ;) |
| 21:01.29 | blast007 | even with IIS |
| 21:01.29 | Winny | k? |
| 21:02.45 | TheRedBaron | blast007: would your nokia ring everytime some one accessed the list server? ;) |
| 21:02.52 | Erroneous | heh |
| 21:03.34 | Paul_PL | Is there currently a centralized system of providing the information - who is playing where? |
| 21:03.47 | TheRedBaron | Paul_PL: not really |
| 21:04.04 | TheRedBaron | there are several user created sites |
| 21:04.21 | Epyon | What the heck is this: |
| 21:04.22 | Epyon | svn: Commit failed (details follow): |
| 21:04.22 | Epyon | svn: MKACTIVITY of '/svnroot/bzflag/!svn/act/2958663e-fb78-11dc-8770-25be8ce8ed1 |
| 21:04.22 | Epyon | 6': 403 Forbidden (https://bzflag.svn.sourceforge.net) |
| 21:05.54 | Epyon | JeffM, any ideas? |
| 21:06.05 | Lukstr | Epyon: I'm reading this: http://www.alagad.com/go/blog-entry/subversion-mkactivity-and-403-forbidden-headaches |
| 21:06.41 | Lukstr | oh yeah |
| 21:06.45 | Lukstr | svn urls are case sensitive |
| 21:06.55 | Lukstr | well, obviously |
| 21:07.10 | Paul_PL | I'm thinking about the easiest way of getting this information on game client demand |
| 21:07.13 | Epyon | Well it worked yesterday, I didn't do any changes since then O.o |
| 21:07.25 | Lukstr | Epyon you're using the exact same command? |
| 21:07.31 | Lukstr | what's the line? |
| 21:07.34 | Epyon | svn commit |
| 21:07.35 | Epyon | :P |
| 21:07.57 | Paul_PL | are there ideas better than asking all the game servers every time? |
| 21:07.57 | Lukstr | hm |
| 21:08.40 | Lukstr | Epyon: "You are trying to commit something to an anonymous checkout." ? |
| 21:08.50 | Lukstr | perhaps? |
| 21:09.07 | Epyon | It's the same checkout I used yesterday |
| 21:09.16 | Lukstr | have you ever done a check in with it though? |
| 21:09.30 | KTL | Paul_PL, i guess you saw http://my.bzflag.org/currentplayers.php |
| 21:09.59 | *** join/#bzflag tupone (n=atupone@pdpc/supporter/active/Tupone) |
| 21:09.59 | *** mode/#bzflag [+v tupone] by ChanServ |
| 21:10.01 | Lukstr | did you do an anonymouse checkout or did you authenticate |
| 21:10.02 | Lukstr | ? |
| 21:10.09 | Epyon | Auth |
| 21:10.15 | Lukstr | hmm |
| 21:10.54 | Lukstr | what's the url of the checkout? |
| 21:11.58 | Epyon | http://pastebin.bzflag.bz/d1fd043e4 |
| 21:13.36 | Lukstr | strange |
| 21:13.40 | Lukstr | so you have auth? |
| 21:13.57 | Epyon | Unless I was removed from the auth list :P |
| 21:15.12 | *** join/#bzflag QuantumBeep (i=QB@72.190.51.216) |
| 21:16.19 | Epyon | Lol, I was :P |
| 21:16.25 | Lukstr | hah |
| 21:16.34 | Lukstr | that'll do it |
| 21:16.36 | Epyon | brlcad, JeffM, is there something I should know about? |
| 21:19.36 | Epyon | definitively thinks that was an "uncool" way to go about things. |
| 21:31.15 | TheRedBaron | JeffM: the vector$d.cxx references in common.vcproj need to be deleted for common to build |
| 21:31.29 | Epyon | TheRedBaron, I did that but cant commit. |
| 21:31.35 | spldart | Huh? What? Did I get nominated for something? |
| 21:31.43 | TheRedBaron | spldart: court jester |
| 21:31.53 | spldart | uh huh |
| 21:31.54 | TheRedBaron | hands spldart the pointy hat with the bells |
| 21:32.15 | spldart | keeps it for Pogovina.. She'd love it |
| 21:39.59 | *** join/#bzflag LongDon (n=LongDon@host-091-097-081-209.ewe-ip-backbone.de) |
| 21:40.34 | Winny | o__o |
| 21:53.23 | Paul_PL | is the http://my.bzflag.org/currentplayers.php subdomain third-party or is it administrated by the BZFlag developers? |
