IRC log for #bzflag on 20080524

00:00.04JeffMbut there is a curve you are trying to approxmate
00:00.23JeffMbasicly make the normal be the vector from the center axis to the vertex
00:00.27JeffMthen normalise it
00:00.34JeffMthat will give you smoother shading
00:00.56JeffMsince it'll then have the normals from a real curve in and shade it like it was curved, not flat segements.
00:15.13KTLhas the impression the normals really have to be normalized
00:17.03*** join/#bzflag ahs3- (n=alvin@adsl-065-005-193-158.sip.rdu.bellsouth.net)
00:41.37*** part/#bzflag Mets (n=Mets@cpe-72-224-55-41.nycap.res.rr.com)
00:42.03JeffMKTL, yep
00:42.20KTLgot normals working, doing ceiling normals now
00:42.53*** join/#bzflag Mets (n=Mets@cpe-72-224-55-41.nycap.res.rr.com)
00:44.01KTLhttp://users.edpnet.be/dpf/bz/bzfi0074.png
00:45.24spldartcorkscrew?
00:45.45KTLmyes, climbable, i just made the normals acceptable -> smoother surface
00:47.01brlcadnifty
00:49.10spldartultra nifty
00:49.14spldartonline?
00:49.45KTLi dont have a server, and its not playable yet, but it is fun
00:50.10KTLhttp://users.edpnet.be/dpf/bz/oe.bzw
00:50.35KTLgood that the server options can be put in the bzw
00:51.01spldartI'll have to try that out later
00:51.10spldartis testing code right now
00:59.39BulletCatcherIt looks like MacOS 10.4 does not have wcsdup() as used on line 205 of src/other/ftgl/src/FTFont/FTBufferFont.cpp .
01:00.55BulletCatcherThe Linux man page says wcsdup() "is a GNU extension."
01:01.32KTLhttp://users.edpnet.be/dpf/bz/bzfi0075.png  <- also climbable
01:13.35BulletCatcherWe could add wcsdup to the AC_CHECK_FUNCS list in the top-level configure.ac and then import the GNU source code for it for when it is not available.
01:29.30*** join/#bzflag Epyon_ (n=Epyon@90-156-93-82.magma-net.pl)
01:33.17*** join/#bzflag ndim_ (i=hun@moooo.n-dimensional.de)
01:45.02*** join/#bzflag JeffM (n=Jeff@pool-71-109-223-10.lsanca.dsl-w.verizon.net)
01:45.26KTLthose meshes
01:45.30KTLis it possible to define one
01:45.44KTLand to reuse it with a transformation over and over?
01:47.48blast007KTL: yeah. with 'group' and 'define', iirc
01:49.57spldartHrm.. Somebody did a nice job cleaning up BUGS but got a little prune happy imho
01:50.13JeffMhowso?
01:51.18spldartOne of my ongoin bugs was removed.. along with another that could have been related.. but I was still seeing the bug last I checked... I guess I'll hafta recheck.. The other bug left that I had submitted I just confirmed has been fixed.
01:52.07spldartgoes back to testing
01:52.24JeffMI pruned a lot of nodupes assuming that ether they were fixed, or would be reported again with more info if they happend with a current person
01:53.30spldartgimme a bit.. maybe this one is fixed.. wanna do a latest windowz and linuz build and check
01:57.02CIA-33BZFlag: 03dtremenak * r17246 10/trunk/bzflag/src/common/Bundle.cxx: prune dead code
01:57.09spldartnvmd.. I can't join a server with my latest windows build. I get token and stop.. But can hear my latest linux client playing on my latest linux server
01:57.38spldartand f12 quit isn't working on windows.. but does on linux
02:01.39*** join/#bzflag forrestv (n=forrestv@unaffiliated/forrestv)
02:02.53JeffMthen get those reported
02:03.03spldart... But I do like the more clearly formatted BUGS layout
02:04.37spldartthe "coding: utf-8" @ the top... is that a demand on a precise formating of the commits to that file?
02:05.50spldartOh crap! My, the one bug, is still there... sry.. but can't confirm or deny it now...
02:06.40*** join/#bzflag BenUrban (n=benurban@unaffiliated/benurban)
02:11.23spldartok.. bout to add those freshly observed bugs
02:11.29KTLwhat is the easiest way to prevent spawning on the ground?
02:12.24KTLwaterlevel
02:12.25CIA-33BZFlag: 03bzspldart * r17247 10/trunk/bzflag/BUGS: Add Win client f12 bug and win client not full join bug.
02:12.57spldartLinux build is doin good right now though :)
02:13.27spldartis anxious for his hardware for his linux test rig to arrive from newegg
02:15.21BulletCatcherA revised freetype/ftgl configure patch is now ready for review at http://pastebin.bzflag.bz/m10ade33f .  It should satisfy the concerns raised by brlcad.
02:30.26forrestvwhere is the source for the listserver?
02:31.56bradsvn
02:33.14bradhttp://bzflag.svn.sourceforge.net/viewvc/bzflag/trunk/db/
02:45.47spldarthrm.. Noticing this... Some peeps with 2.0.x clients are reporting sealing to my bases and some are fine. I can't reproduce but to many odds and ends peeps are having problems to be a non issue on my end.
02:46.23spldartNot that I expect and investigation and new release of 2.0.x..
02:46.25spldartjust noting
02:46.35blast007spldart: are they on a specific OS?
02:46.56spldartI believe the ones I have gotten that much info from are windows
02:47.22spldartWhat's the scuttlebutt on a possible 3. release?
03:03.58short_circuitgoes to bed with sci-fi disaster marathon
03:20.04*** join/#bzflag Penth (n=1000@71.249.148.181)
03:21.02*** join/#bzflag Ride- (n=Ride-@24-197-250-9.dhcp.stpt.wi.charter.com)
03:21.57Ride-hey guys i have a ? about momentum of a tank when jumping backwards and forwards
03:22.57JeffMwhat about it?
03:25.39Ride-well on the server when you jump forward then hold back and jump backward right after forward jump you still go forward but slower
03:26.19JeffMasking the same question more then once will not get you a different answer
03:26.36JeffMthe jump adds a velocity
03:26.42Ride-but is there a variable that can fix that
03:26.46Ride-like friction maybe
03:26.53JeffMthat speed is modified by your current speed
03:26.59Penththat's physics. tanks still have mass
03:27.03JeffMyou go slower when you move backwards
03:27.13JeffMand yes there is acceleration
03:27.25JeffMthere is no variable to 'fix that' other then normal acceleration
03:27.39JeffMbut it's all based on your current speed
03:31.06blast007don't jump until you are moving backwards then
03:36.37Ride-ok
03:36.52Ride-it was only sabar who moved forward on backward jump
03:37.01Ride-we think it might bee his mouth
03:37.15blast007yeah, he talks too much while driving his tank
03:37.17Ride-*mouse
03:37.22Ride-:D
03:37.24blast007;)
03:37.59blast007does the server have acceleration limitations turned on?
04:49.56CIA-33BZFlag: 03dtremenak * r17248 10/trunk/bzflag/ (24 files in 6 dirs): (log message trimmed)
04:49.56CIA-33BZFlag: A first stab at international input.
04:49.56CIA-33BZFlag: * Allow wide-character key events.
04:49.56CIA-33BZFlag: * Request wide-character input on Windows platforms.
04:49.56CIA-33BZFlag: * Strip down a variant of FTUnicodeStringItr for UTF8-only usage and add an additional utility class for converting single wide characters into UTF8 strings.
04:50.00CIA-33BZFlag: * Convert wide-character keystrokes into UTF8 strings where appropriate.
04:50.02CIA-33BZFlag: Issues:
04:50.41DTRemenakloves that CIA chopped off the "Issues" section :)
04:50.59think_tankmaybe that can be a project for the next GSoC
04:51.10DTRemenaknaw, it'll be done by then
04:51.13think_tank;)
04:51.15DTRemenak(famous last words)
04:51.27think_tank"never say never"
04:51.56DTRemenakyou can at least type in local characters, send them to the server, and get them back now :)
04:52.21DTRemenakeven if you do have to press "right" between 1 and 3 times for each character ;)
04:52.59DTRemenakand can't bind local keys (forgot to mention that in issues)
04:56.35CIA-33BZFlag: 03dtremenak * r17249 10/trunk/bzflag/src/platform/SDLDisplay.cxx: I *think* this will put SDL at the same place Windows is at wrt unicode input. Someone will need to try.
05:01.45BulletCatcherSpeaking of SDL, is it really optional?  The configure system thinks so but there is a compile error on MacOS without it.
05:01.58DTRemenakon the mac it is required
05:02.13DTRemenakon other unices, linux, and windows it is optional
05:02.27BulletCatcherOkay.  I'll add that to my list.
05:02.46DTRemenakthe old native mac stuff was very flaky, so it got pulled
05:04.51*** join/#bzflag delusional_ (n=delusion@pool-141-158-72-148.pitt.east.verizon.net)
05:09.18CIA-33BZFlag: 03dtremenak * r17250 10/trunk/bzflag/src/common/WordFilter.cxx: in case other windows folks want to try localized input, disable the wordfilter on windows when building debug. msvcrt throws a ton of asserts that make it difficult to get through.
05:21.38brlcadBulletCatcher: that looks great
05:24.28brlcadoh noes
05:24.39brlcadze beloved wordfilter ist disabled
05:25.01DTRemenakonly when debugging on windows
05:25.09brlcadyeah, I saw
05:25.14DTRemenakdoesn't particularly want to tackle utf8-ifying regex at the moment ;)
05:25.33brlcadis it that the char * is multibyte coming in?
05:25.44brlcador does there also need to be a wchar
05:25.51DTRemenakchar*
05:25.55DTRemenakbut some of them are negative
05:26.00DTRemenakand msvcrt doesn't like that
05:26.42brlcadis it an invalid parameter assert?
05:26.56DTRemenakyeah
05:27.11brlcadthere's a function that will tell the crt to stfu
05:27.15brlcad_set_invalid_parameter_handler()
05:27.22brlcadset that to an empty func and it'll behave
05:28.10brlcadlearned that trick just recently .. crt asserts on the traditional I/O routines were added to msvc recently ..
05:30.34*** join/#bzflag ibot (i=ibot@pdpc/supporter/active/TimRiker/bot/apt)
05:30.34*** topic/#bzflag is http://cia.vc/stats/project/BZFlag || http://my.BZFlag.org/w/Getting_Help || logs at http://ibot.rikers.org/%23bzflag/ || http://ohloh.net/projects/189 || 2.0.10 is the latest public release || GSoC 2008, mentor blog at http://tanks4code.blogspot.com || Congratulations GSoC participants: http://code.google.com/soc/2008/bzflag/about.html || Students should add their projects to the wiki for review
05:30.34*** mode/#bzflag [+o ibot] by ChanServ
05:30.40brlcadfound it required where vc2005 was annoyingly throwing the asserts on code that relied on the posix-defined behavior to just set error codes when things like close() failing
05:32.15BulletCatcherbrlcad: Thanks.  I'll commit the change shortly.
