| 00:37.25 | *** join/#bzflag gubernator (uid284174@gateway/web/irccloud.com/x-tfcctnthpsyuvxqp) |
| 00:42.34 | allejo | oh gd |
| 00:42.36 | allejo | god* |
| 00:42.57 | allejo | nothing has really changed since then |
| 01:19.11 | *** join/#bzflag infobot (ibot@rikers.org) |
| 01:19.11 | *** topic/#bzflag is https://BZFlag.org || http://ohloh.net/p/bzflag || https://bzflag.org/help || Channel Logs: http://infobot.rikers.org/%23bzflag/ || 2.4.12 released! https://forums.bzflag.org/viewtopic.php?f=62&t=19720 |
| 01:19.11 | *** mode/#bzflag [+v infobot] by ChanServ |
| 02:53.58 | *** join/#bzflag BZnotify (BZnotify@gateway/service/github.com/x-gturrpavxkveqozr) |
| 02:53.58 | BZnotify | [13download.bzflag.org] 15blast007 pushed 1 new commit to 06master: 02https://git.io/vA6Ig |
| 02:53.58 | BZnotify | 13download.bzflag.org/06master 14fadcd39 15Scott Wichser: Use an absolute path |
| 02:53.58 | *** part/#bzflag BZnotify (BZnotify@gateway/service/github.com/x-gturrpavxkveqozr) |
| 04:27.48 | *** join/#bzflag AAA_awright (~aaa@2607:5300:60:7d4::1) |
| 04:48.24 | BulletCatcher | src/bzfs/WorldWeapons.cxx:95: warning: declaration of 'worldEventManager' shadows a global declaration [-Wshadow] |
| 04:48.25 | BulletCatcher | Using a different name for the local variable would fix that. |
| 06:06.44 | *** join/#bzflag BZnotify (BZnotify@gateway/service/github.com/x-kqkcgoyasmkwshqe) |
| 06:06.44 | BZnotify | [13bzflag] 15allejo pushed 1 new commit to 062.4: 02https://git.io/vA6nt |
| 06:06.44 | BZnotify | 13bzflag/062.4 141d1e18d 15Vladimir Jimenez: Use global worldEventManager, not local version |
| 06:06.44 | *** part/#bzflag BZnotify (BZnotify@gateway/service/github.com/x-kqkcgoyasmkwshqe) |
| 06:07.20 | allejo | thanks BulletCatcher :) |
| 09:10.03 | *** join/#bzflag L4m3r (~l4m3r@bzflag/developer/L4m3r) |
| 09:10.03 | *** mode/#bzflag [+v L4m3r] by ChanServ |
| 11:34.13 | *** join/#bzflag BZnotify (BZnotify@gateway/service/github.com/x-apqqwzjsxhbajtgr) |
| 11:34.13 | BZnotify | [13bzflag] 15blast007 pushed 1 new commit to 062.4: 02https://git.io/vA6D0 |
| 11:34.13 | BZnotify | 13bzflag/062.4 14ff98516 15Scott Wichser: Remove old BitmapFont renderer. |
| 11:34.13 | *** part/#bzflag BZnotify (BZnotify@gateway/service/github.com/x-apqqwzjsxhbajtgr) |
| 11:55.02 | blast007 | okay.. now gcc 7.1 has another set of warnings :) |
| 11:55.06 | blast007 | http://paste.debian.net/hidden/398312e1 |
| 11:55.28 | blast007 | for the first one I'd probably just remove the motto from the author part of the header, since that's kinda silly anyway |
| 12:11.40 | blast007 | actually just not showing it as part of the message sent to players works. for the appversion I just subtracted 14 from the length so it fits. :) |
| 14:14.01 | blast007 | gcc 7.1 also triggers warnings about switch statements where a case will "fall through" to another case, such as in src/bzflag/clientConfig.cxx where we use the switch statement to run through (potentially) multiple layers of upgrades. |
| 14:15.04 | blast007 | there's a newish [[fallthrough]]; attribute to indicate that is expected behavior |
| 14:16.49 | blast007 | hmm, seems to be a C++17 attribute |
| 14:17.55 | blast007 | so that'd require Visual Studio 2017 |
| 15:36.44 | *** join/#bzflag aazz (~mathiasla@90.115.122.246) |
| 15:38.15 | *** part/#bzflag aazz (~mathiasla@90.115.122.246) |
| 16:19.42 | JeffM | upgrades to 2017 are trivial |
| 16:21.13 | blast007 | yeah, I'm prepping a new VM |
| 16:21.46 | blast007 | might still have to do something for older gcc/clang versions |
| 16:22.08 | JeffM | want me to see if I can change the code to not use sprintf? |