| 21:53.47 | Lukstr | Paul_PL: as in my.bzflag.org ? |
| 21:54.48 | Winny | my.bzflag.org, images.bzflag.org and bzflag.org and .bz are official |
| 21:54.49 | Winny | IIRC |
| 21:55.00 | L4m3r | my.bzflag.org runs the list, forums, wiki, etc. it's administered by the project, yes |
| 21:55.22 | Winny | <something>.bzflag.org can be got by asking Tim for one |
| 21:55.34 | Paul_PL | thanks :) |
| 22:00.57 | JeffM | TheRedBaron, k |
| 22:02.30 | JeffM | Winny, not for much longer, only .net if we can get it |
| 22:02.39 | spldart | Ah.. Jeff Did you have a task for me when I get home n'stuff? |
| 22:02.45 | L4m3r | JeffM: eh? |
| 22:02.58 | JeffM | L4m3r, not giving out .org to non project people |
| 22:03.05 | L4m3r | ah |
| 22:03.09 | JeffM | give them .net |
| 22:03.13 | L4m3r | so l4m3r.bzflag.org is safe :) |
| 22:03.17 | JeffM | keep .org for the offical stuff |
| 22:03.18 | JeffM | nope |
| 22:03.21 | JeffM | you'd be .net ;) |
| 22:03.25 | L4m3r | aww :( |
| 22:04.53 | JeffM | you'll live ;) |
| 22:05.15 | brlcad | woot, done ... finally found and cleaned up the old network usage graphs |
| 22:05.24 | *** part/#bzflag CBG (n=CBG@about/essy/fan/CBG) |
| 22:05.30 | *** join/#bzflag CBG (n=CBG@about/essy/fan/CBG) |
| 22:05.30 | *** part/#bzflag CBG (n=CBG@about/essy/fan/CBG) |
| 22:06.39 | brlcad | http://bzflag.bz/usage/ |
| 22:07.28 | Epyon | brlcad, prv |
| 22:07.55 | *** join/#bzflag a_meteorite (n=a_meteor@unaffiliated/ameteorite/x-000000001) |
| 22:07.55 | Lukstr | brlcad: sexy graphs |
| 22:10.03 | CIA-16 | BZFlag: 03JeffM2501 * r16814 10/trunk/bzflag/MSVC/VC8/common.vcproj: don't build files that do not exist right now. |
| 22:11.43 | JeffM | make files should be fine as it looks like they were never added to them |
| 22:17.03 | *** join/#bzflag Erickj92 (n=erickj92@204.73.103.253) |
| 22:17.26 | Epyon | Goodbye, all. |
| 22:17.29 | *** part/#bzflag Epyon (n=Epyon@90-156-93-82.magma-net.pl) |
| 22:17.30 | spldart | time to go home :-D |
| 22:25.18 | AAA_awright | How about a decentralized global group list? |
| 22:25.27 | JeffM | AAA_awright, group manager |
| 22:25.51 | AAA_awright | I mean such as, you define groups as URLs |
| 22:26.06 | JeffM | http://my.bzflag.org/w/Group_Management_System |
| 22:26.17 | AAA_awright | I'm reading that right now |
| 22:26.31 | JeffM | we want it centralised, but easy to use |
| 22:27.47 | AAA_awright | What if we could define a group such as, http://bzfx.net/group.bzfx.admin.bzg though? |
| 22:27.51 | JeffM | why? |
| 22:28.03 | AAA_awright | You could refer to it as bzfx.admin for short |
| 22:28.05 | JeffM | what advantage does that have over an easy to use system |
| 22:28.12 | AAA_awright | It is easy to use |
| 22:28.24 | JeffM | not everyone has hosting |
| 22:28.39 | AAA_awright | Use the central system then. |
| 22:28.48 | JeffM | then why have more then one system? |
| 22:28.56 | AAA_awright | This is all one system |
| 22:29.14 | JeffM | then why do we need URLS on another system? |
| 22:29.22 | *** part/#bzflag LongDon (n=LongDon@host-091-097-081-209.ewe-ip-backbone.de) |
| 22:29.45 | AAA_awright | The "central" groups would be at, say, http://groups.bzflag.org/what.who.bzg |
| 22:30.06 | JeffM | oh you want to mix and match |
| 22:30.11 | JeffM | groups are tied to auth |
| 22:30.11 | AAA_awright | kinda? |
| 22:30.31 | JeffM | the complexity dosn't realy add any global benifit |
| 22:30.39 | JeffM | why can't bzfx just use our system? |
| 22:30.40 | AAA_awright | The server is only aware of the groups that have permissions attached to them |