05:36.44Flashsince the Windows gurus are out tonight, maybe I can get an answer to the question burning in my mind
05:36.55Flashwhat is the uninitialized value of SOCKET ?
05:37.50DTRemenakINVALID_SOCKET
05:37.55CIA-33BZFlag: 03bullet_catcher * r17251 10/trunk/bzflag/ (4 files in 3 dirs):
05:37.55CIA-33BZFlag: Fully support the ./configure options --enable-freetype-build and --enable-ftgl-build (or their absence).
05:37.55CIA-33BZFlag: Require the same minimum version of freetype in the top-level configure as in FTGL's configure.
05:39.07BulletCatcherFlash: I expect r17251 to be the complete fix for your freetype configure problem.
05:39.20Flashyay
05:39.33DTRemenakwhich is ~0 iirc
05:39.44Flashas in, auto-enable freetype-build for MacOS 10.5?
05:39.46DTRemenakor rather (SOCKET)(~0)
05:39.54DTRemenakbut it's best to use the define
05:40.10BulletCatcherOh.  Almost complete then.  :-(
05:40.25FlashDTRemenak: ok, thanks. One more thing to add to the Unix emulation of Windows support for BSD sockets
05:40.56BulletCatcherYou'll have to explicitly use --enable-freetype-build for now, but it should work.
05:40.57DTRemenakyeah.  on unices a quick #define INVALID_SOCKET -1 will do the trick.
05:48.24BulletCatcherFlash, when you have a few minutes I'd like to discuss your need for using the built-in freetype so I can use the most appropriate trigger.
05:48.32Flashok
05:48.34Flashshoot
05:48.40Flasher.... wait
05:48.46Flashtalk
05:48.55BulletCatcherI recall it has something to do with /usr/X11.
05:49.30FlashIn /usr/X11/lib there is a libgl that has an issue with the Framework OpenGL library
05:49.48Flashso we don't want to include /usr/X11/lib in the link line
05:50.00Flashbut the system freetype-config slams it in there no matter what
05:50.19BulletCatcherGot it.
05:50.35BulletCatcherSo first, I need to detect the buggy libgl.
05:51.01FlashI suppose. I'd just detect OS 10.5 and force --enable-freetype-build
05:51.43BulletCatcherThat would probably be easiest.
05:52.47BulletCatcherIs that good enough for the more experienced developers in the audience?
05:53.24Flashbrlcad is the one who explained it to me... he should have a good vote
05:54.05brlcadavoids platform checks at all costs
05:54.55brlcadand it's not a voting matter, just a matter of someone actually doing something about it :P
05:55.50BulletCatcherSo you think we should find a way to directly detect the buggy libgl?
05:56.25brlcadfreetype-config shouldn't be in the path on the Mac by default
05:56.52brlcadso it should fall back to using bundled freetype regardless
05:57.12Flashit's in /usr/X11/bin... "should" that be in the path? I can't say
05:57.16brlcadif it doesn't, there's probably some *other* logic that needs tweaking in configure.ac
05:57.40brlcadyeah, it's in /usr/X11/bin, but that's not a default system path unless your in an xterm
05:57.47BulletCatcherIt will be in the path if configure is run from xterm.
05:57.54brlcador /usr/X11R6/bin on 10.4-
05:58.51FlashX11R6 is a link to X11 on 10.5
05:58.59BulletCatcherWe could just remove the directory from PATH and go on, but we'd only want to do tha on MacOS --> platform check.
05:59.06brlcadFlash: which is bogus
05:59.29FlashI see that it is explicitly added to my path in .tcshrc, possibly because I was playing with Tcl/Tk
05:59.30brlcadit's R7 now, the link is just for compat
05:59.35Flashtrue
05:59.50AAA_awrightIs the BZFlag source ever gonna become standardized in tabs vs. spaces?
05:59.52*** join/#bzflag delusional (n=delusion@151.201.237.195)
05:59.59BulletCatcheradds it in .cshrc too
06:00.12AAA_awright(tabs for indenting, spaces for aligning, get it straight please!)
06:00.26brlcadit is already
06:00.27Flashhands AAA_awright a real editor
06:00.58AAA_awrightvim, MSVC, doesn't work by default.
06:01.14brlcadso?
06:01.19Flashgives AAA_awright emacs
06:01.22DTRemenakvim does.  it reads the ex modeline at the bottom of each file.
06:01.26brlcadthere is a vi-line
06:01.40DTRemenakunless you have it in some wierd mode
06:01.41AAA_awrightdoesn't seem to work on my system, I'll check on another
06:01.45brlcadand DTRemenak wrote a snazzy ex modeline compat for msvc
06:02.03AAA_awrightPoint is anyways, it isn't portable, even if it /can/ work
06:02.23Flashif it CAN work, doesn't that make it portable?
06:02.32DTRemenaksure it is.  every editor worth its salt has the ability to set indent style.
06:02.37brlcadAAA_awright: the point is that the *code* is consistent (and if it's not, it's usually isolated cases that have slipped in or where others haven't configured their tools)
06:03.03DTRemenakmsvc can be set correctly with no tools, so can vi.  if you use different defaults, then we've offered several techniques to set it on a per-file basis automatically.
06:03.11AAA_awrightIt is consistent with respect to the fact there is no consistency :p
06:03.25brlcadthat's utter BS, it works just fine
06:03.41DTRemenakthe entire codebase is 2-space indent, 8-space tab, tabs not expanded.  how is that inconsistent?
06:04.32brlcadthat style is written *into* every single file, and into the developer guide .. what more do you want? (rhetorical question)
06:04.51brlcadus to set up your own damn tools for you? (another rhetorical question)
06:04.53AAA_awrightHm, I don't know what I am seeing then
06:06.10BulletCatcherOn MacOS 10.4 I moved aside my .cshrc file and logged out.  After logging in again I still have /usr/X11R6/bin in my path within xterm.  It seems to be a feature.
06:07.12brlcadBulletCatcher: I did say "unless you[']r[e] in an xterm" :)
06:07.44brlcadit's not a system default, it's just an xinit default
06:08.12brlcadpoint only being that it should be easy enough to get it working for Terminal
06:09.04brlcadgetting it to work with freetype-config in the path could be more tricky .. but then for that there really should be some other way to break the libgl vs openGL framework cycle in 10.5
06:09.43forrestvwould anyone be interested in a python version of bzflag?
06:10.20BulletCatcherTerminal, without my .cshrc, indeed does not have the X bin directory in its path.
06:11.14Flashon 10.5, I explicitly set PATH to /usr/bin:/bin then (simulated) su - anotheruser and X11/bin is in the path
06:11.23Flashlooks like they REALLY want to help
06:11.51Flash/Library/Frameworks/Python.framework/Versions/Current/bin:/usr/bin:/bin:/usr/sbin:/sbin:/usr/local/bin:/usr/X11/bin
06:11.55Flashinteresting
06:12.48BulletCatcherforrestv: Rewriting BZFlag in Python (or any other language) would be a lot of work.  Is there a compelling reason for it?
06:13.12AAA_awrightBecause it is fun
06:13.18forrestv:)
06:13.23AAA_awrightBZFlag port to PHP woo!
06:13.36forrestvBulletCatcher, besides that, it would be easier to maintain
06:14.02*** join/#bzflag L4m3r (n=l4m3r@about/essy/warning/L4m3r)
06:14.02*** mode/#bzflag [+v L4m3r] by ChanServ
06:14.11AAA_awrightNo kidding, If there is one good thing about python, it is that proper indenting is required. I really don't like python at all usually.
06:14.12brlcadforrestv: you're familiar with pybzflag?
06:14.32BulletCatcherAAA_awright: lol
06:15.05forrestvbrlcad, i looked... i don't think it's functional
06:15.15Flashit certainly isn't up to trunk
06:15.33brlcadof course not, just wondering if you'd heard of it'
06:15.38brlcadas you're proposing the same thing
06:15.50brlcadthe python bzflag effort several years ago, during late 1.7 days
06:16.04forrestvbrlcad, screenshot looks nice: http://static.bzexcess.com/pybzflag/python%202008-01-29%2022-40-25-54.png
06:16.10brlcada couple of the devs made great progress in a pretty short timeframe
06:16.19AAA_awrightIt isn't too hard if you know how the protocol works, I have a from scratch PHP script that can join and play as a robot, and another that can function as a server... somewhere
06:17.04AAA_awrightHas anyone ever talked about a NDS port?
06:17.49brlcadbut that progress spoke more of the devs doing the work than the language itself imho .. and the fact that they didn't get to implementing anything more than proof-of-concept (graphics+net, pretty much no game logic)
06:18.34brlcadforrestv: it would be a heck of a lot easier to update pybzflag than to start over yet again if you're interested in working on it
06:20.12brlcadyou're not going to get buy-in from the other devs, though .. that's about a fruitful discussion as picking just about any random language and/or forking for whatever reason -- you're welcome to it, but it wouldn't be "BZFlag"
06:21.06BulletCatcherIf you just want a challenging Python project there must be many other better choices.
06:21.27macsforme~donny_baker++
06:21.48brlcadwrite a 2D bzflag client in python, or a network protocol testing framework
06:22.24forrestvbrlcad, i've already started actually. i have the textures and models working. i'm adding network support now...
06:22.38AAA_awrightDitch python altogether, use Ruby I would say. Just my opinion.
06:23.38FlashAAA_awright: you just don't like enforced indentation ;-)
06:23.54AAA_awrightI do actually
06:24.28AAA_awrightIf I wrote a language I would take that hint, but require tabs. And require objects too, which is the part I don't like.
06:24.37AAA_awrightabout Python.
06:24.54AAA_awrightOh, they broke the = operator too. That is what got me upset.
06:25.06purple_cowwhen all you have is OO, everything looks like a design pattern?
06:25.29AAA_awrightI believe so?
06:26.04FlashI'm convinced that academics invent new labels every so often so they can talk about what the rest of us are already doing and sound smarter than us
06:26.22Flashsuch as design patterns
06:26.25AAA_awrightHeh :)
06:27.16BulletCatcherIf we would just stick with FORTRAN-66 we wouldn't have to have these arguments. :-)
06:27.30purple_cowsighs
06:27.31BulletCatcherBack to freetype-config on MacOS 10.5, should we just leave it as an open bug for now?  The workaround isn't too onerous.
06:27.57FlashI'm lazy, but I have no power, so I can tolerate whatever you come up with ;-)
06:28.06AAA_awrightbrlcad: About the code style, vim doesn't change anything for me, so idk, and in many cases, blocks of code are not indented, but un-indented to apparently reduce whitespace, which makes it very hard to read
06:28.31DTRemenakthat sounds like bad tabwidth settings
06:28.36DTRemenaktabs are equivalent to 8 spaces
06:28.41AAA_awrightThere are no tabs I can count
06:28.56Flashwhich file(s)?
