| 01:40.04 | starseeker | reaches the conclusion he half expected - going to have to create Find*.cmake files on steroids to deal with the Apple X11/OGL/AGL duality situation |
| 01:40.31 | starseeker | minimally for Tcl/Tk, probably X11 and GL too... |
| 01:40.40 | starseeker | actually, certainly for GL |
| 01:41.16 | *** join/#brlcad IriX64 (~MarioDUli@bas2-sudbury98-1177680812.dsl.bell.ca) |
| 01:41.54 | starseeker | wonders if he could actually get the CMake devs to incorporate enhanced versions, or if they need to be BRL-CAD specific... hmm... |
| 05:12.27 | *** join/#brlcad IriX64 (~MarioDUli@bas2-sudbury98-1096600947.dsl.bell.ca) |
| 05:31.35 | IriX64 | mc |
| 06:52.41 | *** join/#brlcad merzo (~merzo@smartbussiness.mobicom.net.ua) |
| 07:11.37 | *** join/#brlcad Stattrav (~Stattrav@117.192.155.229) |
| 08:22.45 | kanzure | which version of opennurbs is included in brlcad? |
| 08:23.08 | kanzure | i just compiled the 2010-04-09 version of opennurbs and i'm trying to link against something that used the brlcad opennurbs headers (i know, this is really dumb) |
| 09:37.45 | *** join/#brlcad Stattrav (~Stattrav@117.192.150.156) |
| 10:45.09 | louipc | kanzure: check out the svn log |
| 10:45.30 | louipc | it's 201004095 (201004099 if _DEBUG defined) |
| 10:53.55 | kanzure | my bad :) |
| 10:55.37 | d-lo | mernin |
| 11:22.30 | *** join/#brlcad Zaebos (~irc@pd95b7f5e.dip0.t-ipconnect.de) |
| 12:36.49 | *** join/#brlcad d_rossberg (~rossberg@BZ.BZFLAG.BZ) |
| 12:37.56 | d_rossberg | starseeker: this is what i get with cmake on linux: http://www.pastebin.org/471285 |
| 12:38.17 | d-lo | brlcad: Looking at pkg_permserver. What is the reason for passing the the port to be used to listen on as a 'const char*' as apposed to an 'unsigned short' ? |
| 12:38.34 | starseeker | d_rossberg: where is your tcl/tk installed? |
| 12:38.51 | starseeker | or rather, do you have a system install? |
| 12:39.20 | starseeker | if not, this looks to be expected - I don't have a build of the local tcl/tk going yet |
| 12:40.30 | starseeker | my options for local tcl/tk building are either to use ExternalProject and try to get it working with the native build system, or write a CMake system for tcl/tk |
| 12:41.02 | starseeker | both are potentially thorny, so I've been using system Tcl/Tk up until now |
| 12:41.28 | starseeker | I've just run into my own issues with doing so though, so it's time to figure something out :-/ |
| 12:41.34 | d_rossberg | starseeker: i think tcl will be build from the brl-cad source tree |
| 12:41.56 | d_rossberg | i you like i could install the tcl devel packages |
| 12:42.14 | d_rossberg | ... and see what happens then |
| 12:42.21 | starseeker | d_rossberg: it's not urgent - in fact, maybe better not to |
| 12:42.54 | starseeker | then you can test once I get local tcl/tk build going |
| 12:43.16 | d-lo | brlcad: especially since _pkg_permserver_impl runs the 'char*' through atoi then casts it to an 'unsigned short' |
| 12:43.57 | d_rossberg | ok, i'll be a tester for "no tcl devel packages installed" |
| 12:44.02 | starseeker | d_rossberg: thanks! |
| 12:44.05 | d-lo | brlcad: sorry, ment char* -> atoi -> unsigned short -> htons |
| 12:44.42 | starseeker | d_rossberg: My hunch for Windows is that things will fail early in the configure stage... |
| 12:45.40 | CIA-42 | BRL-CAD: 03davidloman * r40097 10/rt^3/trunk/src/libPkgCpp/ (PkgClient.cxx PkgClient.h PkgServer.cxx PkgServer.h): Start wiring up libpkgcpp functionality to libpkg |
| 12:47.30 | d_rossberg | starseeker: i'm still in the process of checking out the cmake branch on windows, it stops every few files (probable an issue with the proxy) |
| 12:48.00 | starseeker | ah, fun |
| 12:49.37 | d_rossberg | i probable should have copied the trunk checkout and switched it to the cmake branch ... now it's to late |
