00:38.41 | *** join/#brlcad IriX64 (n=IriX64@bas3-sudbury98-1168056909.dsl.bell.ca) |
00:39.04 | IriX64 | that datarootdir thing this fixed it for me. |
00:39.13 | IriX64 | http://www.pastebin.ca/361223 |
00:52.58 | brlcad | there's already a fix in for that that doesn't outright override the datadir |
00:53.22 | IriX64 | hah new way to distribute binaries, compile with --prefix=/usr/local/whatever, make install and tar up /usr/local/whatever and tell em to unpack accordingly :) |
00:53.34 | IriX64 | where can i get this fix? |
00:54.42 | brlcad | cvs |
00:54.49 | IriX64 | ty |
00:55.39 | IriX64 | at the moment tho, my forge is melting bits :) |
01:00.01 | IriX64 | mine won't hurt anything tho will it? |
01:00.59 | IriX64 | not that ill distribute it or anything but i fear shooting myself in the foot here. |
01:02.24 | IriX64 | i figure as long as im on *this setup it should work forever :) |
05:10.54 | *** join/#brlcad digitalfredy_ (n=digitalf@200.71.62.161) |
07:20.45 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/autogen.sh: |
07:20.45 | CIA-5 | BRL-CAD: collapse the version checks into one routine. improve the version checks so |
07:20.45 | CIA-5 | BRL-CAD: that they work more reliably for various version string formats. this |
07:20.45 | CIA-5 | BRL-CAD: refactoring should fix a bug in the script where automake 1.10 was being |
07:20.45 | CIA-5 | BRL-CAD: incorrectly processed as 1.1.0 among other benefits and code simplification. |
07:27.56 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/autogen.sh: simplify some more, make sed a requirement and abort if it seems to be unavailable |
09:12.02 | *** part/#brlcad digitalfredy_ (n=digitalf@200.71.62.161) |
09:18.38 | *** join/#brlcad clock_ (i=clock@84-72-93-122.dclient.hispeed.ch) |
09:28.20 | *** join/#brlcad cad57 (n=54a11c8a@bz.bzflag.bz) |
09:28.57 | *** part/#brlcad cad57 (n=54a11c8a@bz.bzflag.bz) |
12:05.55 | *** join/#brlcad cad72 (n=530d8e42@bz.bzflag.bz) |
15:12.49 | *** join/#brlcad cad45 (n=4b3d409d@bz.bzflag.bz) |
16:40.59 | *** join/#brlcad clock_ (i=clock@84-72-61-113.dclient.hispeed.ch) |
16:48.38 | *** join/#brlcad louipc (n=louipc@bas8-toronto63-1096782827.dsl.bell.ca) |
19:04.11 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/include/sysv.h: add an initial interface file for libsysv. this compatibility interface shouldn't really be included throughout the package, but may ultimatley get added to common.h as part of configuration compatibility. |
19:04.44 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/include/Makefile.am: add the new sysv.h header to the noinst header list. |
19:06.02 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/libsysv/ (bsdselect.c strdup.c): deknrify the function |
19:06.34 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/libsysv/strsep.c: ws |
19:12.18 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/libsysv/ (bsdselect.c strchr.c strdup.c strtok.c): ws |
19:13.34 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/libsysv/basename.c: initial addition of basename() so we don't need to test for its availability |
19:14.01 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/libsysv/Makefile.am: add the new basename.c file, extracted from openbsd (bsd licensed) |
19:14.21 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/include/sysv.h: add basename() |
19:57.30 | *** join/#brlcad IriX64 (n=IriX64@bas3-sudbury98-1168056909.dsl.bell.ca) |
19:58.28 | IriX64 | cygwin+brlcad=windows cad system :) |
20:04.11 | IriX64 | tried to fix autogen.sh myself, mucked it up completly, started reporting version 0.0 of the tools. |
20:04.25 | louipc | whoah |
20:04.43 | IriX64 | thank God for backups. :) |
20:04.45 | louipc | yeah I just noticed that it detected automake incorectly |
20:05.01 | IriX64 | 1.1.0? |
20:05.02 | louipc | it thinks v 1.10 is 1.1.0 |
20:05.07 | IriX64 | ditto |
20:05.21 | IriX64 | good job. |
20:05.36 | louipc | nice |
20:06.25 | brlcad | would one of you test and make sure it works: http://brlcad.cvs.sourceforge.net/*checkout*/brlcad/brlcad/autogen.sh |