06:29.02DTRemenakany line indented more than four times should have a tab rather than spaces at the beginning of the line
06:29.26AAA_awrightMost all of them I have ever seen, I am looking at ListServerConnection.cxx right now
06:30.38AAA_awrightthat makes sense though
06:30.57Flashpick a poorly formatted line number
06:31.12CIA-33BZFlag: 03dtremenak * r17252 10/trunk/bzflag/ (3 files in 2 dirs): rewrite HUDuiTypeIn to handle multibyte strings
06:31.32DTRemenakwoot, now it at least behaves mostly appropriately :)
06:31.32AAA_awright219
06:31.48Flashhas 2 tabs
06:31.52AAA_awright>_>
06:32.13Flash== 16 spaces or 8 levels of indent ;-)
06:32.37AAA_awrightI have never seen anyone use that convention, I always use tabs for indenting, spaces for aligning. Makes things portable *and* customizable.
06:32.38Flashsounds like you don't have your tab width properly configured
06:33.04DTRemenakAAA_awright: "set modeline" in ~/.vimrc
06:33.27DTRemenakprobably your distro disabled it from security concerns
06:33.40AAA_awrightI'll have to do that now, but my point is do not rely on tabs to be a specific setting. aka do not mix tabs and spaces.
06:33.50AAA_awrighter, e.g. do not mix them
06:33.57DTRemenaktabs are eight spaces.  you would be *shocked* how many utilities depend on that
06:34.21Flashit's a religious point, I've heard it before. it doesn't bother me
06:34.27brlcadBulletCatcher: if it works with Terminal (i.e. there's a working path), then I'd say it's fine for now at least
06:34.28AAA_awrightI don't work on too much C/++ code though
06:34.43BulletCatcherLine 431 of the top-level configure.ac already has some Darwin-specific OpenGL checks.  Could we adjust that to detect the 10.5 problem?
06:34.57Flashk
06:35.05Flash(oops... wrong window)
06:35.34DTRemenakis excited
06:36.02DTRemenakI can type russian characters, send it to the server, and get it back :)
06:36.20purple_cowcongrats
06:36.23AAA_awrightDTRemenak: Is that some published standard? The only tools that I know of that have that set is notepad.exe and vi/m
06:36.23BulletCatcherBut can you read what they say? :-)
06:36.45DTRemenakBulletCatcher: no, unfortunately
06:37.01DTRemenakAAA_awright: what is the antecedent of "that" in your question?
06:37.03brlcadAAA_awright: then you just haven't been exposed to much .. there's plenty of projects for just about every indentation style imaginable, major popular projects
06:37.15AAA_awrightthe tab=8 spaces
06:37.30DTRemenakunix machines have done that for many many years
06:37.39brlcadit's pretty darn common to find tab-width of 8 and an indent width of 2/4/5
06:37.54DTRemenak8/5 is just silly :)
06:38.05brlcadgood ol original k&r
06:38.41AAA_awrighto.O
06:39.06Flashmy personal favorite is c-set-style=stroustrup but as long as the editor can enforce it, I don't care
06:39.07purple_cowDTRemenak: nonsense, 40 is an awesome number
06:39.42DTRemenakheh...if you indent that far, you probably need to consider refactoring :)
06:40.01purple_cowit kind of wigs me out that we use 3-space indent at work
06:40.18DTRemenakI've always found 3-space to be quite aesthetically pleasing
06:40.23DTRemenakbut I'm weird like that
06:40.44brlcadiirc the reason was because if you use if (something ) { in k&r and then have a something that spans multiple lines .. the something lines up evenly with the if-body making them hard to distinguish at 4 chars
06:41.07purple_cowI guess it'd be a lot worse if we didn't disallow all tab characters
06:42.10brlcadFlash: that's actually what we use for BRL-CAD (random factoid)
06:42.22brlcadeven on the C code as it's the closest to k&r with 4-char indents
06:42.36brlcadjust have to set/fix the case-indentation
06:42.40BulletCatcherFlash: at this point it seems that the best thing would be for you to open a SF bug about the MacOS 10.5 configure problem.
06:42.58Flashok
06:43.05brlcador even better, find a fix
06:43.33brlcadfind a fix that lets it use system freetype
06:44.06BulletCatcherTo paraphrase JeffM: MacOS is broken; fix it! :-)
06:44.26Flashbefore I make the bug, I thought I remembered hearing that build system problems weren't bugs
06:44.32brlcad10.5 does have lots broken wrt /usr/X11
06:44.58brlcadthey jumped to latest and greatest xorg sources with a ton of issues unresolved
06:45.21BulletCatcherThe same thing happened with Fedora 9.
06:46.18FlashI believe there is (currently) no fix that can allow OS 10.5 to use the system freetype
06:46.34FlashThat's because I'm assuming that the GL libs need to follow freetype in the link order
06:47.08Flashand the system freetype puts -L /usr/X11/lib in the link line just before freetype
06:49.15AAA_awrightAha, that works.
06:49.24AAA_awrightBut makes no sense
06:49.32AAA_awrightYet it works
06:51.21AAA_awrightWhat was the reasoning behind indent=2 spaces, tabs=8 spaces? If you set indent=1 tab, you can change the spaces per indent by changing n in tabs = n spaces, which makes so much more sense.
06:51.38Flash(to you)
06:51.42AAA_awrightPlus it may be easier to compress?
06:52.27DTRemenakthe fixation on tabs=8 spaces is why that was chosen.  and 8 spaces is too much for indentation on 80-column monitors.
06:52.36DTRemenakso something smaller was picked for that
06:53.16brlcadFlash: there's nothing that prohibits modifying the result from freetype-config .. hence why I said "find" a solution.. there are lots of ways to end up with something generic that'll work
06:53.18AAA_awrightPlus, I also like the concept for the elastic tabs for aligning content
06:54.39brlcadAAA_awright: seriously, that's the convention established .. bikeshed decision that is done and not likely going to change anytime soon
06:54.56brlcadmore important that it's consistent than it fit anyones personal favorite
06:55.01AAA_awrightI just find this fascinating
06:56.24BulletCatcherdoesn't understand the obsession with form over content
06:56.31purple_cowah, the exuberance of youth
06:56.51brlcadpurple_cow: you must be getting old :)
06:57.05CIA-33BZFlag: 03dtremenak * r17253 10/trunk/bzflag/TODO: remove some finished items
06:57.39purple_cowbrlcad: yeah, I've become a crotchety, er, mid-twenties guy
06:58.01Flash~lart purple_cow for being a young kid
06:58.01ibotbeats the living hamstercrap out of purple_cow for being a young kid
07:10.20CIA-33BZFlag: 03dtremenak * r17254 10/trunk/bzflag/src/bzfs/bzfs.cxx: Fix bug #1970976. Don't crash when setting a single global bzdb var from the client. Note to self: check for everywhere that packs a message before you go and change the format of it.
07:12.50CIA-33BZFlag: 03dtremenak * r17255 10/trunk/bzflag/src/date/buildDate.cxx: numbers are cheap, bump the protocol for that bugfix.
07:13.27BulletCatcherDon't forget to increment the number in configure.ac, too.
07:13.43DTRemenakbah, nobody checks that number anyway :P
07:14.11CIA-33BZFlag: 03dtremenak * r17256 10/trunk/bzflag/configure.ac: fine, fine, this number too
07:14.23BulletCatcherI am probably the only one who does "make dist" at this point, but that's where it shows up.
07:15.37AAA_awrightIs there any way to get CIA to call a 3rd party script/URL on an update? It could be possible. Or is pulling updates (RSS) the only way? I can't find any such method to do so.
07:18.52brlcadshould unify the version number like he's done elsewhere
07:18.59brlcad"there can be only ONE!"
07:19.45BulletCatcherThat is now on my list, but at low priority.
07:20.40FlashDTRemenak: there should only be one set of code to pack (or unpack) a given message
07:20.43Flashhas a dream
07:21.09DTRemenakI'd tend to agree.  I see why this was done though.
07:21.32BulletCatcherI hope to give your net.patch a try this weekend, Flash.
07:21.56FlashI'm debating uploading a windows-friendly update, but Jeff's last comment somewhat deters me
07:22.07DTRemenakone of them packs a single variable in a single message from a callback, the other packs the entire state database (n variables in m messages).
07:22.47DTRemenakwhich comment is that?
07:23.37FlashMy recommendation would be to work on this not as a patch but as a branch
07:23.37Flashin some VCS so that it can be done incrementally and kept up to sync.
07:24.02DTRemenaksounds like a good idea to me
07:24.14Flashbut beyond my ability....
07:24.40DTRemenaknot really
07:26.06Flashas I understand it, the point of a VCS would be so others could collaborate
07:26.31DTRemenakmmm...not so much.  more to make the review process easier.
07:26.44DTRemenakalthough if you wanted others to contribute, that might be valuable too
07:26.48Flashthen how would it differ from a patch?
07:27.00DTRemenakit's a series of patches with commit messages
07:27.30DTRemenakrather than just "I changed these 30,000 lines" it's "I changed these 50 lines on this day for this reason, and then these 160 on this day for that reason" and so on
07:27.47AAA_awrightUnrelated to the tabs/spaces issue, MSVC breaks/doesn't have multi-line indenting >:( . Related to the issue, it breaks un-indenting when there are no spaces, only tabs indenting the code.
07:27.59FlashI'm not prepared to set up or host a project for this. Maybe it's easier than I think, but it's just another skill I haven't acquired yet
07:28.30AAA_awrightBranching ftw
07:28.31DTRemenakit's pretty easy to throw up a new project on google code or sf.net, or you could ask to work in a branch of ours
07:28.53DTRemenakor you can just keep working the way you have been.  uploading more often is better
07:29.05FlashI think I'm waiting for Jeff to decide if he trusts me or not ;-)
07:29.09DTRemenakthat's part of the advantage of the vcs...uploads multiple times a day
07:29.10AAA_awrightFlash: If you like Git, I have a Git import I keep in sync with the source, you can branch from that
07:29.23FlashI don't know git
07:29.35Flashthinks that sounds too close to something else
07:29.41BulletCatcheris glad that isn't a typo!
07:29.42AAA_awrightIt's simply brilliant, but I won't get into that here
07:30.37AAA_awright(This is *exactly* the sort of situation I talk about when I say Git would be ideal for BZFlag if it worked well on all the platforms BZFlag runs on)
07:31.02BulletCatcherThat seems like a critical "if".
07:31.08AAA_awrightIndeed
07:31.29AAA_awrightActually, there is git-cvsserver, so clients can check out with that
07:31.30DTRemenakyes, this situation is exactly what dvcs' are intended for.  not that we run into it that often.