| 16:22.48 | blast007 | is that referring to the motto stuff above? |
| 16:22.57 | JeffM | yeah |
| 16:23.17 | blast007 | I don't think that's really the issue |
| 16:23.20 | JeffM | I can probalby change client config too |
| 16:23.29 | JeffM | ok |
| 16:23.45 | blast007 | clientConfig could just use if statements instead of a fallthrough switch |
| 16:23.55 | JeffM | yeah |
| 16:24.08 | blast007 | there was another spot where a fallthrough happened |
| 16:24.10 | JeffM | fallthrough switches are kinda bad programing practice. |
| 16:24.51 | blast007 | yeah, definitely makes it a bit more confusing to read :) |
| 16:25.08 | JeffM | most modern languages don't allow them |
| 16:25.10 | blast007 | especially when it only *sometimes* does it |
| 16:25.28 | JeffM | yeah the main problem is allowing it makes it easy to hide bugs when it's not intended. |
| 16:25.49 | JeffM | and the valid uses for it are rare and have other methods of doing it. |
| 16:25.55 | blast007 | https://github.com/BZFlag-Dev/bzflag/blob/2.4/src/bzflag/SegmentedShotStrategy.cxx#L483-L504 that's another spot we do it |
| 16:26.15 | blast007 | so again, probably better to use a set of if/else statements |
| 16:26.31 | JeffM | yeah |
| 16:26.47 | blast007 | I'll refactor those when I get home |
| 16:27.26 | blast007 | for the sprintf stuff I just removed the motto from the output for the first one, and reduced the size of the appVersion for the second |
| 16:28.10 | blast007 | I don't really see the motto of the one who saved a replay as hugely useful information :) |
| 16:28.23 | blast007 | but I'll leave it in the replay format for now |
| 16:28.36 | JeffM | well it used to be email address ;) |
| 16:28.53 | blast007 | heh... forgot about that |
| 16:36.34 | JeffM | christ... when your IF test is that long, put it in a friggin function. https://www.irccloud.com/pastebin/vRuEGmT5/ |
| 16:39.42 | blast007 | we probably don't even need the first 3 client config upgrade steps anymore |
| 16:40.09 | blast007 | if (configVersion < 3) resetConfig(); :P |
| 16:40.58 | blast007 | or maybe 2 |
| 16:42.28 | JeffM | does anyone really feel that if (){ is more readable than |
| 16:42.28 | JeffM | if () |
| 16:42.28 | JeffM | { |
| 16:42.48 | JeffM | you didn't remove bitmap font from the vc project |
| 16:43.05 | blast007 | yeah, I was just doing it this morning quick from my linux system |
| 16:43.11 | blast007 | windows and mac will need to get updated |
| 16:43.25 | JeffM | you can edit the XML on linux ;) |
| 16:43.36 | JeffM | I'd get told I forgot it in the makefile |
| 16:44.29 | blast007 | I seem to recall issues with line ending when I edit it on linux though |
| 16:44.36 | blast007 | endings* |
| 16:44.53 | JeffM | it shoudl be flagged the same as the source files |
| 16:44.57 | JeffM | and go checkout native. |
| 16:45.13 | JeffM | the SLN has that issue but the project should not |
| 16:45.20 | JeffM | just use an editor that doesn't screw it up |
| 16:46.40 | blast007 | it adds a Windows newline to the end of the file |
| 16:46.52 | blast007 | even though it's using linux line endings for the file |
| 16:47.07 | JeffM | what crappy editor are you using? notepad++ would not do that |
| 16:47.11 | blast007 | vim |
| 16:47.19 | JeffM | then vim sucks ) |
| 16:47.40 | JeffM | an editor should be able to look at the file to determine newlines, if it has at least one newline in it |
| 16:50.00 | *** join/#bzflag BZnotify (BZnotify@gateway/service/github.com/x-ouqjwkropzlvcniu) |
| 16:50.00 | BZnotify | [13bzflag] 15JeffM2501 opened pull request #101: fix fall though switches, remove bitmap font. (062.4...06switch_fixes) 02https://git.io/vAi2X |