20:07.56 | Maloeran | Grr. "cvs [commit aborted]: there is no version here; do `cvs checkout' first" |
20:08.08 | Maloeran | And a cvs checkout of course doesn't solve anything |
20:08.17 | louipc | brlcad: looks good so far |
20:08.45 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/libbu/brlcad_path.c: progname is often not the same as argv[0], so modify bu_getprogname() accordingly and expose the bu_argv0() function to distinguish between the two. |
20:09.19 | Maloeran | Do you have 30 seconds to help with that cvs message, brlcad? |
20:09.31 | Maloeran | I sure often get obscure cvs errors when I move files around between desktop and latop |
20:10.11 | brlcad | hm |
20:10.46 | brlcad | they both the same OS and version of CVS? |
20:11.02 | Maloeran | Same OS, 32 and 64 bits though, perhaps not quite the same version of CVS |
20:11.23 | brlcad | sounds like you're commiting from a dir outside the cvs tree |
20:11.38 | brlcad | you only do a checkout once |
20:11.47 | Maloeran | I'm right in ~/rayforce |
20:11.57 | brlcad | and there's a ~/rayforce/CVS dir? |
20:12.08 | Maloeran | Yes |
20:12.21 | brlcad | cat ~/rayforce/CVS/Root |
20:12.34 | Maloeran | Looks fine |
20:12.55 | brlcad | can you pm it to me |
20:13.02 | brlcad | as well as CVS/Repository |
20:13.41 | Maloeran | I think it's throwing the error because of the other/ directory actually, there's no ./CVS there |
20:14.08 | brlcad | could be |
20:14.32 | brlcad | all of the D entries in ~/rayforce/CVS/Entries will get traversed |
20:14.56 | brlcad | you can avoid the problem by commiting files individually of course, or fix your Entries file |
20:15.08 | Maloeran | ./other is listed |
20:15.15 | brlcad | ahh |
20:15.18 | brlcad | then that's why |
20:15.35 | brlcad | you probably copied the file over from some other place that didn't have a CVS into a current checkout |
20:16.00 | Maloeran | Right, I think I'll manage to fix it from there. Thanks |
20:16.05 | brlcad | s/probably/almost guaranteed |
20:16.29 | brlcad | you should be able to mv other and cvs update -dP |
20:16.39 | Maloeran | There we go, it works. Thanks |
20:17.04 | brlcad | you should update -dP just to make sure you didn't hide the problem for down the road ;) |
20:17.17 | ``Erik | <-- points out that it's WAY easier to leave a checkout in place, have seperate checkouts on the lappie and workstation, and when you want to move between the two, make sure you're all committed up and just update on the other... |
20:17.43 | brlcad | yep |
20:17.53 | brlcad | frequent commits and all that jazz |
20:18.32 | ``Erik | to usurp the meme, commit early, commit often |
20:21.10 | ``Erik | the Entries file will hold what your side thinks is in the repo... you can always co to a new place to verify it |
20:23.45 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/libbu/brlcad_path.c: basename() is guaranteed by libsysv now, so no need to check for HAVE_BASENAME |
20:24.35 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/nirt/str_manip.c: remove the unnecessary definition of basename() .. libsysv provides a better version if it truely is not provided by the system. |
20:27.24 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/misc/enigma/.cvsignore: ignore more ac_config_subdir/configure crap dumped in here |
20:28.55 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/proc-db/.cvsignore: ignore newly added brep_cube if it exists |
20:29.16 | IriX64 | ok, I'm beat how do i get that autogen.sh? |
20:29.58 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/proc-db/Makefile.am: sort lines, vars first |
20:30.02 | brlcad | IriX64: er, with any browser? |
20:30.16 | IriX64 | can't open page. |
20:30.25 | brlcad | then you probably copied the url wrong |
20:30.34 | IriX64 | cut and paste brlcad. |
20:30.45 | IriX64 | will try again. |
20:30.48 | brlcad | then you probably cut and paste wrong :) |
20:31.13 | brlcad | else, just type it in: http://brlcad.cvs.sourceforge.net/*checkout*/brlcad/brlcad/autogen.sh |
20:31.39 | brlcad | it might even just download immediately |