07:32.01AAA_awrightThe main devs wouldn't, but smaller contributers do
07:32.32DTRemenaksmall contributors typically have neither need nor desire for a vcs.  it's the people who do bigass patches that do :)
07:33.38AAA_awrightWhich is exactly why I love Git so much, it encourages you to do branches and merges that on other VCS's would be one large commit or something
07:35.23FlashBulletCatcher: I went ahead and updated it. Have fun
07:35.53BulletCatcherTommorow... I must sleep soon.
07:36.01Flashsleep is highly overrated
07:36.15FlashI think I have 7 graduation parties to go to tomorrow/today
07:36.55BulletCatcherI get to fly across the country for one next week.
07:37.26Flash<PROTECTED>
07:37.58BulletCatcherIndy 500 party!
07:43.50*** join/#bzflag delusional_ (n=delusion@pool-151-201-221-6.pitt.east.verizon.net)
07:50.27brlcadwonders if Flash fell asleep
07:50.43DTRemenak'tis about that time
07:50.53DTRemenakis on the way to falling asleep himself
07:51.20BulletCatcherOn Linux: iswspace() not declared in scope in KeyManager.cxx
07:52.07DTRemenakyeah, doesn't surprise me...that's one of the "issues" I mentioned in my original commit message
07:52.13DTRemenakinclude wctype.h iirc
07:52.32BulletCatcherYes, wctype.h looks like the answer.
07:53.39CIA-33BZFlag: 03dtremenak * r17257 10/trunk/bzflag/ (include/Makefile.am src/common/Makefile.am): add new files to *nix build system
07:54.10brlcadDTRemenak: that's crazy talk
07:54.17brlcadit's nowhere near that late :)
07:54.28DTRemenakBulletCatcher: and I have no idea if wctype is unavailable on some platforms or if it's universal, or what has and does not have the appropriate functions
07:54.43DTRemenakI figured someone else would figure it out if I broke the build bad enough ;)
07:55.05BulletCatcherFedora and MacOS both seem to have it.
07:55.15DTRemenakyeah, all reasonably modern gccs should
07:55.34BulletCatcherIt'll be less broken if you just assume it is there.
07:55.43DTRemenakgo ahead then :)
07:55.59DTRemenakdidn't keep track of which files he added wchar functions too
07:56.11BulletCatchersummons the power of svn
07:57.18BulletCatcherIt'll be a few minutes while I determine the scope of the problem.  "make -k" is my friend.
07:57.51DTRemenakdon't forget to re-autogen since I added the new files to Makefile.am
07:58.07DTRemenaksorry I forgot that earlier
07:58.18DTRemenakwas too excited :)
07:58.39BulletCatcherYup.  I almost always start with a clean tree and autogen.sh.
07:59.13DTRemenakI can type and funny characters come out, and nothing breaks! :) фывапролджэ!
07:59.24BulletCatcher:-)
07:59.44DTRemenakhopes he didn't just inadvertently insult someone's mother
08:00.51AAA_awrightWhat happened to the "Start Server" menu on BZFlag trunk? It was there at least a week ago.
08:01.09BulletCatcherHow do you make it go into Russian mode?  LANG=ru or some such?
08:01.42DTRemenakhowever you change your locale on your os, for input.  the options menu, for gui l10n.
08:01.49BulletCatcherThat menu item was removed within the last week, AAA_awright.
08:01.55AAA_awrightMakes sense
08:02.23AAA_awrightI was using it, so it is just a bit inconvenient
08:02.48DTRemenakI think the idea was to encourage people to work on BZWTestLauncher
08:03.07AAA_awrightAs a stand alone program, makes sense
08:03.18AAA_awrightBrilliant actually
08:03.36AAA_awrightWho thought of that?
08:04.06DTRemenakjeff wrote one in...'97 or so?  crs didn't like it.
08:04.31AAA_awrightRemoving the menu item that is
08:04.33ManuFlash: welcome ;)
08:04.40DTRemenakjeff did that too
08:05.07DTRemenakpeople were complaining that it was broken
08:05.35DTRemenakand since we were moving away from that idea anyway, it made sense
08:05.45*** join/#bzflag QuantumT1m (n=tim@62-249-230-206.no-dns-yet.enta.net)
08:06.10AAA_awright~JeffM++
08:06.29AAA_awrightSomeone needs to merge the karma for JeffM and JeffM2501 sometime
08:06.49DTRemenakmost of us with split personalities also have split karma pools ;)
08:07.58AAA_awrightI need a second online ego that is evil and attracts negative karma like Cannon Ball Guy... (!) Or maybe not
08:08.28AAA_awright;)
08:09.43BulletCatcherWarnings in HUDuiTypeIn.h are going to annoy those who configure with --enable-debug ...
08:09.54DTRemenakfix'em :)
08:10.02DTRemenakwhat'd I do?
08:10.36BulletCatcher'CountUTF8StringItr::counter' will be initialized
08:10.52DTRemenakah, got something out of order
08:11.57CIA-33BZFlag: 03dtremenak * r17258 10/trunk/bzflag/src/bzflag/HUDuiTypeIn.h: init base class before member functions so gcc doesn't cry
08:12.54AAA_awrightAlright, this should remove most all of the password logic in BZFS, I don't know what is in the client and such, and I have a plugin in progress that replaces the commands http://bzflag.pastebin.ca/1027762
08:13.38AAA_awrightAnyone want to check it and I'll post it to sf.net later on
08:13.59DTRemenakis going to bed
08:14.16brlcadnight!
08:14.21brlcadwuss.. :)
08:14.24DTRemenakpfft
08:14.56Manubood night DTRemenak :)
08:14.59Manugood*
08:15.09brlcadbood morning manu :)
08:15.26Manu</irony>
08:15.28Manu:)
08:15.39Manuhola brlcad :)
08:17.12Manubrlcad: hi bood ;)
08:20.11Flashis there a norang admin about?
08:26.13CIA-33BZFlag: 03bullet_catcher * r17259 10/trunk/bzflag/src/ (3 files in 2 dirs): Include wctype.h as needed.
08:27.13BulletCatcherSeveral warnings remain, but that should allow a successful non-debug compile.
08:36.59CIA-33BZFlag: 03jwmelto * r17260 10/trunk/bzflag/README.MacOSX: add special instructions for 10.5
08:37.44*** join/#bzflag QuantumTim (n=tim@62-249-230-206.no-dns-yet.enta.net)
08:37.55brlcadwoot!
08:45.39BulletCatcherCongrats, Flash!  I didn't realize you had been promoted.
08:47.13Flashnow I'm a potato peeler
08:47.20BulletCatcher:-)
08:47.32Flashbut don't trust me with a sharp knife
08:47.59brlcadthey don't call him "mr. stabby" for nothing
08:48.56BulletCatcherr17259 compiles successfully for me.  'night guys.
08:49.30Flashwonders if r17260 will compile....
08:49.52BulletCatcherIf it doesn't, you get to fix it!
08:54.59CIA-33BZFlag: 03jwmelto * r17261 10/branches/v2_99_net_branch/: lock me in a closet
08:56.42*** join/#bzflag Upsetter (n=Ups@i577B7129.versanet.de)
08:57.15L4m3rit seems that my daily bzfs build has not built successfully since the 20th
08:57.40L4m3r...and I've modified the script to try again if the build segfaults, so that is not the problem :P
08:58.08BulletCatcherthinks Flash's branch is more like a fishbowl
08:58.21Flashgo to sleep
08:58.30L4m3r(fortunately the script does not kill and restart bzfs if the build fails)
08:58.33AAA_awrightYeah, go to sleep
08:58.40AAA_awrightplunks to the ground
08:58.53Flash~sleep BulletCatcher
08:58.57ibotACTION sends BulletCatcher to bed without supper
08:59.06BulletCatchercomplies
08:59.07L4m3rimpales AAA_awright with a rebar rod to keep him down
08:59.15AAA_awrightNot as cool as Divibot:
08:59.31AAA_awrightAAA_awright plunks to the ground. Goodnight, AAA_awright!
08:59.49AAA_awrightSo about that patch...
09:00.38AAA_awrightAnyone want to examine it, I have not really tested it besides compiling
09:00.52AAA_awrightAnyways, laters
09:00.57Flash~wake BulletCatcher
09:00.57ibotACTION throws a barrel-full of ice water on BulletCatcher and shouts "GOOD MORNING!!!!"
09:01.47FlashBulletCatcher: never mind. go back to sleep
09:01.52Flashthrows BulletCatcher a towel
09:08.48FlashDTRemenak: FTFont/FTBufferFont.cpp doesn't compile. Don't know if it's your change or BulletCatcher's
09:08.51FlashFTFont/FTBufferFont.cpp:205: error: 'wcsdup' was not declared in this scope
09:08.53Flashsleeps
10:44.55*** join/#bzflag SpazzyMcGee (n=SpazzyMc@unaffiliated/spazzymcgee)
10:54.07*** join/#bzflag Wits (n=Warinthe@212.119.33.65.cfl.res.rr.com)
11:10.59*** join/#bzflag kirun (n=kirun@222.5.gr6.adsl.brightview.com)
11:19.06tsHow is wcsdup() defined? My system does not have it, as it is not standard.
11:25.50Wyk3dJBdiGriz: ping
11:35.42jorgenptts:        The  wcsdup() function is the wide-character equivalent of the strdup(3) function.  It allocates and returns a new
11:35.46jorgenpt<PROTECTED>
11:35.48jorgenpt<PROTECTED>
11:36.22tsHmm
11:36.41tswonders whether it is possible to work around with wcsncpy
11:37.40jorgenptts: What about malloc & wcslen & wcsncpy?
11:39.19tsI thought this is enough:
11:39.22ts<PROTECTED>
11:39.22ts{
11:39.22ts<PROTECTED>
11:39.22ts<PROTECTED>
11:39.22ts}
11:39.33jorgenptwchar_t *wcsdup(const wchar_t *s) { size_t len = wcslen(s); wchar_t *new = malloc(sizeof(wchar_t) * (len + 1)); wcsncpy(new, s, len + 1); }
11:40.11jorgenptts: not sizeof(wchar_t) in wcsncpy - it takes a number of wide characters, not bytes.
11:40.35jorgenptAnd I doubt you need to specialcase for 0, not that you'd want that.
11:40.54jorgenptYou'd still want a string of the length is zero. OTOH, you should probably check for s == NULL
11:41.13jorgenptwchar_t *wcsdup(const wchar_t *s) { if (s == NULL) return NULL; size_t len = wcslen(s); wchar_t *new = malloc(sizeof(wchar_t) * (len + 1)); wcsncpy(new, s, len + 1); }
11:41.37jorgenptwchar_t *wcsdup(const wchar_t *s) { if (s == NULL) return NULL; size_t len = wcslen(s); wchar_t *new = malloc(sizeof(wchar_t) * (len + 1)); wcsncpy(new, s, len + 1); return new; } /* Of course, my bad. */
11:42.14jorgenptHmm.
11:42.44jorgenptYes. That should work.
11:44.11jorgenptAnd perhaps even check if 'new' is NULL after you malloc it, to ensure correct behaviour when OOM.