| 16:50.01 | *** part/#bzflag BZnotify (BZnotify@gateway/service/github.com/x-ouqjwkropzlvcniu) |
| 16:50.10 | JeffM | but there is a PR that fixes the switches and the project files. |
| 16:51.07 | blast007 | nano does the same thing |
| 16:51.20 | JeffM | linux needs a good editor then |
| 16:52.58 | blast007 | most text editors will add a trailing slash to your text files by default |
| 16:54.26 | JeffM | I'm surprised there is no np++ port to linux. |
| 16:54.37 | JeffM | it's based off some very nice open source components |
| 16:57.06 | allejo | or you know, you can use git to handle line endings and use whatever editor you want |
| 16:57.28 | JeffM | svn did that with autoprops |
| 16:57.56 | JeffM | all the VC projects should be fine to work the same as the source files |
| 16:58.04 | JeffM | it's all just XML now, no odd binary stuff. |
| 16:59.22 | blast007 | allejo: it would still add a line ending |
| 17:00.00 | allejo | the editor or git? |
| 17:00.04 | blast007 | editor |
| 17:00.49 | allejo | are we talking about LF vs CRLF or a new line at the end? |
| 17:02.42 | allejo | if the former, git can take care of it with pull/push, e.g. https://help.github.com/articles/dealing-with-line-endings/ |
| 17:11.27 | blast007 | the files do not have a trailing line ending at all |
| 17:27.32 | blast007 | should that be casting the return to a bool instead of an int? https://github.com/BZFlag-Dev/bzflag/blob/2.4/src/bzfs/bzfsAPI.cxx#L1725-L1732 |
| 17:35.15 | *** join/#bzflag youtah (~youtah@67.139.187.132) |
| 17:35.49 | *** part/#bzflag youtah (~youtah@67.139.187.132) |
| 17:39.25 | JeffM | it should return bz_fireServerShot > 0 |
| 18:54.06 | blast007 | regarding your earlier question about if () { vs if ()\n{, I'm not sure which I prefer. The former is what I'm used to. I do dislike having 'else' on the same line as the closing bracket for the matching 'if', because I prefer to document above the if and else statements (which gets weird if else isn't on it's own line), I prefer if and else lining up horizontally, and then it's also consistent |
| 18:54.12 | blast007 | alignment wether or not curly brackets are used. |
| 18:57.31 | blast007 | is there a good standard that would be easy to enforce in editors across our platforms? and ideally something that would also be machine reviewable so that a webhook could run a check? |
| 19:10.20 | JeffM | allman it's a BSD style |
| 19:10.42 | JeffM | clang-format should be able to do it |
| 19:11.10 | JeffM | if() on it's own line has the advantage of seperating the conditon from the code too, so you can // out the IF and have the code still run |
| 19:11.28 | blast007 | ah |
| 19:35.38 | tupone | DEVINFO rules |
| 19:37.06 | JeffM | good thing it's just a file that can be changed ;) |
| 19:37.38 | blast007 | yup, that was the point of the discussion :) |
| 19:38.06 | JeffM | I'll make a branch in my repo with what I think would look good, for review |
| 19:39.46 | blast007 | I'm more concerned with something that is easier to keep consistent (so, editor support, automatic checks possibly via TravisCI) |
| 19:39.59 | JeffM | sure |
| 19:44.22 | JeffM | blast007: do you still want to notify bzflag-users@lists.SourceForge.net of a build? :) |
| 19:44.40 | JeffM | oh god, and freshmeat |
| 19:44.46 | blast007 | is that something that is happening somewhere? |
| 19:44.56 | JeffM | ass end of devinfo |
| 19:45.00 | blast007 | heh |
| 19:48.37 | JeffM | you are also supposed to notifiy some guy who reviewed bzflag once. |
| 19:54.59 | blast007 | and add the version to the wiki :) |
| 19:55.30 | JeffM | isn't that so much prettier :) https://usercontent.irccloud-cdn.com/file/xA9KfTlW/image.png |