20:32.44 | IriX64 | not letting me in and its exactly what you just typed. |
20:33.17 | IriX64 | http://brlcad.cvs.sourceforge.net/*checkout*/brlcad/brlcad/autogen.sh |
20:33.24 | IriX64 | right? |
20:33.38 | brlcad | IriX64: that looks right |
20:33.39 | IriX64 | im using ie. |
20:33.58 | ``Erik | <-- can't try ie, is totally windows free here at home |
20:34.16 | IriX64 | heh a dream come true eh? |
20:34.38 | IriX64 | maybe i should install firefox. |
20:34.48 | brlcad | IriX64: then go to http://sourceforge.net/projects/brlcad -> click on Code -> click on Browse CVS -> click on brlcad -> scroll down to autogen.sh and click on the name -> click on "download" |
20:35.18 | brlcad | i suspect it's just downloading it already to your desktop or something |
20:36.53 | brlcad | yep, I just tried it in IE |
20:37.10 | brlcad | it shows a "cannot display this page" and downloads the file |
20:37.20 | Maloeran | Neat.. |
20:37.44 | brlcad | which for me also includes a dialog prompt of where to save the file , but I suspect you turned that off at some point |
20:38.01 | brlcad | of you've got a bunch of dialogs underneath or something |
20:38.08 | IriX64 | anim, awf but no autogen.sh |
20:38.35 | IriX64 | downloads it to where? |
20:38.59 | brlcad | you're asking me? |
20:39.01 | IriX64 | thanks tho lets not get caught up in it. |
20:39.07 | brlcad | it's your machine, your set up ;) |
20:39.20 | IriX64 | she built with automake 1.9.6 im happy. |
20:39.24 | brlcad | autogen.sh is at the very bottom |
20:40.45 | brlcad | no matter, it worked for louipc yes? |
20:41.07 | IriX64 | thanks give me ten i gotta install automake 1-10 again takes me a few minutes. |
20:43.17 | IriX64 | bing bang you're a winner, my thanks. |
20:43.43 | IriX64 | cept libtool is coming up 1.5.23a is that proper? |
20:44.29 | louipc | brlcad: yep |
20:44.37 | IriX64 | thanks then. |
20:44.56 | IriX64 | lets see if it actually builds with those makefiles. |
20:46.16 | louipc | It is a weird URL with the asterisks |
20:51.58 | *** join/#brlcad docelic (n=docelic@212.15.183.80) |
20:52.19 | brlcad | asterisks are valid |
20:52.19 | CIA-5 | BRL-CAD: 03brlcad * 10brlcad/src/proc-db/brep_cube.cpp: footer and ws |
20:52.36 | brlcad | IriX64: what does glibtoolize --version report? |
20:53.15 | IriX64 | 1.5.23a |
20:53.41 | IriX64 | so you're right its proper. |
20:54.27 | IriX64 | configuring it now. |
20:54.32 | brlcad | even better, it correctly compares now even though there's an alpha in there |
20:54.39 | IriX64 | yeah |
20:54.46 | IriX64 | good work. |
20:54.55 | louipc | brlcad: you really spiffied up that script |
21:00.01 | Maloeran | Gez, 20 minutes to do a cvs commit of 50mb from a 10mbits line |
21:01.08 | ``Erik | 50mb? O.o |
21:01.12 | ``Erik | what is it? |
21:01.45 | IriX64 | the whole freaking project (ducking and running) |
21:01.46 | Maloeran | All code! :) No, just a now decent geometry group file format |
21:01.56 | Maloeran | And +8202-3147 lines of code |
21:02.38 | ``Erik | it's an ascii format? |
21:02.52 | Maloeran | No, the roter-lowe galleon model is heavy |
21:03.23 | ``Erik | if it's not ascii, you should've done cvs add -kb <file> so it doesn't attempt to search it for tags |
21:03.29 | ``Erik | and try to do diffs on it |
21:03.30 | ``Erik | heh |
21:03.37 | Maloeran | Oh! So that's what it was doing |
21:04.19 | Maloeran | And you are right, all the model files weren't uploaded as binary |
21:04.44 | ``Erik | plus changing every 0x0d to 0x0d0a |
21:05.05 | ``Erik | or trying to see if it needs to change 0x0d0a to 0x0d, or something |
21:05.10 | ``Erik | newline fixing |
21:05.13 | Maloeran | That will only happen if someone checkouts from a windows box |
21:05.45 | ``Erik | windows/dos will use 0x0d0a, *nix and osX will use 0x0d, and os[1-9] will use 0x0r |
21:06.01 | ``Erik | I don't recall what things like beos will do |
21:07.26 | ``Erik | (the history of all that predates unix, actually, and unix was deviant in faking the operations... newline and carrier feed map directly to old tty operations, just like old typewriters themselves) |