11:45.11tsWell, I don't really know what I am doing :)
11:45.49jorgenptWell, try this:
11:45.53jorgenpthttp://pastebin.ca/1027850
11:46.13jorgenptPut that in some common.c or whatnot, and add a #ifndef _GNU_SOURCE \n #
11:46.19jorgenptwchar_t *wcsdup(const wchar_t *s);
11:46.21jorgenpt#endif
11:46.23jorgenptto some common.h ;-)
11:47.44jorgenpthits the shower
12:12.10*** join/#bzflag whodaman- (n=whodaman@pdpc/supporter/student/whodaman-)
12:18.37tsjorgenpt: I don't get it to work and other work awaits me :(
12:18.53jorgenptHow does it fail?
12:19.03jorgenptOh well, guess we'll postpone it. No rush, was just a suggestion :)
12:31.00*** join/#bzflag hwilde (n=hwilde@pool-68-162-175-181.pitt.east.verizon.net)
12:31.21hwildehow can one player have vertical shot velocity if nobody else on the server has it
12:32.53hwildeand super bullet and wall walking
12:33.00hwildeand he can fly with laser flag
12:35.06hwildepretty cool cheat, I thought some of that stuff was server side
12:38.27hwildecheck out frederix on badgerking
12:42.50spldartmissenden now.. all the cheats you listed I got him
12:44.19hwildewings and vertical velocity are pretty dominating if nobody else has them
12:46.08spldartI noticed.. I was following missended through 5 or 6 kills
12:46.13spldartbtw
12:46.28spldarthe was invincible. At lease selectively.
12:46.58spldartI'm goin to poke in again
12:53.50spldarton badgerking TRIPITRAPI player isn't moving.. isn't responding to me... But wherever he is.. Has only moved once he's not being noticed/ playing
12:54.01spldartoh well
12:55.18hwildei think that guy is ok
12:55.31*** join/#bzflag whodaman- (n=whodaman@pdpc/supporter/student/whodaman-)
12:58.43*** join/#bzflag Wits (n=Warinthe@212.119.33.65.cfl.res.rr.com)
13:30.06*** join/#bzflag whodaman- (n=whodaman@dslb-088-066-017-253.pools.arcor-ip.net)
13:35.01Winnywoo, new this week in photography
13:39.02spldart?
13:39.32Winnyhttp://twipphoto.com/ ...
13:40.57spldartmeh.. My last 35mm camera was a half decent Olympus... Haven't brushed the dust of it in years
13:41.16spldartI think it 8 years to be more exact
13:42.23Winny:o
13:43.30spldartI remember it like yesterday.. Took pictures of other competitors cars when I last competed at world finals. Ah the memories :)
14:02.25*** join/#bzflag TD-Linux (n=wheeeeee@about/essy/indecisive/TD-Linux)
14:41.31*** join/#bzflag whodaman- (n=whodaman@pdpc/supporter/student/whodaman-)
14:56.24*** join/#bzflag me1 (n=me1@unaffiliated/me1)
15:24.35*** join/#bzflag Think_Differentl (n=shedelj@unaffiliated/thinkdifferentl/x-838543)
15:52.55*** join/#bzflag whodaman- (n=whodaman@pdpc/supporter/student/whodaman-)
16:16.12*** join/#bzflag KTL (n=KTL@213.219.162.66.adsl.dyn.edpnet.net)
16:21.49spldartQuestion about behaviour of stable client. In this case 2.0.10. I open client full screen. F1 to window then hit windows key to see something else real quick but find that bzflag hasn't just lost highlight but has crashed out. No message from windows and it doesn't remember the server you were on or setting you had.. It's as if it hasn't been opened since the time you had played before that.
16:22.29spldartThat isn't designed behaviou is it
16:22.46BulletCatcherIt saves settings when you exit cleanly, so a crash would bypass that.
16:23.23spldartthought so
16:23.37tsWhere is the question?
16:23.57spldartadd a question mark to 'That isn't designed behaviour is it'
16:24.27tsWell, crashes are not intentionally I would guess.
16:24.45spldartAs BulletCatcher just confirmed for me
16:25.01*** join/#bzflag JeffM (n=Jeff@pool-71-109-223-10.lsanca.dsl-w.verizon.net)
16:26.00*** join/#bzflag bryjen (n=bryjen@CPE-76-177-176-191.natsoe.res.rr.com)
16:26.00*** mode/#bzflag [+v bryjen] by ChanServ
16:28.15CIA-33BZFlag: 03t-s * r17262 10/trunk/bzflag/src/other/ftgl/src/FTFont/FTBufferFont.cpp: Hack around wcsdup usage
16:41.53CIA-33BZFlag: 03t-s * r17263 10/trunk/bzflag/ (4 files in 3 dirs): name conflict at class UTF8Char with system headers, rename our class to bzUTF8Char
16:45.10*** join/#bzflag gearth (n=1893fd30@gateway/web/cgi-irc/zeebrothers.net/x-761c4942037132be)
16:58.17Flashts: thanks, that seems to compile now
17:00.06JeffMdon't forget to get FTLG changes moved upstream
17:00.15KTLhttp://my.bzflag.org/w/LeftSkyboxMaterial  <- /me can't get this to work
17:00.31tsFlash: yeah, it's kind if hackish but...at least it works
17:00.43KTLand when i try another picture, i changed the downloadacces file
17:01.50JeffMKTL: does it draw anything?
17:02.01KTLnot one of the pics
17:02.04KTLblue sky
17:02.52KTLbzfs -world thatfile.bzw
17:03.32JeffMput the pictures in your data dir and then just try using those names
17:03.44KTLthe client debug output mentions the files, ... Image not found or unloadable: mountain4
17:05.14JeffMthat can be normal
17:05.20JeffMwhat version aer you using?
17:05.32KTL2.0.10.20080404-MAINT-linux-gnu-SDL
17:05.50JeffMhas it downloaded his images?
17:06.15KTLit did not mention them in the messagewindow of the client
17:06.27JeffMtry it with just simple local images
17:06.29*** join/#bzflag Slon02 (n=pusik10@c-71-224-170-197.hsd1.pa.comcast.net)
17:06.57KTLplaced 1 texture more in /usr/share/bzflag, no effect yet
17:07.11KTLwhere is the user space data directory supposed to be?
17:07.18KTLjust .bzf?
17:07.22JeffMhuh?
17:07.27JeffMuser space data?
17:07.41KTLwell, /usr/share/bzflag is only accessible by root
17:08.00JeffMis that where all the game textures are?
17:08.10KTLyes
17:08.20JeffMthen how does the game run normaly?
17:08.22KTLwritable by root i mean
17:08.27JeffMahh yeah that's normal
17:08.28KTLreadable by everyone
17:08.34JeffMjust throw them in there as as TEST
17:08.48JeffMand change the map to just use the file name
17:08.52JeffMno png or URL
17:10.16KTLweird weird
17:10.22KTLdid not help
17:11.00JeffMdude that is NOT helpfull
17:11.01JeffMpastebin
17:11.06KTLhttp://pastebin.com/m73d5fe73
17:11.26KTLi tried with the .png removed too
17:11.56JeffMno
17:12.07JeffMpng has to be removed
17:12.15JeffMadd a newline between material and name
17:12.21JeffMI don't belive they are on the same line
17:13.05KTLstill not, is the matref line needed?
17:13.19KTLi remove diffuse and metref too
17:13.20JeffMI'd take that out
17:13.27JeffMI don't think the wiki is a valid sample
17:14.06KTLstill nothing, i go search for a map that uses texturing online then
17:14.41BulletCatcherThe file should be /usr/share/bzflag/skybox1.png but in the bzw it should just be "texture skybox1".
17:15.42KTLboth checked, no result
17:16.03Flashwhy is it that when I svn up to ts's latest change, my trunk builds, but when I merge the same change to my branch it still fails in the same place?
17:17.10BulletCatcherDid you autogen.sh recently in your branch?
17:18.11Flashlast night, when I created it
17:19.07Flashtakes a memo to make sure trunk compiles before branching in the future
17:19.14BulletCatcherautogen.sh is required again after r17257
17:19.46BulletCatcherOh.  Your branch is later than that.
17:19.49Flashmy branch was r17261...
17:20.15Flashbut I'm trying that now too, just because nothing can be too weird to explain compile errors
17:23.24Flashok, now I'm deleting everything except SVN files and trying again
17:23.46FlashI see no reason that 17261 + 17262:17263 should not compile
17:25.55Flashgrrr
17:26.16Flashchecks to make sure trunk compiles after all
17:37.01spldarttrunk compiles for me :)
17:37.37BulletCatcherFlash: what is the output of "sdl-config --cflags" for you?
17:48.32*** join/#bzflag Saturos (n=Saturos@i59F76931.versanet.de)
18:08.44*** join/#bzflag Paul_PL (n=pawelbl6@chello087206248103.chello.pl)
18:50.14*** join/#bzflag Saturos (n=Saturos@i59F76931.versanet.de)
18:55.25*** join/#bzflag Wyk3d (n=middle@cl-86-125-183-148.cablelink.mures.rdsnet.ro)
19:02.46CIA-33BZFlag: 03dtremenak * r17264 10/trunk/bzflag/src/other/ftgl/src/FTFont/FTBufferFont.cpp: rewrite StringCopy. wcscpy is not a standard function; we have logic for doing it ourselves already (minus the len = 0 case), so handle len = 0 and do it all ourselves.
19:04.15CIA-33FTGL: 03dtremenak * r1163 10/trunk/src/FTFont/FTBufferFont.cpp: from BZFlag r17264: rewrite StringCopy. wcscpy is not a standard function; we have logic for doing it ourselves already (minus the len = 0 case), so handle len = 0 and do it all ourselves.
19:04.55Wyk3dJBdiGriz: ping
19:05.42DTRemenakerr, both of those should read wcsdup, not wcscpy :)
19:08.03DTRemenakthinks the r17261 commit message is just *begging* for bad "coming out of the closet" puns
19:14.14*** join/#bzflag tallelfin (n=tallelfi@24.96.228.122)
19:16.58*** part/#bzflag tallelfin (n=tallelfi@24.96.228.122)
19:17.19Wyk3dJeffM/brlcad, anyone: ping ?
19:17.27DTRemenakWyk3d: pong
19:18.23Wyk3dwell the coding period should start on monday and i don't even know if i'm supposed to do it in a branch or not :)
19:18.36DTRemenakheh :)
19:19.04Wyk3di saw that flash got a branch for his netcode changes
19:19.33Wyk3dthing is, the first step i had planned for the project included some netcode refactoring
19:20.03DTRemenakwhat for?
19:20.03Wyk3dnot as extensive as flash's stuff.. just moving it to the net project, separating game logic
19:20.23Wyk3dto reuse it for the daemon
19:21.08DTRemenakah, THERE's your application
19:21.17DTRemenakhad trouble finding the darn page
19:21.32Wyk3dso now i'm not sure if i should wait until flash finishes it to a point where i can work with it / help him with it
19:21.49Wyk3dor if i should do it separately ..