| 22:30.52 | AAA_awright | So it doesn't matter at all |
| 22:30.56 | JeffM | the gorup manager dosnt' store permmisions |
| 22:31.26 | AAA_awright | No, but the only way a server knows about a group is when it is told "This group <a> has permissions x,y,z" |
| 22:31.31 | JeffM | yes |
| 22:31.32 | JeffM | so? |
| 22:32.00 | JeffM | the problem is, how does your group auth a user? |
| 22:32.02 | AAA_awright | It might be useful to grab groups from remote locations |
| 22:32.07 | JeffM | why? |
| 22:32.42 | AAA_awright | Why not? It makes it decentralized, open, fills the purpose of global groups, and you are not tied to the central server (as much) |
| 22:33.13 | AAA_awright | ...It sounded like a good idea |
| 22:35.22 | JeffM | well the auth of who is in what group happens at the list server |
| 22:35.26 | JeffM | cus you need the token for it |
| 22:35.44 | JeffM | your system would not be able to auth a user |
| 22:35.55 | JeffM | since it would not be tied to the list server, and could not reuse the one token |
| 22:36.20 | AAA_awright | How do local groups work? Great, now instead of getting them from the local fs, grab them from the network |
| 22:36.48 | JeffM | local groups use local users |
| 22:36.54 | JeffM | they suck |
| 22:37.21 | Erroneous | has plenty of local groups that have only global users |
| 22:37.31 | JeffM | does it use BZID? |
| 22:37.37 | JeffM | cus that just came into my head |
| 22:37.52 | JeffM | you could have remote groups that just had BZID |
| 22:38.02 | *** join/#bzflag Ride- (n=Ride-@24-197-250-9.dhcp.stpt.wi.charter.com) |
| 22:38.08 | JeffM | and then you could check membership based on ID |
| 22:38.12 | mibocote | what role doe groups play in bzflag? |
| 22:38.24 | JeffM | mibocote, many servers use them for permision groups |
| 22:38.36 | JeffM | such as all the users in a league can spawn, or start games |
| 22:38.52 | mibocote | thanks |
| 22:38.54 | JeffM | or a group of users have kick and ban |
| 22:39.23 | JeffM | AAA_awright, you could do your idea as a plugin |
| 22:39.28 | JeffM | it'd not be too hard |
| 22:39.36 | AAA_awright | Interesting idea JeffM |
| 22:39.37 | JeffM | now that I think about it |
| 22:39.42 | JeffM | so it could work with the global system |
| 22:39.49 | JeffM | dosn't need to be part of it, but it could work with it |
| 22:40.11 | AAA_awright | Great way to prototype |
| 22:40.23 | JeffM | there is no reason to built it into BZ |
| 22:40.25 | AAA_awright | or.. |
| 22:40.30 | JeffM | as none of our standard services would use it |
| 22:40.34 | JeffM | but it could be a plugin we ship |
| 22:40.36 | AAA_awright | True |
| 22:40.40 | JeffM | that'd be fine |
| 22:40.48 | JeffM | for those that are untrusting |
| 22:41.01 | JeffM | I'd also like to have some form of automation API for the group manager |
| 22:41.13 | AAA_awright | I thought Constitution was working on that |
| 22:41.17 | AAA_awright | It uses MySQL though |
| 22:41.17 | JeffM | so that leauges could automate geting people into and out of groups |
| 22:41.34 | JeffM | he is working on the group manager that is speced in that wiki article |
| 22:41.39 | AAA_awright | I have plans to work on something similar to that, after I finish everything else :p |
| 22:41.40 | JeffM | I don't know if he did an automation API |
| 22:41.59 | JeffM | it should have an API page that just takes requests and does operations |
| 22:42.05 | JeffM | after auth of course |