21:09.03 | Maloeran | :) I always found awkward the whole concept of "text mode" files, creates more problems than it solves |
21:10.41 | ``Erik | plenty of C functions to nice things up for you |
21:18.42 | brlcad | jeez.. a 12000 line commit differential.. if that's your idea of frequent small commits .. |
21:19.12 | brlcad | if there was a bug in there, you'd never find it via cvs history that's almost for certain |
21:28.04 | Maloeran | Indeed brlcad, but I'm quite alone on this anyway, and a CVS history is of no help for one's own code |
21:28.30 | brlcad | of no help for one's own code??? |
21:28.49 | brlcad | that's probably the dumbest thing I've heard you say in a long time :) |
21:28.51 | Maloeran | You wrote the code, you know it better than any history :) |
21:29.08 | *** join/#brlcad IriX64_ (n=who@bas3-sudbury98-1168056909.dsl.bell.ca) |
21:29.31 | brlcad | maybe you do today, what about in two weeks, or in two months or two years later.. when that bug is finally exposed |
21:29.50 | ``Erik | if you're working on tiny toy projects, sure... :D I hope most things I do are nontrivial and I forget minor details that machines find so important |
21:30.28 | brlcad | there is almost zero tracability on large commits, if the code is going to be more than a pet project .. that matters |
21:31.08 | IriX64_ | nmg-bot.exe, so far automake behaved :) |
21:31.31 | brlcad | I've frequently review cvs logs to see stuff that I did years ago, if anything it shows you the motivation for why something was changed, and if there's detail it can include references and other leads when there's a problem |
21:31.34 | ``Erik | I know with my personal projects, I rely heavily on tagging and good comments in the cvs history for preparing changelogs and making release statements |
21:31.39 | Maloeran | Right, years later, one might not even recognize the code as his own. If I knew a few extra CVS commands, I think I would recognize the value of a proper history |
21:32.02 | ``Erik | "log" "status" and "ann" are three I like |
21:32.06 | ``Erik | ann being annotate |
21:32.17 | brlcad | or just use annotate and you don't have to explain it :P |
21:32.36 | Maloeran | *nods* Three commands I have never learned or used |
21:32.46 | ``Erik | well, I'm thinking with my fingers, not my brain... :D I'd say 'co' instead of 'checkout' |
21:32.51 | brlcad | because all your code is "still in context" |
21:33.07 | ``Erik | hehehe |
21:33.10 | brlcad | just wait until you eventually do have a context switch and have to figure it all out again |
21:33.32 | Maloeran | I know that, brlcad. I really need to learn these commands to see the value of a good history |
21:33.36 | brlcad | that's where code size and maintenance adds up exponentially |
21:33.36 | ``Erik | mal, pick something funky in brlcad and muck around with those three commands for a bit... |
21:34.02 | ``Erik | then think about poor little me, trying to understand rayforce when all I see is "+5123 -128 Stuff." |
21:34.26 | Maloeran | Ahah @ Stuff :) |
21:36.14 | ``Erik | erik@vidar ~/devel$ cvs log 2>/dev/null | grep -i stuff | wc -l |
21:36.14 | ``Erik | <PROTECTED> |
21:37.44 | ``Erik | very little code is ever complete... no code is ever adequately tested... and no code is ever bug-free :D |
21:38.01 | brlcad | unless your code is but a handful of lines, there will almost *always* undoutedly be bugs no matter how much you test and feel comfortable with it |
21:38.20 | ``Erik | for initial development, being able to compile is a good tough goal for nightly cvs state |
21:38.21 | IriX64 | autogen.sh says you're a liar, so does 10 goto 10 ;) |
21:38.51 | brlcad | being able to compile after every cvs commit, regardless of testing and functionality is generally a good goal |
21:39.04 | brlcad | if you can get that much working right, you'll be better off than 90% of projects |
21:39.27 | brlcad | what bugs? |
21:39.39 | ``Erik | (also; testing fresh checkouts every once in a while is good practice) |