19:23.23DTRemenakI'd say do it separately, even if it means redoing some of it when we merge.  You'll get a better feel for how things work earlier.
19:23.50DTRemenakalternatively if you have something that doesn't depend on it, you could work on that first
19:24.11DTRemenakand you will most likely be working in a branch
19:24.14*** join/#bzflag JeffM (n=JeffM@cpe-75-83-217-37.socal.res.rr.com)
19:24.16bryjen~bzpastebin
19:24.17iboti heard bzpastebin is http://pastebin.bzflag.bz or a place to put large chunks of text so as not to flood a channel
19:24.36DTRemenaksince I think we're still "trying to get 3.0 out" or some such
19:25.08Wyk3dwell last time i spoke to jeffm, i was told that if the changes aren't too extensive they could go into trunk
19:25.33Wyk3dwaves @ JeffM
19:25.35DTRemenakyeah, but looking at your app, I'd say they're extensive enough they should not
19:25.37JeffMif they don't break anything or are seperate things
19:25.43JeffMyeah yours is VERY involved
19:25.55JeffMit would be tough for you to keep i t working
19:26.03JeffMso much that it would slow you down
19:26.09JeffMso  it's simpler for you to use a branch
19:26.13bryjenDTRemenak: i had to make a couple of tweaks to get trunk to compile here http://pastebin.bzflag.bz/da05a2dd
19:26.32bryjenwhere here == slackware
19:27.01DTRemenakyou sure it works?
19:27.16*** join/#bzflag Saturos (n=Saturos@i59F76931.versanet.de)
19:27.25DTRemenakdoesn't think HUDuiTypeIn will work right with a non-virtual UTF8StringItr::operator++()
19:27.30Wyk3dwell .. compared to other projects, i wouldn't say "very" .. but sure a branch it is then
19:28.05bryjenit didn't crash in the two minutes i ran it ;)
19:28.17bryjengoes to try some more
19:29.53bryjeni can't erase the 1st letter in a field
19:29.56bryjen:/
19:30.23DTRemenakhm, I didn't check normal characters :)
19:32.26CIA-33BZFlag: 03dtremenak * r17265 10/trunk/bzflag/src/bzflag/HUDuiTypeIn.cxx: constants need to match variable type
19:32.38DTRemenakthat should fix your first two issues, without casting
19:33.01DTRemenakthe third one is fine to commit.  second one I still think is not-quite-right.
19:38.00bryjengcc 4.1.2 complains "'class UTF8StringItr' has virtual functions but non-virtual destructor"
19:38.08bryjengcc 4.2.3 doesn't mind
19:38.14DTRemenakmake the destructor virtual then
19:38.26*** join/#bzflag Ride- (n=Ride-@24-197-250-9.dhcp.stpt.wi.charter.com)
19:42.49bryjengah. am i blind, or just not seeing one?
19:43.06DTRemenakthere might not be one :)
19:43.12DTRemenakcreate one if you want
19:43.22DTRemenakvirtual ~UTF8StringItr() {}
19:46.31CIA-33BZFlag: 03bryjen * r17266 10/trunk/bzflag/include/bzUnicode.h: gcc warned about NULL in 'math'
19:50.28CIA-33BZFlag: 03bryjen * r17267 10/trunk/bzflag/include/bzUnicode.h: gcc 4.1 likes classes with virtual functions to have virtual destructors
19:52.33tsDTRemenak: According to "man wcscpy" wcscpy is a standard function.
19:53.08Chestalts: C99 only I think?
19:53.32tsYep, C99
19:54.08Chestalwe use C++ though :-)
19:54.51DTRemenakts: as I said right after those commits, I meant wcsdup
19:57.34tsDTRemenak: Heh, I looked in backlog before I commented but didn't saw it
19:58.57*** join/#bzflag Tupone (n=Tupone@pdpc/supporter/active/Tupone)
19:58.57*** mode/#bzflag [+v Tupone] by ChanServ
19:59.05tsCiao Tupone
19:59.24Tuponehallo ts
20:01.50KTLeverything compiled withou warnings
20:01.51tsDTRemenak: Is UTF-8 based chat planned?
20:02.06DTRemenakts: planned?  it's all working
20:02.14DTRemenakthat was the whole point of those commits
20:02.14tsDTRemenak: Not here.
20:02.43DTRemenakworks for me.  debug it if it doesn't for you.
20:02.44tsDTRemenak: The interface is displayed ok..but chat isn't changed
20:03.09DTRemenakare you using an input mode that uses extended characters?
20:03.15tsLike when I type "abcßd" in my window "abcd" shows up
20:03.32DTRemenakdoes your font have the ß glyph?
20:03.36tsErr, what do you mean by input mode?
20:03.51tsMy font is Lucida Grande and yes, it has the char
20:04.17DTRemenakI mean, are you using a keyboard layout that sends the right characters
20:04.17DTRemenakon windows you can switch on the fly.  dunno how other platforms work.
20:05.30*** join/#bzflag ibot (i=ibot@pdpc/supporter/active/TimRiker/bot/apt)
20:05.30*** topic/#bzflag is http://cia.vc/stats/project/BZFlag || http://my.BZFlag.org/w/Getting_Help || logs at http://ibot.rikers.org/%23bzflag/ || http://ohloh.net/projects/189 || 2.0.10 is the latest public release || GSoC 2008, mentor blog at http://tanks4code.blogspot.com || Congratulations GSoC participants: http://code.google.com/soc/2008/bzflag/about.html || Students should add their projects to the wiki for review
20:05.30*** mode/#bzflag [+o ibot] by ChanServ
20:05.43DTRemenakdebug it then.  src/platform/<yourplatform>Display.cxx, set a breakpoint in getKey and make sure the character gets set to the correct unicode codepoint.
20:06.58tsI'll at wednesday..other priorities until then.
20:07.47DTRemenakyou can join a server and I can send you some text if you want, to prove it works :P
20:08.02DTRemenakthe SDL platform changes are all untested though
20:17.17JBdiGrizappears and catches up with the conversation
20:18.04JBdiGrizWyk3d: It looks like your questions were answered. Was there anything else?
20:19.25tsDTRemenak: Ok, name one server :)
20:21.36Wyk3dactually, no, that was about it :)
20:22.22DTRemenaktsÖ join one and Iäll join zou
20:22.28DTRemenakwhoops
20:22.33DTRemenakts: join one and I'll join you
20:23.39*** join/#bzflag LongDon (n=LongDon@p5B060346.dip.t-dialin.net)
20:24.14tsDTRemenak: 0 servers found and I am behind a NAT
20:24.30DTRemenakmmm, ok, give me a minute then and I'll update swordofkahless.
20:24.46blast007I'm updating mine as well  :)
20:29.32DTRemenakbets blast007 will finish before he does
20:29.46blast007never know
20:29.52blast007I haven't updated in a while
20:29.58DTRemenakneither have I
20:30.01blast007is yours a single core machine?
20:30.04DTRemenakyup
20:30.05DTRemenakP4 2.0
20:30.08blast007mine too  :)
20:30.11DTRemenakhah
20:30.11blast0072.53Ghz Celeron
20:31.13blast007I last built on the 3rd  :)
20:31.22DTRemenakmore recently than me :)
20:31.30DTRemenakmine's building in src/common right now
20:41.18*** part/#bzflag LongDon (n=LongDon@p5B060346.dip.t-dialin.net)
20:41.22*** join/#bzflag DTRemenak (i=DTRemena@about/essy/CrazyCoder/DTRemenak)
20:41.22*** mode/#bzflag [+v DTRemenak] by ChanServ
20:41.26blast007ts: do you see mine now?
20:42.11DTRemenaksee, mine's still building :(
20:42.22blast007;)
20:43.04blast007needs to build a client now too
20:44.47tsblast007: I'll check
20:46.56tsblast007: yes, I do
20:48.55*** join/#bzflag me1 (n=me1@unaffiliated/me1)
20:52.13KTLhttp://users.edpnet.be/dpf/bz/bzfi0079.png <- added textures
20:52.29Winnywhoa
20:52.52KTLdetects enthousiasm
20:57.38*** join/#bzflag Pimpi_ (n=frank@gondolin.pimpi.org)
21:06.39*** join/#bzflag whodaman- (n=whodaman@pdpc/supporter/student/whodaman-)
21:09.20bryjenfoo
21:09.27DTRemenakts: so y'see, it does work...on windows :)
21:12.12KTLgot a problem: http://users.edpnet.be/dpf/bz/bzfi0082.png <- is this due to my graphics card?
21:13.19tsDTRemenak: Not in my windows
21:14.01tsKTL: Wow, that looks Picasso-like
21:14.25KTLi'm not such a fan of picasso so how can i avoid this?
21:14.42KTLit might be because of the transparant png i used
21:15.31tsDTRemenak: Looks like it computes the 223 for ß
21:16.30DTRemenakthat's right, I think...U+00DF?
21:17.02tsDTRemenak: wikipedia says U+00DF..and DF_{16} = 223_{10}
21:17.07DTRemenakyeah
21:17.31DTRemenakthat's what it gets set to here too
21:17.49DTRemenakand at the end of the function key.chr is still 0xDF?
21:18.07tsI don't see where the value is returned
21:18.31DTRemenakit's not, it's passed by reference
21:19.07DTRemenakgetKey(..., BzfKeyEvent& key)
21:19.39tsI guess you mean callbacks
21:20.21DTRemenakno, no callbacks, just variables passed by reference.  like passing a pointer...if you modify it, the original gets modified
21:20.27tsDid I say I am not familiar with callbacks today?
21:20.52tsAh, that's the easy stuff
21:20.56DTRemenakyeah
21:21.02DTRemenakno funkiness
21:21.11tslooks what is at the end of the fctn
21:22.15tsYeah, the value remains 223 in getKey until the last return
21:23.30DTRemenakok...now how about in playing.cxx:916 (doEvent)?
21:23.55DTRemenakactually playing.cxx:786
21:24.05DTRemenakno reason to step through doEvent, try in doKey
21:24.27DTRemenakagain checking key.chr
21:27.06tssame there
21:27.10KTLhttp://users.edpnet.be/dpf/bz/bzfi0083.png <- hoho, i love this one
21:27.28Winny~KTL++
21:27.36KTLjawohhll my second karma point
21:28.19tsDTRemenak: Where could it be dropped?
21:28.35DTRemenakdunno, that should be after all the platform code.  and you're getting the same as I am.
21:28.57*** join/#bzflag JeffM (n=Jeff@pool-71-109-223-10.lsanca.dsl-w.verizon.net)
21:28.58DTRemenakcheck in HUDuiTypeIn::doKeyPress
21:29.04DTRemenakright at the beginning
21:30.45DTRemenakand if it's ok there, check at HUDuiTypeIn.cxx:174: what is c, and what is ch.buf
21:31.26tsOut of curiousity, what is reverted at playing.cxx:796?