| 19:55.35 | blast007 | "We will likely move to OGG someday" *10 years later* |
| 19:56.00 | JeffM | yeah after the MP3 patent has expired. |
| 19:56.32 | blast007 | I think opus is probably more useful now anyway |
| 19:56.49 | blast007 | opus is crazy good |
| 19:57.25 | JeffM | really it's just what the libs support |
| 19:57.38 | JeffM | I don't think you should waste time doing your own audio |
| 19:58.08 | blast007 | there's already opus libs in debian stable |
| 19:58.31 | JeffM | cool so 4% of the players can use it ;) |
| 19:58.39 | JeffM | those 2 guys |
| 19:59.10 | blast007 | well, what I mean is that for Windows (and probably soon mac) we build dep packs |
| 19:59.18 | JeffM | sure |
| 19:59.31 | JeffM | does it work with openAL? |
| 19:59.42 | JeffM | I don't think you should do any of your own sound ;) |
| 20:00.53 | JeffM | ooooo git does NOT like the size of this commit :) |
| 20:02.12 | JeffM | https://github.com/JeffM2501/bzflag/tree/beautification |
| 20:10.25 | *** join/#bzflag nadir (uid134094@gateway/web/irccloud.com/x-sekokehqvvxltaji) |
| 20:53.47 | allejo | oh god. if we're talking style guide |
| 20:53.50 | allejo | PLEASE. PLEASE |
| 20:53.55 | allejo | stop the mixing of tabs and spaces |
| 20:55.00 | blast007 | allejo: I'm assuming the suggestion is to go to all spaces? |
| 20:55.27 | blast007 | s/the/your/ |
| 20:55.38 | allejo | that's what I prefer because we have finer control of spacing things versus tabs. but i'd be fine with just tabs |
| 20:55.53 | allejo | but no more 8 spaces = 1 tab but everything else is indented with spaces |
| 20:57.34 | allejo | and me personally, I like having all braces on a new line no matter what. instead of differences between functions and ifs and other stuff |
| 20:59.19 | JeffM | The tool I ran fixed it all to make it consistent |
| 20:59.34 | allejo | my problem with tabs is depending on your IDE, you'll get stuff like this: https://imgur.com/a/SjKrZ |
| 20:59.58 | allejo | whereas with spaces, it'll always look exactly across anything |
| 21:00.12 | JeffM | Unless you scew that up |
| 21:00.28 | JeffM | I find spaces much easier to mess up and get out of sync |
| 21:01.17 | allejo | in which cases? |
| 21:01.27 | L4m3r | I like tabs because everyone can have their preferred tab width set in their IDE |
| 21:01.39 | L4m3r | but I agree, mixing them is a mess |
| 21:02.03 | allejo | if we use tabs for indentation only and spaces for aligning things to look pretty like enums, I'd be fine with that |
| 21:02.31 | blast007 | why not tabs for aligning stuff too? |
| 21:03.05 | allejo | see my image |
| 21:03.08 | JeffM | If the ide is compliant nobody has problems, so we are talking about cases where ide are not compliant. With spaces you get a bunch of extra characters when stuff gets out of sync. With tabs you just have to fix the editor |
| 21:03.19 | allejo | it's not consistent across all IDEs/editors |
| 21:03.52 | JeffM | Spaces burn the data into the file, tabs are there for the express purpose of defining an indent |
| 21:04.04 | blast007 | k |
| 21:04.21 | allejo | but we don't live in an age where extra data in files is affective |
| 21:04.33 | JeffM | Itâs harder to clean up |
| 21:04.59 | JeffM | We should also live in an agent where people can have good editors |
| 21:05.06 | JeffM | Not vim |
| 21:05.55 | JeffM | If the editor isnât going to read the editor config, then itâs going to screw up anyway, tabs are easier to fix |
| 21:06.17 | JeffM | Just make all the source bitmapa |
| 21:06.27 | allejo | as long as aligning things aren't done with tabs |
| 21:06.31 | JeffM | Then all the formatting is burned in ;) |
| 21:06.35 | allejo | lol |