| 12:50.56 | ``Erik | d-lo: getservbyname() can use the strings in /etc/services instead of declaring port yourself |
| 12:51.28 | ``Erik | :564 |
| 12:51.36 | d-lo | ?? not exactly what I was asking, but thanks for that info! |
| 12:51.58 | ``Erik | pkg_permserv() can take either the number of the name and figure out what to do |
| 12:52.22 | ``Erik | pkg_permserv(iface, "telnet", ...) |
| 12:53.00 | d-lo | ah, okay. so could pass "telnet" as well as "45199" |
| 12:53.02 | d-lo | ? |
| 12:53.13 | ``Erik | yeah |
| 12:53.26 | d-lo | ah. C-fu whoops my butt again :) |
| 12:53.34 | ``Erik | line 564 |
| 13:03.05 | *** join/#brlcad Stattrav (~Stattrav@117.192.150.82) |
| 13:09.25 | d-lo | ``Erik: is there a way to obtain a complete list of brlcad libraries currently installed? |
| 13:09.40 | d-lo | brlcad-config --libs only gives a few, not all. |
| 13:23.30 | CIA-42 | BRL-CAD: 03davidloman * r40098 10/rt^3/trunk/cmake/FindBRLCAD.cmake: FindBRLCAD.cmake: Cleaned up logic flow. Added a BRLCAD_LIB_DIR for aid in finding libraries not reported by brlcad-config. |
| 13:26.57 | ``Erik | um |
| 13:26.59 | ``Erik | ls? |
| 13:27.01 | ``Erik | :D |
| 13:28.35 | d-lo | :P |
| 13:28.50 | d-lo | thats what I am doing until i find a better way :) |
| 13:28.51 | ``Erik | for the fbsd port manifest, I update it by hand (sorta, with find and sed wired in) |
| 13:29.22 | ``Erik | I've been thinking about having a make target to automatically generate the manifest and manpage sets |
| 13:29.50 | d-lo | that would be good :) |
| 13:30.05 | d-lo | brlcad-config is a good idea, but it doesnt report all the libs :/ |
| 13:37.22 | CIA-42 | BRL-CAD: 03d_rossberg * r40099 10/brlcad/trunk/src/ (libged/bot.c librt/primitives/bot/bot.c): |
| 13:37.22 | CIA-42 | BRL-CAD: some compilers (as MSVC 2008) do not like these C99 idioms |
| 13:37.22 | CIA-42 | BRL-CAD: put variable declarations to the begin of the section |
| 13:38.50 | CIA-42 | BRL-CAD: 03d_rossberg * r40100 10/brlcad/trunk/src/libged/CMakeLists.txt: synced with Makefile.am |
| 13:42.26 | CIA-42 | BRL-CAD: 03erikgreenwald * r40101 10/brlcad/branches/bottie/ (82 files in 34 dirs): MFC |
| 13:42.28 | d-lo | does 48 libs sound about right?> |
| 13:44.48 | ``Erik | hm, I see 48 .a files with tcl and png, but no libz... somewhere in the neighborhood, but "it depends" |
| 13:45.27 | d-lo | I'm going off of .so file sinstalled after an --enable-all |
| 13:45.34 | d-lo | so as long as I am close. |
| 13:46.40 | *** join/#brlcad Stattrav (~Stattrav@117.192.130.128) |
| 13:47.10 | d-lo | whats libutahrle ? |
| 13:47.29 | ``Erik | the utah compression crud for images, iirc |
| 13:47.54 | ``Erik | src/other/libutahrle :D |
| 13:48.15 | ``Erik | thinks it's just used for the *-rle and rle-* pix utils |
| 13:51.44 | d-lo | kk thanks. |
| 13:52.54 | CIA-42 | BRL-CAD: 03bob1961 * r40102 10/brlcad/trunk/src/tclscripts/archer/ArcherCore.tcl: Membership operator was missing for leaves. |
| 13:55.52 | CIA-42 | BRL-CAD: 03erikgreenwald * r40103 10/brlcad/branches/bottie/src/ (4 files in 2 dirs): move tie engine into librt |
| 14:03.46 | *** join/#brlcad Nohla (~Nohla@201.255.222.78) |
| 14:09.41 | CIA-42 | BRL-CAD: 03bob1961 * r40104 10/brlcad/trunk/src/tclscripts/archer/CombEditFrame.tcl: Activate the old membership view and tie it to the apply/reset buttons. |
| 14:19.29 | d-lo | question: if we install headers into 2 places: include/ include/brlcad, we can easily find the include/brlcad/ dir by searching for bu.h, but what header file in include/ can we be absolutely sure will be present for a brlcad install (that we can search for to determine the brlcad *base* include dir? |
| 14:21.05 | d-lo | that was poorly worded, i know :/ |
| 14:21.16 | ``Erik | none |
| 14:21.34 | ``Erik | or; brlcad/bu.h :) |
| 14:21.51 | d-lo | the only guranteed dir is include/brlcad/ then? |