| 22:43.37 | *** join/#bzflag eagle (n=45f4ec24@bz.bzflag.bz) |
| 22:44.31 | JeffM | cus the only reasons I could see to not want to use the global system is lack of automation, or lack of trust |
| 22:44.46 | *** join/#bzflag bz60 (n=62e29fa2@bz.bzflag.bz) |
| 22:45.23 | bz60 | -public "Jack" |
| 22:45.34 | JeffM | this isn't a server bz60 |
| 22:46.12 | *** join/#bzflag bz11 (n=62e29fa2@bz.bzflag.bz) |
| 22:46.37 | bz11 | how do you get your server running |
| 22:46.53 | JeffM | depends on what OS you run |
| 22:47.02 | bz11 | vista |
| 22:47.11 | Ride- | go to ##BZW that place is for maps and map making |
| 22:47.15 | JeffM | then you use the command line terminal |
| 22:47.17 | Ride- | or just ask jeff M |
| 22:47.28 | JeffM | looks for more asprin |
| 22:47.35 | Ride- | lol |
| 22:47.43 | bz11 | gol |
| 22:47.56 | Ride- | gives Jeff his last asprin pill |
| 22:47.59 | Ride- | for free |
| 22:48.02 | JeffM | I wasn't joking |
| 22:48.04 | *** mode/#bzflag [+v Erroneous] by ChanServ |
| 22:48.12 | bz11 | what command line |
| 22:48.14 | JeffM | bz11, there are a number of threads on the fourms about how to start a server |
| 22:48.19 | JeffM | read them first |
| 22:48.24 | bz11 | ok |
| 22:48.25 | *** mode/#bzflag [+o Erroneous] by ChanServ |
| 22:48.28 | JeffM | they tell you how to start the command line terminal |
| 22:49.02 | AAA_awright | http://my.bzflag.org/w/Creating_A_Server |
| 22:49.06 | AAA_awright | Eh, too late |
| 22:51.35 | *** join/#bzflag TheRedBaron (n=kemhof@75.144.224.155) |
| 22:54.11 | *** join/#bzflag hwilde_ (n=hwilde@pool-151-201-56-160.pitt.east.verizon.net) |
| 22:54.17 | hwilde_ | cheater on Incoming |
| 22:54.20 | *** join/#bzflag whodaman (n=whodaman@pdpc/supporter/student/whodaman-) |
| 22:54.27 | hwilde_ | anybody got ops ? |
| 22:54.34 | hwilde_ | giancarlo borromeo wings cheat |
| 22:57.26 | hwilde_ | anybody home? |
| 22:57.32 | JeffM | yes |
| 22:57.39 | JeffM | it is best to tell the server owner |
| 22:57.52 | hwilde_ | who owns it |
| 22:57.59 | JeffM | dunno |
| 22:58.06 | hwilde_ | who owns bztank.net |
| 22:58.08 | Winny | benfish or something |
| 22:58.27 | hwilde_ | it's actually a pretty good cheat |
| 22:58.31 | hwilde_ | it is faster than any i've seen |
| 22:58.53 | hwilde_ | and he can have gm too |
| 22:59.33 | *** join/#bzflag Think_Differentl (n=shedelj@unaffiliated/thinkdifferentl/x-838543) |
| 22:59.45 | hwilde_ | he left anyways too late |
| 23:03.45 | KingofCamelot | Hey JeffM |
| 23:03.46 | TheRedBaron | hrm - after i build bzflag on VC9, do i just execute the .exe file ? |
| 23:03.50 | JeffM | yes? |
| 23:03.54 | TheRedBaron | using trunk |
| 23:03.55 | JeffM | TheRedBaron, sure |
| 23:04.00 | JeffM | if it works ;) |
| 23:04.02 | TheRedBaron | JeffM: ok - i thought so |
| 23:04.09 | TheRedBaron | which brings me to the next point - it doesn't :) |
| 23:04.11 | JeffM | VC9 in branch has "issues" |
| 23:04.14 | *** join/#bzflag Xenathi (n=5a9c5d52@bz.bzflag.bz) |
| 23:04.21 | TheRedBaron | I used vc8 proj file |
| 23:04.27 | JeffM | I mean the compiler |
| 23:04.36 | JeffM | are you in branch or trunk? |
| 23:04.42 | CIA-16 | BZFlag: 03javawizard2539 * r16815 10/trunk/bzflag/ (11 files in 5 dirs): Added a new flag, Cloaked Shot, which makes your shots invisible out-the-window but still visible on radar. |
| 23:04.50 | JeffM | ibot javawizard2539++ |
| 23:04.53 | TheRedBaron | error: "The Procedure Entry Point WSAPoll could not be located in the dynamic LInk Library WS@_32.dll" |
| 23:05.00 | TheRedBaron | JeffM: trunk |