21:32.31DTRemenakplaying.cxx:793-798 make sure that the menus get the keypresses if they're up
21:32.43DTRemenakdoes that answer your question (I didn't really understand it)
21:33.05tschecking 174
21:34.01DTRemenakch.buf[0] should be 0xc3, ch.buf[1] should be 0x9f
21:34.29tserroring..
21:34.44ts‘char* bzUTF8Char::buf’ is private
21:34.56DTRemenakch.str() if gdb won't override the permissions
21:35.41ts‘char* bzUTF8Char::buf’ is private
21:35.43tsoops
21:35.48DTRemenakch.str().c_str()[0] and [1]
21:35.56tsThat was supposed to be a make
21:36.02tsin a different window
21:36.07DTRemenakheh
21:36.23tsDTRemenak: I don't use gdb, I am using evil std::cout
21:36.35DTRemenakah
21:36.44DTRemenakgdb is a good tool, well worth the time to learn
21:36.47tsNot risking a locked up machine
21:37.35blast007run in a window with mouse grabbing turned off
21:37.36tsDTRemenak: It is especially well, when the window is in fullscreen mode, the event fires and I can't get out of fullscreen mode :)
21:37.48DTRemenakyeah, run in a window :)
21:37.58blast007it's not that hard
21:38.07Tuponerun with ssh from another computer
21:38.15blast007or that too  :)
21:39.05tsDTRemenak: after     std::cout << "HUDuiTypeIn::doKeyPress: bzUTF8Char: c: "<< c << " ch: buf" << ch.str() << "\n";
21:39.09tsarg
21:39.22tsafter bzUTF8Char ch(c); nothing is sample
21:39.24tsd
21:39.42tsLike the function isnt called
21:40.13tsso "something must eat the input" before
21:40.28DTRemenakdoes it output what you expect with a normal character (like 'a')?
21:40.54DTRemenakkey.chr should be 'a', ch.str()[0] should be 'a' also
21:41.12tsDTRemenak: yes
21:41.19DTRemenakinteresting
21:41.32tsc: 120 ch: bufx
21:42.07DTRemenakdon't forget the [0] and [1] when outputting ch.str()
21:42.12DTRemenak.str() returns a std::string
21:42.20tsOh
21:42.51DTRemenakso either do .c_str() after it (ch.str().c_str()), or output both ch.str()[0] and ch.str()[1]
21:43.31tsI see
21:44.50DTRemenakI don't see why you aren't getting there with ß though
21:45.09tsyep, ch.str()[0] is a
21:45.49DTRemenakdo you reach :166 with ß?
21:45.54tsDTRemenak: Obviously I get to the beginning of the function..I'll put some steps inside and see where the output stops
21:46.02DTRemenakk, thanks
21:46.35DTRemenakahh...maybe iswprint is returning 0?
21:47.40*** join/#bzflag bryjen (n=bryjen@CPE-76-177-176-191.natsoe.res.rr.com)
21:47.40*** mode/#bzflag [+v bryjen] by ChanServ
21:49.54tsit stops between 137 and 169..looking further
21:50.13DTRemenakyeah, check at 139
21:50.20DTRemenakbets it's iswprint being silly
21:52.42tsstops between 137 and 140
21:52.53tsmeaning it's this block:
21:52.54ts<PROTECTED>
21:52.54ts<PROTECTED>
21:53.04tss/block/box/
21:55.14tsiswprint(c) returns 0
21:55.27CIA-33BZFlag: 03dtremenak * r17268 10/trunk/bzflag/src/bzflag/bzflag.cxx: enable non-ascii locale handling on non-windows platforms. should make iswprint and friends behave.
21:55.33DTRemenaktry that
21:55.48tsHeh, that calls for conflicts
21:56.18tsah, it's only bzflag.cxx
21:56.32DTRemenakyeah, we never called setlocale on non-windows.  not sure why.
21:56.52DTRemenakso everything was behaving like the locale was ASCII
21:58.05tsWee, it works now
21:58.12DTRemenakcool :)
21:58.34tsNow I have to delete onethousand std::cout's
21:58.43DTRemenakdon't try binding those keys though (!)
21:58.46DTRemenakthat's still broken
21:58.56DTRemenakbut you should be able to type with them
21:59.28tsI also see new problems comming
21:59.40DTRemenakohh definitely :)
21:59.44DTRemenakwordfilter is broken also
21:59.52DTRemenakand certainly other things will need testing
22:00.01tsLike web browsers have their fun with fishing we can now happily avoid word filter :)
22:00.04DTRemenakcallsigns and such
22:00.26DTRemenakyeah, it'll be good times
22:00.29tsheh, callsign fishing >:-)
22:00.44DTRemenakbut I think it's worth it
22:01.10ts~DTRemenak++
22:01.38tsDTRemenak: It could save me a lot of keystrokes in chat
22:01.44DTRemenakI'm sure
22:01.55DTRemenakand it lets some folks chat, period...russians and poles and so on
22:01.58KTLdoes anyone know how i can hide the green fields around a map? i want it all to be water somehow
22:02.28DTRemenakKTL: afaiik using a skybox is the only way
22:02.46KTLhehe mmm i will try again to get them working
22:03.35tsDTRemenak: Still mathematicans have a slight problem..we need to implement LaTeX typesetting for them :)
22:03.46DTRemenakhah
22:04.14DTRemenakactually I'm not sure how we handle multistroke input either
22:04.18DTRemenakI think we probably choke on it
22:04.47tsmultistroke?
22:04.49DTRemenakâ in your layout for instance
22:05.32DTRemenakbut I have no real plans to fix that
22:05.32tsAh, that...germans don't use them. You have to ask the french or spanish people :)
22:05.32DTRemenakheh :)
22:05.40DTRemenakspanish layout works fine
22:05.46DTRemenakthe french use all the funny accents
22:05.51blast007can we limit callsigns to ascii?
22:05.53DTRemenaknot sure how that'll get handled
22:06.03DTRemenakblast007: sure
22:06.06DTRemenakthey may already be
22:06.11DTRemenaktry it
22:06.17blast007goes to build a client
22:06.20tsDTRemenak: I had spanish in school..they also have multistroke chars, altough not many
22:06.26DTRemenakI think we use isprint() on each character, so utf-8 will get trimmed out
22:06.50DTRemenakts: I don't remember any.  had plenty of spanish myself.
22:07.05DTRemenakportuguese does
22:07.07DTRemenakfrench does
22:08.34tsDTRemenak: Well, qué for example I thing. However Manu should know better. I forgot lots of stuff after school..I had english, french and spanish..nowadays I can only speak very very little.
22:08.44DTRemenakcould be that some layouts do
22:09.01DTRemenakyeah, spanish intl does here
22:09.17DTRemenakáéíóú
22:10.01DTRemenakand ý
22:10.21DTRemenakso yeah.  I dunno how to handle those
22:13.13CIA-33BZFlag: 03t-s * r17269 10/trunk/bzflag/src/bzflag/HUDuiTypeIn.cxx: whitespace
22:14.37Manuts: ?
22:15.08DTRemenakwe were just discussing multistroke input
22:15.22DTRemenakhe mentioned that germans don't have to worry about it, but you might ;)
22:15.39Manuin spanish: áéíóúÁÉÍÓÚñÑ
22:16.17Manuand in caps up
22:16.31DTRemenakhm, on most layouts i've seen at least ñ and Ñ have their own keys
22:16.47Manuyes, the is a ñ key
22:16.55DTRemenakbut yeah, the accented letters need two keystrokes
22:17.01DTRemenakand bzflag right now may not be able to handle that
22:17.10ManuDTRemenak: or three ;)
22:17.14Manufor example: ü
22:17.37Manushitft+"key + u
22:17.57DTRemenakstill just two as far as we{re concerned
22:18.13DTRemenakwe're, been too long since I used a spanish layout Ñ=
22:18.29DTRemenakheh, should proofread before I send :)
22:18.40Manutwo keys for accents in vowels
22:18.44Manuthree keys for "dieresis"
22:19.36Manulol
22:19.51BulletCatcherblast007: I would test with the bzexcess.com servers if you would unhostban me.
22:20.01whodaman-hah
22:20.15DTRemenaktwo events though, one for ¨ and one for the vowel
22:20.27Manuyes
22:20.44DTRemenakso we will have to figure out how to handle it
22:21.17DTRemenakbut it's better than before at least
22:21.36Manusee the key code, check if shift is pressed.. if it's some special key.. add it to a buffer and check the next key :)
22:22.31DTRemenakyeah, the "some special key" thing is harder than you'd think though
22:22.33Manuif the next key pressed is allowed to be mixed with the last key. .then handle it :)
22:22.49ManuDTRemenak: no harder than I think ;)
22:22.58DTRemenakheh Ñ=
22:23.00DTRemenakdoh
22:23.06DTRemenak:)
22:23.36DTRemenakswitches back to english-us layout
22:23.54Manuhas english and spanish all in one
22:25.32ManuDTRemenak: allowing to type 'ñ' would be a big step ahead in bzflag ;)
22:25.36Manufor the spanish users :)
22:25.41DTRemenakwould be?  is :)
22:25.44DTRemenakgo try it
22:25.49Manuaccents aren't so important for the mean of the words ;)
22:26.02tsI am not so happy with my keyboard layout by the way. In QWERTZ the signs used for coding are terribly placed.
22:26.08ManuDTRemenak: try to write 'year' in spanish without the 'ñ'
22:26.39ManuDTRemenak: with 'n' has a different mean ;)
22:26.55DTRemenaksometimes they are, ¿sí?
22:27.08DTRemenakyeah, true
22:27.30ManuDTRemenak: if you say'si' or 'sí' the other guy will understand you perfectly :)
22:27.40tsManu: Are the accents in spanish only for pronounciation?
22:27.51Manumainly yes
22:27.53DTRemenaktrue...context usually makes that obvious :)
22:28.15tsManu: I thought for n versus ñ applies the same
22:28.27Manuts: no
22:28.27DTRemenakheh, no
22:28.33tsMeanies!
22:28.39ts:)
22:29.03Manuts: the nletter 'ñ' was originally a short to write in latin sime words with two letters 'n'
22:29.13Manuanno -> año
22:29.27DTRemenakaño is the quickest example, it's true other places to though (cañon for instance)
22:29.28tsManu: I know how it is pronounced, I had spanish in school.
22:29.44tsManu: However..I thought it is only different pronouncing.
22:29.56Manuts, you can omit an accent, but you can't replace a 'ñ'
22:30.14Manucor example: 'cono' and 'coño' or 'ano' and 'año'
22:30.18Manudifferent means ;)
22:30.24tsManu: I can? Heck, I ommited ALL accents.