| 14:22.11 | ``Erik | if we can find everything we need on the system and don't touch src/other, then the only two things in that dir should be brlcad/ and tie/ |
| 14:23.21 | ``Erik | huh, chinese t69's are being dumped in the gulf of thailand to form an artificial reef |
| 14:23.30 | d-lo | kk. so assuming that openNURBS/ would be: include/brlcad/../openNURBS is probably a bad idea... |
| 14:24.56 | ``Erik | theoretically as bad of an idea as assuming include/brlcad/../zlib.h (though I don't think anyone actually packages opennurbs yet) |
| 14:33.23 | d-lo | okay. Does --enable-all make the build system assume there is NOTHING on the system and needs to make everything in src/other ? |
| 14:33.35 | ``Erik | it should |
| 14:33.50 | d-lo | okay, so next assumption: |
| 14:34.24 | d-lo | if I were to do a --enable-all and set the --prefix to a fresh (empty) dir |
| 14:34.54 | d-lo | then the resultant headers in that fresh dir is representative of header brlcad needs ? |
| 14:35.06 | ``Erik | um, probably? :D |
| 14:35.19 | d_rossberg | starseeker: this is my cmake output on windows: http://pastebin.org/471875 |
| 14:35.29 | ``Erik | (aside from OS specific stuff) |
| 14:36.13 | d_rossberg | and on the second run: http://pastebin.org/471886 |
| 14:45.29 | starseeker | d_rossberg: ah, excellent - thanks! |
| 14:45.47 | starseeker | hmm, looks like the day/month assumptions are different - must be locale specific |
| 14:45.51 | starseeker | doggone it |
| 14:46.05 | starseeker | not sure what to do about that - detect it somehow? |
| 14:46.55 | starseeker | what version of CMake are you using on Windows? I'm surprised that if clause fails on Windows but not on Linux... |
| 14:58.13 | d_rossberg | version 2.8.2 |
| 14:58.48 | starseeker | sigh... ok - I'll try nexting that test |
| 14:59.16 | starseeker | d_rossberg: do you know of any way to either tell the Windows data command a specific format or determine what region Windows is set to from the command prompt? |
| 15:06.28 | CIA-42 | BRL-CAD: 03starseeker * r40105 10/brlcad/branches/cmake/ (CMakeLists.txt src/libfb/CMakeLists.txt): Tweak the if logic for data directory settings |
| 15:06.50 | starseeker | d_rossberg: that may resolve the unknown arguments error |
| 15:07.24 | starseeker | looks into what to do about lex/yacc on Windows... |
| 15:08.24 | d_rossberg | starseeker: no, i used date and time until now as simple time stamping strings not caring how they look like |
| 15:08.36 | starseeker | nods |
| 15:09.00 | starseeker | what little I can find on the Web suggests that this isn't something that can be done from the Windows command line |
| 15:09.11 | starseeker | ponders... hmm... |
| 15:12.07 | d_rossberg | the BRLCAD_PREFIX isn't set for windows |
| 15:14.47 | starseeker | ah, whoops |
| 15:20.56 | CIA-42 | BRL-CAD: 03starseeker * r40106 10/brlcad/branches/cmake/CMakeLists.txt: Put some kind of default in for Windows - probably will need something smarter later. |
| 15:30.06 | brlcad | d-lo: if brlcad-config doesn't report all of the libs, that's a bug that should be fixed |
| 15:30.28 | brlcad | what's not reporting? |
| 15:32.32 | d-lo | most of them, actually. *compiles list* |
| 15:32.40 | d-lo | the one I am specifically after is libpkg |
| 15:32.42 | starseeker | brlcad: uh. What do I do about lex/yacc code on Windows? A search suggests that Visual Studio doesn't support lex/yacc, and that means either a) we require something like gnuwin32's ports of flex and bison to be installed or b) we build our own copies of lex/yacc on Windows (in which case we may as well build our own optionally on all platforms...) |
| 15:33.21 | brlcad | _psilva: the dll is pretty snazzy if you just want to link against one dll and get access to a lot of our library goodies |
| 15:38.22 | d_rossberg | starseeker: it often works this way: the generated files will be checked in too |