| 23:05.16 | TheRedBaron | er "WS2_32.dll" |
| 23:05.17 | JeffM | TheRedBaron, are you on the same machine as you developed on? |
| 23:05.20 | TheRedBaron | yup |
| 23:05.33 | TheRedBaron | XP x64 |
| 23:05.36 | JeffM | oh |
| 23:05.40 | JeffM | well still |
| 23:05.43 | TheRedBaron | but i specified the 32bit |
| 23:05.47 | JeffM | well we have heard of issues in VC9 |
| 23:05.52 | JeffM | issues just like that |
| 23:05.57 | TheRedBaron | ok |
| 23:06.00 | JeffM | and we haven't taken the time to work on them |
| 23:06.10 | JeffM | I HIGHLY recomend VC 2005 |
| 23:06.12 | JeffM | VC8 |
| 23:06.14 | TheRedBaron | ok |
| 23:06.28 | TheRedBaron | If you need any help debugging VC9 |
| 23:06.29 | JeffM | I'm sure before we release we will do what we can to make it build on VC9 |
| 23:06.49 | JeffM | TheRedBaron, we need help understanding why it dosn't like to load the correct runtimes ;) |
| 23:07.07 | JeffM | so it's more then debuging, it's understanding what VC9 changed from VC8 |
| 23:07.34 | KingofCamelot | Oh, btw JeffM, got BZFlag up and running on VC 2005 without too many problems :) |
| 23:07.37 | brlcad | Lukstr: yeah, hotness .. |
| 23:07.49 | brlcad | shame those mods didn't make it into the sources |
| 23:07.52 | JeffM | KingofCamelot, yeah 2005 is prety straigit forward |
| 23:08.07 | JeffM | thats why I recomend it' |
| 23:08.11 | TheRedBaron | JeffM: okies :) |
| 23:08.21 | KingofCamelot | Only thing which wasn't mentioned in the wiki or anything was the need to get Glew, but that wasn't too hard to figure out |
| 23:08.25 | JeffM | 2008 is very new |
| 23:08.31 | JeffM | KingofCamelot, you do not need glew |
| 23:08.41 | JeffM | you probalby built one of the distrobution targets |
| 23:08.43 | KingofCamelot | Hmm, couldn't seem to get mine to compile without it |
| 23:08.44 | JeffM | not the normal ones |
| 23:08.50 | JeffM | there is debug and debug_distrobution |
| 23:08.59 | JeffM | distrobution needs glew and curses |
| 23:09.05 | JeffM | the regular one does not |
| 23:09.09 | JeffM | glew is optional |
| 23:09.19 | KingofCamelot | ah, thats it :) |
| 23:09.25 | KingofCamelot | you're right, says debug_distrobution |
| 23:09.43 | JeffM | there is a section in the readme.windows that talks about that |
| 23:09.52 | KingofCamelot | pft...readme :P |
| 23:09.53 | JeffM | under **Targets** |
| 23:09.59 | brlcad | DTRemenak: do you remember what the perplayer graphs showed? ... |
| 23:10.07 | brlcad | ( http://bzflag.bz/usage/ ) |
| 23:10.16 | brlcad | scratches his head |
| 23:11.01 | brlcad | ooh, maybe it's the first derivative.. |
| 23:11.01 | KingofCamelot | So I'm not sure if you know much of anything about the server listing, but I've got an odd bug going on there after some code changes I made |
| 23:11.10 | KingofCamelot | I've made it so it will sort the servers by ping, but for some reason it only does the ping for about 4/6 of them one the first attempt |
| 23:11.30 | KingofCamelot | If I tell it to do it again it usually adds the 5th one, then one more time and the list acts like it should |
| 23:11.37 | JeffM | KingofCamelot, you ping all the servers? |
| 23:11.39 | JeffM | eww |
| 23:11.39 | KingofCamelot | sometimes it only does 3/6 on the first attempt |
| 23:11.49 | KingofCamelot | How else can you sort them by ping? :) |
| 23:11.49 | JeffM | many don't respond to pings |
| 23:12.02 | JeffM | KingofCamelot, I woudln't sort them by ping |
| 23:12.10 | JeffM | thats a lot of servers |