| 21:06.52 | L4m3r | alignment is a tricky one |
| 21:07.00 | JeffM | Iâm less |
| 21:07.21 | JeffM | Concerned with what is used for alignment of comments than I am for indents |
| 21:08.35 | allejo | same, but might as well discuss that if that's the current topic :) |
| 21:12.10 | JeffM | The 2 space indent is bs |
| 21:13.10 | allejo | agreed. 4 is better |
| 21:35.10 | macsforme | if (case) { seems much more conventional than if (case)\n{... I would personally rather keep the brace on the same line... but then, it doesn't make sense to do it one way for classes and a different way for if statements, loops, etc. |
| 21:36.16 | macsforme | I agree to stop the mixing of spaces and tabs... personally I would prefer to indent with spaces... what is wrong with a two-space indent? |
| 21:36.34 | JeffM | it's very small |
| 21:36.41 | JeffM | I have a very large display ;) |
| 21:36.54 | macsforme | we tell people to stay within the 80 column... with a four-space indent, you run out of space (no pun intended) quickly |
| 21:37.04 | JeffM | it's quite wide and bzflag code is always scrunched up in the corner, like it thinks i'm using a terminal |
| 21:37.21 | JeffM | well it's 2018, people have larger monitors now, not terminal emulators |
| 21:37.43 | JeffM | with spaces I have 0 adujustment options |
| 21:37.52 | JeffM | and am forced to work in a small space |
| 21:40.36 | macsforme | you can't line up arguments on different lines using tabs only; e.g.: |
| 21:40.38 | macsforme | void function (int arg1, int arg2, |
| 21:40.38 | macsforme | <PROTECTED> |
| 21:40.52 | JeffM | don't make your functions that large ;) |
| 21:40.57 | JeffM | use all your screen |
| 21:42.52 | JeffM | I mean is bzflag really going to be programed on an 80 column punch card? |
| 21:43.17 | macsforme | on my 1080p screen in Xcode, with the sidebar and two source files open side-by-side, I can display each page 103 characters wide |
| 21:43.44 | JeffM | no, you must put it on a punch chard! |
| 21:43.49 | JeffM | that is the unix way! |
| 21:44.11 | JeffM | like daylight savings time |
| 21:44.29 | allejo | what's wrong with a new 120 char limit *cough* |
| 21:44.39 | JeffM | 'murica, DST, inches, and 80 cols! |
| 21:45.27 | JeffM | nothing is wrong with it |
| 21:45.45 | macsforme | furthermore, some things are easier to read on multiple lines than on one long line... with spaces, you are able to line everything up the way you like, and it shows up consistently for everyone |
| 21:46.34 | JeffM | for that case, it should be put into a well named function |
| 21:55.48 | macsforme | this example isn't from our code, but how would you format something like this? https://pastebin.com/SnBgEHJ1 |
| 21:56.26 | JeffM | I'd write a friggin matrix object and pass that in |
| 21:56.43 | JeffM | even formated right that is hard to read |
| 21:56.58 | macsforme | that's what it is :P |
| 21:57.10 | JeffM | you need sub methods to peel off stuff |
| 21:57.58 | JeffM | fill out the object on multiple lines then do the operator on the objects. |
| 21:58.48 | JeffM | the very fact that you made a method that takes 12 arguments is where the problem started..... |
| 21:59.33 | JeffM | methods to peel off vectors from the matrix, then pass them in as vector objects, etc... |
| 22:07.30 | macsforme | even Matrix(Vector(), Vector(), Vector(), Vector()) will have the same challenges, so I'm not sure what difference that makes... but regardless, I'm not comfortable with this kind of limitation on how statements can be written, for the sake of configurable tab width |
| 22:08.20 | macsforme | although I agree with writing readable code, and rearranging statements where necessary to enhance that (to an extent) |