21:39.50 | Maloeran | *nods* All right, so noted, I'll do so. Basically, before now, I was only commiting when there was something new to show |
21:39.53 | ``Erik | that "sh autogen.sh" not matching that you fixed back then |
21:40.35 | IriX64 | speak too. |
21:40.35 | ``Erik | I hope autogen.sh is posix sh, not bash... |
21:40.54 | IriX64 | posix is just a name like bash. |
21:40.58 | ``Erik | cuz, uh, I use it with non-bash shells :) a modified bsh, ksh, zsh, ... |
21:41.06 | ``Erik | bash is a piece of software, posix is a set of standards... |
21:41.08 | IriX64 | cmon |
21:41.38 | IriX64 | cmd.exe is not posix but can be made to run autogen.sh |
21:41.39 | brlcad | ah, yes.. that was working, then broken then working re sh |
21:41.52 | brlcad | it wasn't a posix vs bash thing, just a basename/lookup issue |
21:42.31 | ``Erik | <-- tries to make his work with a strict posix sh for portability |
21:42.44 | IriX64 | sometimes it pays to learn the scripting languages, but there are so bloody many of them. |
21:42.58 | brlcad | can't imagine what it's break on even with strict posix at the moment |
21:43.05 | brlcad | i'd be interested in seeing a report if it fails |
21:43.17 | IriX64 | passed here |
21:43.20 | brlcad | only thing maybe might be some [] operations |
21:43.34 | brlcad | don't remember if posix allows -ne and -eq |
21:44.30 | ``Erik | since it's using /bin/sh, it'll be pretty damn close if not strictly posix on bsd, solaris, I THINK irix, ... only linux assumes /bin/sh is bash, iirc |
21:45.39 | IriX64 | ``erik bat file script i know, might take me a while but i could code an equivelent. |
21:46.38 | IriX64 | ``E too, sorry :) |
21:46.38 | ``Erik | equivelant to what? |
21:46.51 | IriX64 | to autogen but *you have to write the tools that it uses ok? ;) |
21:48.03 | IriX64 | bwthresh.exe sweet. |
21:48.04 | ``Erik | uh... if you're gonna use it on windows, you might as well install cygwin and have a working sh? |
21:48.16 | IriX64 | uhhh ;) |
21:48.24 | ``Erik | or use the msvc files and ignore the auto* stuff... |
21:48.47 | IriX64 | msvc is pricey. |
21:51.06 | ``Erik | didn't ms release a free version? lacking the fancy ide, I think, but should be able to build using the project build files, no? |
21:51.30 | IriX64 | unknown to me ``Erik. |
21:52.53 | ``Erik | http://msdn.microsoft.com/vstudio/express/visualc/download/default.aspx |
21:57.49 | IriX64 | http://pastebin.com/883914 |
21:59.05 | ``Erik | hm, I've seen similar screens before :D |
21:59.39 | ``Erik | of course, if I see that one, I say "oh, huh, I didn't break it" and try harder... |
22:00.16 | brlcad | heh |
22:06.28 | IriX64 | i think automake-1.10 works brlcad. |
22:06.59 | IriX64 | gonna install and runtest it. |
22:08.27 | IriX64 | whoa---/usr/cygcad/include/brlcad ... no such file or directory, what gives here? |
22:12.44 | IriX64 | http://pastebin.com/883927 |
22:18.25 | IriX64 | it looks like its not making the dir but i don't understand all this stuff in makefile, sorry :( but i can tell you this, with automake1.9.6 it works. |
22:44.40 | IriX64 | just rgressed to 1.9.6 and make install works again. |
22:45.30 | brlcad | IriX64: hard to say what the cause is, but on the surface it looks like an automake bug |
22:46.10 | IriX64 | too bad i don't speak perl, id try to help, but i leave that to those who are fluent. |
22:47.22 | IriX64 | ``Erik did I try hard enough? :) |
22:47.59 | IriX64 | btw is it safe to reconfigure and just do a make install thats how im trying this thing out. |
22:49.21 | brlcad | depends on what you're trying to do |
22:49.41 | brlcad | in general if you need to reconfigure, you probably should recompile |
22:49.48 | brlcad | but that's not always the case |
22:50.52 | IriX64 | same configure options, codes already compiled, should be able to get away with it no? |
22:51.46 | brlcad | then why are you rerunning configure? |
22:52.15 | IriX64 | did an autogen, wanted my makefiles fixed after i regressed to 1.9.6 of automake. |
22:52.33 | brlcad | for that, yeah, should be fine |
22:52.36 | IriX64 | ty |
23:26.07 | IriX64 | havoc renders, i'm all happy again :) |