| 23:12.22 | JeffM | if I did a connection quality, I'd do it on the server |
| 23:12.28 | JeffM | the list server |
| 23:12.32 | JeffM | and send it allong |
| 23:12.47 | JeffM | but yea a lot of servers won't respond to them |
| 23:12.53 | KingofCamelot | Ah, interesting |
| 23:12.57 | JeffM | and that's going to be a boatload of data |
| 23:13.08 | JeffM | you've made a flood machine :) |
| 23:13.26 | KingofCamelot | How many servers are there normally on the regular client? I only see the 6 or so test servers |
| 23:13.36 | JeffM | 250+ |
| 23:13.46 | JeffM | ibot bzfslist |
| 23:13.47 | KingofCamelot | Heh, you're right, that is a lot to ping :P |
| 23:13.57 | JeffM | bahh I can never remember the command |
| 23:14.07 | JeffM | ibot bzflist |
| 23:14.09 | AAA_awright | ~bzfsquery |
| 23:14.17 | JeffM | thanjks |
| 23:14.26 | Winny | yay I'm winning |
| 23:14.29 | JeffM | so 261 servers |
| 23:14.54 | AAA_awright | dangit. Attention everyone, I am losing! |
| 23:15.27 | KingofCamelot | Well, the server enhancement listing in the gsoc part of the wiki suggests sorting by ping...but ya, I see the problems :P |
| 23:15.51 | JeffM | not all sugestions are good ones ;) |
| 23:15.57 | JeffM | but that's just my opionon |
| 23:15.59 | KingofCamelot | Hehe, indeed |
| 23:16.02 | JeffM | it may be viable to do a slow ping |
| 23:16.07 | JeffM | or just ping the one you have selected |
| 23:16.08 | JeffM | I dunno |
| 23:16.14 | JeffM | that'd all be the stuff to research |
| 23:16.17 | AAA_awright | KingofCamelot: I operate a server list scanner, I made it intelligent though, it very accurately predicts where servers are, and only needs to get new info from about 5 servers a minute |
| 23:16.51 | KingofCamelot | AAA_awright, sounds interesting :) |
| 23:16.54 | brlcad | KingofCamelot: you're proposing list server updates, right? |
| 23:17.11 | KingofCamelot | Enhancing the server listing, ya |
| 23:17.35 | JeffM | brlcad, that's not the same as list server updates ;) |
| 23:17.39 | JeffM | the order matters |
| 23:17.48 | KingofCamelot | Heh, confusing terms :P |
| 23:17.53 | KingofCamelot | scratches head slightly |
| 23:18.17 | brlcad | er, yeah, what you said :) |
| 23:18.38 | brlcad | just dropped one word.. what a difference in meaning ;) |
| 23:18.57 | brlcad | actually transposed |
| 23:19.17 | brlcad | s/list server/server list/ |
| 23:20.34 | KingofCamelot | Maybe we should call it the server browser to avoid confusion? :P |
| 23:20.49 | brlcad | nah |
| 23:20.59 | Winny | the server listerizer! |
| 23:21.33 | brlcad | i would have slapped myself had I actually read what I wrote :) |
| 23:21.41 | javawizard2539 | If it's an enhancement to locating the nearest servers, you could try something like GeoLite City |
| 23:22.44 | javawizard2539 | It's what sourceforge uses to select the nearst mirror |
| 23:22.58 | *** join/#bzflag O-Neil (n=Miranda@dslb-088-070-246-167.pools.arcor-ip.net) |
| 23:23.07 | brlcad | KingofCamelot: it would be really cool to have an "about" or "get info" page for servers in the list that could show the server's map (in 3d of course), player lists, settings, a more informative description, etc |
| 23:23.30 | javawizard2539 | I agree |
| 23:23.36 | KingofCamelot | As do I :) |
| 23:24.13 | brlcad | could actually join as an invisible observer and get reduced updates (like 1/min) -- so you could show a "snapshot" of what it actually looks like |
| 23:24.41 | AAA_awright | invisible observer? |