| 22:09.13 | JeffM | at least upgrade to VC2017 and move the VC Projects to a place where the editor config works ;) |
| 22:12.23 | blast007 | yeah, I'll work on that. getting a VM updated yet. |
| 22:12.41 | JeffM | you don't have to update the deps if you don't want to |
| 22:12.45 | blast007 | in situations like that, what if it was tab indented up to the matching start of the first line and tab aligned afterwards? |
| 22:12.50 | JeffM | they work fine in 2017 |
| 22:13.01 | blast007 | sorry, spaced asligned aftewards |
| 22:13.04 | blast007 | space* |
| 22:13.08 | JeffM | then you end up with mixed |
| 22:13.13 | JeffM | mixed for any reason is crappy |
| 22:13.23 | JeffM | I'd rather deal with spaces than mixed |
| 22:13.46 | JeffM | gimme \n{ and I'll deal with spaces ;) |
| 22:14.24 | JeffM | then at least everything is consistent |
| 22:14.27 | blast007 | what if we just don't have a line length limit? |
| 22:14.39 | JeffM | that'd be cool too |
| 22:14.48 | JeffM | leave it up to the developer? |
| 22:14.50 | blast007 | cuz word wrap is a thing :) |
| 22:14.57 | JeffM | that generaly looks like butt |
| 22:15.08 | JeffM | if it's going to wrap then allignment is nice. |
| 22:15.26 | JeffM | if there is code like what macsforme posted, then allignment does look best. |
| 22:15.58 | JeffM | but I'd not force people to do multi-line if they just happen to be over an arbitary limit, it should make sense to multi-line it. |
| 22:16.31 | macsforme | agreed |
| 22:17.09 | JeffM | and 4 spaces :) |
| 22:17.11 | blast007 | at 1920 wide terminal with default font size (monospace 12), I can get 212 characters on a single line |
| 22:17.13 | JeffM | or 3 |
| 22:17.34 | JeffM | 2 is just too small |
| 22:20.23 | JeffM | I can get 208 in my default visual studio setup on a HD display, the solution explorer does take up some room. |
| 22:20.44 | macsforme | I can live with a 4-space indent |
| 22:20.46 | JeffM | on my extra wide one at work I don't use the width for code, I have additonal sidebars open on the right, profiler, properties, etc.. |
| 22:21.17 | blast007 | 219 in Qt Creator for me |
| 22:22.51 | JeffM | you mean you don't program in Jurassic park unix? https://www.youtube.com/watch?v=dFUlAQZB9Ng |
| 22:23.57 | blast007 | no, I removed the remnants of Irix support recently ;) |
| 22:26.29 | blast007 | (or at least, the obvious stuff - I'm sure there's a bunch of other old code that was written in a way that favors super old compilers) |
| 22:29.34 | blast007 | JeffM: I think OpenAL Soft expects you to decode the audio files yourself |
| 22:29.58 | JeffM | ahh then you could do that in a background thread with whatever lib you wanted. |
| 22:30.04 | JeffM | good thing ram is cheap :) |
| 22:30.38 | JeffM | and hopefully the decompression of the stream doesn't have a lot of latency |
| 22:30.40 | blast007 | their example uses SDL_sound |
| 22:30.53 | blast007 | https://github.com/kcat/openal-soft/blob/master/examples/alplay.c |
| 22:31.12 | JeffM | yeah but do they care about the time betwen when they start playing and when enough data is sent to the audio provider? |
| 22:31.37 | JeffM | the main problem I've seen with streaming formats is that they take a bit to get the decompression setup |
| 22:31.49 | JeffM | where with a game, you are going to want to play a sound RIGHT NOW |
| 22:32.08 | JeffM | so you need a fast decompressor, or you need to cache the uncompressed sound. |
| 22:32.13 | JeffM | just things to watch out for. |
| 22:32.48 | JeffM | nobody cares if it takes 500ms for your hipster podcast to get going, but if it's a shot sound.... |