22:30.40Manubetter don't translate ;)
22:31.22ManuDTRemenak: also accented vowels in caps up need three keys :(
22:31.38DTRemenakshift doesn't count, it's not sent as a separate event
22:32.00ManuI know, it's just an state
22:32.19DTRemenakso á and Ü take the same number of events as far as we{re concerned (and are equally impossible at the moment)
22:32.26Manubut for example ü in caps up need to check shift for the two pulsations
22:32.30DTRemenakgah, keep forgetting about the stupid '
22:33.03Manu:)
22:33.08DTRemenaksure, you check it on the first event, check it again on the second event.  the problem is that we have no support for characters that take more than one event
22:33.16bryjenwraps DTRemenak up in a curly brace and pokes him with an apostrophe
22:33.16tsHow do japanese people type? I imagine that even more difficult.
22:33.29DTRemenakkanji does require 3 events for some characters
22:33.34Manuts: they use their fingers ;)
22:33.44Winnyehe
22:34.10DTRemenakI am SO not getting us into CJK though, unless someone complains really loudly or does the work for me :)
22:34.13tsManu: Heh, what I meant is 1 char = 1 word..so they have loooots of chars.
22:34.40Manuthey also have letters, not only figures ;)
22:35.00tsDunnu
22:35.06DTRemenakwe SHOULD be able to do hebrew and greek, but I haven't tested them
22:35.20DTRemenakcyrillic languages and single-event characters in latin languages work fine
22:35.22Manuand arabian...
22:35.41ManuDTRemenak: arabian and hebrew.. from right to left ;)
22:36.07DTRemenakheh.  we don't handle the RTL/LTR markers yet, good point
22:36.32DTRemenakdunno if they'd even draw correctly, ftgl might assume LTR too
22:37.22ManuDTRemenak: and chinese from up to down ;)
22:37.39Manubut they allow to from left to right
22:38.50DTRemenakbut we can do way more than we could
22:39.05DTRemenakand you got your ñ :)
22:39.36Manu:)
22:40.15ManuDTRemenak: also allowed in callsigns? ;)
22:40.29Manuhehe
22:40.32DTRemenakdunno
22:40.42Manuactually no I guess
22:40.48*** join/#bzflag KTL (n=KTL@213.219.151.14.adsl.dyn.edpnet.net)
22:40.53AAA_awrightI am having trouble using bz_getPlayerByCallsign in a bzfs plugin in MSVC, it compiles but doesn't link
22:40.53DTRemenakI think bzfs is still locale'd to ascii, and should reject unprintable callsigns
22:41.04*** join/#bzflag Qew (n=qew@82-69-126-225.dsl.in-addr.zen.co.uk)
22:41.20Manuunprintable or two bytes chars :)
22:41.40DTRemenakit's multibyte, but we process the callsign as single-byte
22:41.47DTRemenakso both bytes are >128 == unprintable
22:42.08AAA_awright..In ASCIIi
22:42.09DTRemenakthat's my theory anyway.  haven't tested :)
22:42.20DTRemenakAAA_awright: as I said, bzfs is stil locale'd to ascii
22:42.21Manuprobably you are right ;)
22:42.50AAA_awrightIf you accepted literal byte strings, ignoring characters, it wouldn't matter correct?
22:43.06tsDTRemenak: I tested with callsigns and UTF-8 is rejected in callsigns. Can't enter them through UI at least.
22:43.12DTRemenakAAA_awright: we scan callsigns for unprintable characters
22:43.15ManuAAA_awright: probably dangerous ;)
22:43.42DTRemenakwe could change it to scan over UTF8, sure.  but then people can use characters that admin's can't type
22:43.50DTRemenakwhich is probably not such a good idea
22:44.03ManuDTRemenak: admins don't need to type callsigns ;)
22:44.09bryjenslot numbers
22:44.16AAA_awrightTab completion
22:44.17Manuor auto completion
22:44.26DTRemenakit's useful to be able to type callsigns in any event
22:44.34bryjennot if you can't type the first letter
22:44.36ts*Some* different chars can also be pretty similar to other ones, so confusion comes up.
22:44.41AAA_awrightTake a hint from internationalized domain names
22:44.50ManuDTRemenak: that's why I added the slots usage ;) I'm admin
22:45.18AAA_awrightHave an ASCII form, which is optionally rendered with UTF
22:45.38tsManu: I changed back to using names for banning, after some people with low slot number went away and quickly later came a new guy in that was punished accidently. Partially I also read the wrong slot number.
22:45.48DTRemenaktoo much complexity imho.  and those names are a mess.
22:46.19Manuts: you still can use slots, or not?
22:46.21DTRemenakwe've been using ascii callsigns for years.  you can now chat in your native character set, that's a step forward.
22:46.38tsManu: Well, I can *but* slots are dangerous.
22:46.38ManuDTRemenak: I agree
22:47.02Manuts: nah :) admins can solve any issue
22:47.08Manutrue admins I guess :)
22:47.24DTRemenaka huge step forward actually.  multibyte input and rendering both.
22:48.04ManuDTRemenak: yes, now spanish speakers can swear properly ;)
22:48.22El_Gringo_Feothere must exist a string comparator which can match ascii and multibyte strings...
22:48.42ManuEl_Gringo_Feo: using tables :)
22:49.07El_Gringo_Feodoesn't think banning should be an issue
22:49.07DTRemenakdon't look at me to maintain it :)
22:49.39ManuDTRemenak:  is not so difficult ;)
22:49.54DTRemenakdon't care how difficult it is, just don't look at me ;)
22:49.58Manubut better is go step by step :)
22:50.26El_Gringo_FeoSpanish swearing is fun. There are only a few people at work who don't know Spanish :P
22:50.42Manuheh
22:51.05El_Gringo_Feoone of whom a few co-workers and I openly refer to as "Pendejo"
22:51.30ManuEl_Gringo_Feo: ven aqui pendejo ;)
22:51.42Manupendejo can be used in a friendly way :)
22:51.55El_Gringo_Feotrue
22:52.27AAA_awrightWhat about a solution that would search callsigns for the closest match? Tab completion, without the tab character. Equally dangerous, but at you never refer to anyone else then who you intended
22:52.28Manuthe same word applied to a man or a woman makes a difference
22:52.52El_Gringo_FeoI'd say the best one, though, was when I called him a gringo for being completely unable to pronounce "rapido" properly
22:53.07El_Gringo_Feo...I'm white as a dinner plate, and he is half-black
22:53.19ManuEl_Gringo_Feo: say: carro ;)
22:54.03El_Gringo_Feoyeah yeah :P
22:54.23El_Gringo_Feomy pronounciation is actually decent, but I've always had trouble with rr
22:54.37ManuEl_Gringo_Feo: say-> un tigre, dos tigres, tres tristes tigres trigaban trigo en un trigal ;)
22:54.44ManuL4m3r: ^----
22:54.53L4m3reek
22:55.28L4m3ryeah, that's worse than any English tongue-twister I've seen. yeesh.
22:56.01L4m3rthen again I also can't speak very fast to begin with
22:56.49ManuL4m3r: this tongue-twister is not very difficult.. there are worse ones ;)
22:57.51AAA_awrightDTRemenak: Any clue why bz_getPlayerByCallsign isn't working? bz_getPlayerByIndex is...
22:58.09AAA_awrightMSVC complains of an unresolved external
22:58.14DTRemenakhm, actually multi-stroke might work on SDL.  I see why it doesn't work on windows.
22:59.15DTRemenakAAA_awright: make sure you're linking to the correct lib.  they're defined the same way, they should work the same way.
22:59.19KTLbzfs "couldn't find PhysicsDriver: death" ??
22:59.54AAA_awrightIsn't it in the same library? So it is odd for one to fail and one to work...
23:00.14AAA_awrightOr... hm. Maybe that isn't the case at all.
23:00.50DTRemenakold versions will have byIndex but not byCallsign
23:01.06DTRemenakmake sure your library and header match and you should be good
23:01.18*** join/#bzflag Ride- (n=Ride-@24-197-250-9.dhcp.stpt.wi.charter.com)
23:07.30*** join/#bzflag bier (n=bier@p54A57EA7.dip.t-dialin.net)
23:10.26AAA_awrightDTRemenak: I can't figure out where I do that
23:14.05*** join/#bzflag Ride288 (n=Ride-@24-197-250-9.dhcp.stpt.wi.charter.com)
23:15.25CIA-33BZFlag: 03dtremenak * r17270 10/trunk/bzflag/src/common/CommandManager.cxx: make windows not choke on UTF8 sequences in the client config file. since we aren't actually doing anything with the characters this routine doesn't actually need to be multibyte-aware.
23:19.57KTLhttp://users.edpnet.be/dpf/bz/bzfi0091.png  <- and so it continues
23:27.00*** join/#bzflag blast007` (n=blast@24-180-190-113.dhcp.mdsn.wi.charter.com)
23:27.26*** mode/#bzflag [+v blast007`] by ChanServ
23:31.11TD-LinuxKTL.................... whoa!
23:31.16TD-Linuxis that a map you're working on?
23:31.19KTLmyes
23:31.29KTLbut i want all the playable parts high up in the sky
23:31.41KTLbefore i get there i really will need to invest my time in other things :D
23:32.35me1define map <insrtmp> enddef   group map shift 0 0 alot end?
23:33.21KTLis that a sentence?
23:33.30TD-Linuxit's how to make your map in the sky :P
23:33.35me1define map
23:33.39me1<insert map>
23:33.41me1enddef
23:33.42me1group
23:33.44KTLaaa
23:33.46me1group map
23:33.47KTLmmm...
23:33.50me1shift 0 0 alot
23:33.51TD-LinuxI just got an awesomely horrible idea
23:33.51me1end
23:33.52KTLand where do you get the map from?
23:33.59KTLtell about it.
23:34.01me1KTL, you :P
23:34.07TD-Linuxgoes to work
23:34.29me1goes to dinner
23:34.31KTLmakes me think of an xkcd joke about solving travelling salesman using ebay: complexity 1
23:39.54*** join/#bzflag bier|tp_ (n=bier@p54A57DB3.dip.t-dialin.net)
23:50.26KTLis it possible to include other bzw files ?
23:50.37KTLto modularise the input into files
23:50.48DTRemenakyes, the include keyword will do it.  it is undocumented and unsupported.
23:50.58KTLhehe thanks :)
23:51.13DTRemenaki.e. ymmv :)
23:51.40KTLjust perfect
23:53.54*** join/#bzflag blast007_ (n=blast@24-180-190-113.dhcp.mdsn.wi.charter.com)
23:54.02CIA-33BZFlag: 03dtremenak * r17271 10/branches/v2_99_shot_branch/ (1261 files in 112 dirs): sync with trunk r17268
23:56.40*** join/#bzflag bier_ (n=bier@p54A57DB3.dip.t-dialin.net)
23:59.18CIA-33BZFlag: 03dtremenak * r17272 10/trunk/bzflag/src/platform/ (WinDisplay.cxx WinDisplay.h): hopefully support multi-wchar UTF-16 input. also remove silly ((*WinDisplay)this)->... syntax, and mark translated and charCode as mutable.

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.