| 23:25.04 | javawizard2539 | Yes, a change could be made to the server so that when the listserver user joins, it doesn't show up |
| 23:25.09 | javawizard2539 | That's the idea, right? |
| 23:25.15 | JeffM | brlcad, as oposed to those vissable observers? ;) |
| 23:25.38 | KingofCamelot | Just cause you've never seen them doesn't mean they aren't visible :P |
| 23:26.26 | JeffM | there are lots of thigns we can do |
| 23:26.34 | *** join/#bzflag BenUrban (n=benurban@unaffiliated/benurban) |
| 23:26.58 | javawizard2539 | Random question, is there any particular reason that users can't use the mouse to navigate through bzflag menus? |
| 23:27.13 | brlcad | yeah, I mean having another class of observers that don't show up in black even |
| 23:27.24 | brlcad | except maybe to admins |
| 23:27.33 | KingofCamelot | Anyways JeffM, would my little sort servers by ping change be sufficient as a patch for a gsoc application? Obviously not practical for real use, but just to show that I'm competent and have started playing around with the current capabilites? |
| 23:27.36 | JeffM | or just make it so you can get the world before you send a join ;) |
| 23:27.46 | JeffM | KingofCamelot, probably |
| 23:27.50 | brlcad | that could work mebbie |
| 23:27.58 | brlcad | but I was thinking they'd stay connected |
| 23:28.00 | JeffM | you may want to try something smaller |
| 23:28.03 | brlcad | get updates |
| 23:28.07 | JeffM | it dosnt' have to be a big thing |
| 23:28.16 | brlcad | then admins could actually see people on their "about" page |
| 23:28.21 | KingofCamelot | True, but I've already got this code done :P |
| 23:28.24 | brlcad | (just not interact with them) |
| 23:28.30 | JeffM | brlcad, may as well just connect as a real observer then |
| 23:28.43 | KingofCamelot | Need to start work on the actual application, heh |
| 23:28.44 | JeffM | then if you "join" you go and just change teams to a regular player |
| 23:28.59 | JeffM | and the internal interface pops to full |
| 23:29.16 | JeffM | KingofCamelot, it shold be fine |
| 23:29.24 | jorgenpt | brlcad: You're fully stocked on primary mentors, btw? |
| 23:29.27 | JeffM | it's mostly just to make sure you conform to our standards |
| 23:30.00 | brlcad | JeffM: yeah, but don't want to have server-list observers dragging down all the network updates .. nasty load |
| 23:30.05 | brlcad | and there are only so many observer slots |
| 23:30.17 | KingofCamelot | JeffM, awesome. :) |
| 23:30.20 | brlcad | I was thinking server-list observers would be like the prejoin info clients, unlimited |
| 23:30.29 | JeffM | maybe |
| 23:30.48 | brlcad | so they'd only get position/player updates like once a minute, but stay connected |
| 23:30.50 | JeffM | have to see how it was presented |
| 23:30.51 | KingofCamelot | Anyways...I've got stuff to do, but I'll be back. I'm so tired...in all the excitement of compiling bzflag last night I forgot to go to bed. :P |
| 23:30.54 | brlcad | till they leave the menu |
| 23:30.59 | *** join/#bzflag spldart (n=vircuser@c-98-197-19-35.hsd1.tx.comcast.net) |
| 23:31.55 | jorgenpt | Yes, I'm following Arthur's lead (wasn't that his name?). Good nigt! |
| 23:31.57 | jorgenpt | night too. |
| 23:32.25 | KingofCamelot | King Arthur was the king of Camelot, yes ;) |
| 23:47.29 | *** join/#bzflag Dys (n=ca044962@gateway/web/cgi-irc/zeebrothers.net/x-eaff096ed54fb70a) |
| 23:48.01 | *** join/#bzflag O-Neil (n=Miranda@dslb-088-070-246-167.pools.arcor-ip.net) |
| 23:49.29 | *** join/#bzflag TD-Linux (n=wheeeeee@about/essy/indecisive/TD-Linux) |