| 15:39.06 | starseeker | gah |
| 15:41.24 | d_rossberg | the "datadir:" message is red (maybe because of a missing "STATUS") |
| 15:41.57 | starseeker | does the read clear up after one more configure? |
| 15:42.03 | starseeker | red rather |
| 15:46.36 | d_rossberg | no, it stays red |
| 15:47.18 | starseeker | must not be a valid path |
| 15:47.31 | starseeker | I'll have to scare up another system intended to build on Windows and see what they do |
| 15:48.25 | d_rossberg | as i said: the option argument is missing in this message |
| 15:48.33 | d_rossberg | (?) |
| 15:49.01 | brlcad | d-lo: the reason it takes a string is that you can request for a service by name instead of by port number |
| 15:49.26 | brlcad | d-lo: so you can connect to 128.32.32.128:smtp for example |
| 15:49.38 | brlcad | and it'll do the proper mapping for smtp |
| 15:49.46 | brlcad | getservbyname() |
| 15:49.58 | d-lo | gotcha. erik skooled me already |
| 15:50.45 | brlcad | ahh, yeah, .. what ``Erik already told you ;) |
| 15:53.18 | brlcad | catches up with the log |
| 15:54.20 | brlcad | d-lo: er, it has libpkg |
| 15:55.08 | brlcad | run brlcad-config without any arguments, it'll list the libraries it's set up for |
| 15:57.53 | d-lo | brlcad-config --libs doesnt show it. |
| 16:00.30 | brlcad | starseeker: some check them in, some rely on dist prep (so you have to make dist on a platform with lex/yacc, but then can compile that source tarball anywhere) |
| 16:01.35 | brlcad | d-lo: notice in the Libraries list, it's merely defaulting to the libbrlcad libs if you don't specify one |
| 16:02.19 | brlcad | you shouldn't be relying on the default, specify the libs you want |
| 16:03.48 | d-lo | Well, I was trying to fix part of Dr Rossberg's FindBRLCAD.cmake-fu, but have since abandoned the notion of using brlcad-config to obtain a list of installed libraries and header locations. |
| 16:04.06 | d-lo | ...since last i checked brlcad-config isn't available on windows. |
| 16:09.03 | brlcad | making brlcad-config work on windows is less work than any maintainable alternative work-around that I can think of |
| 16:26.46 | yukonbob | morning #brlcad |
| 16:27.29 | d-lo | mernin! |
| 16:29.12 | yukonbob | mawnin |
| 16:29.13 | yukonbob | :0 |
| 16:29.17 | yukonbob | :) |
| 16:29.38 | yukonbob | irssi consuming the ) after a : |
| 16:29.50 | yukonbob | nom nom nom *burp* |
| 16:29.55 | yukonbob | mmmmmmmm.... parens. |
| 16:40.53 | _psilva | :) |
| 16:41.15 | _psilva | no consumption here :\ |
| 16:41.49 | brlcad | bad silva, no alco fer u |
| 16:44.18 | _psilva | the russians here have a stockpile of vodka |
| 17:21.59 | d-lo | haha, just got an email from a friend: "Just wanted to let you know - today I received my 2010 Obama Stimulus Package. It contained two watermelon seeds, cornbread mix, and 10 Coupons to KFC. The directions were in Spanish." |
| 17:27.35 | _psilva | ok |
| 17:43.20 | *** join/#brlcad IriX64 (~MarioDUli@bas2-sudbury98-1177872371.dsl.bell.ca) |
| 17:48.58 | *** join/#brlcad mafm (~mafm@83.45.73.176) |
| 18:00.01 | CIA-42 | BRL-CAD: 03starseeker * r40107 10/brlcad/branches/cmake/ (3 files in 2 dirs): |
| 18:00.01 | CIA-42 | BRL-CAD: OK, take a very serious stab at getting SOME kind of robust date information. |
| 18:00.01 | CIA-42 | BRL-CAD: Use configure_file, try_run, and every other trick I know of to get this working |
| 18:00.01 | CIA-42 | BRL-CAD: - probably not cross platform yet but should update time.c.in accordingly to try |
| 18:00.02 | CIA-42 | BRL-CAD: and get this compiling everywhere. |
| 18:52.47 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) |
| 18:54.50 | CIA-42 | BRL-CAD: 03starseeker * r40108 10/brlcad/branches/cmake/CMakeLists.txt: Clear stray debugging printout. |
| 19:39.28 | CIA-42 | BRL-CAD: 03erikgreenwald * r40109 10/brlcad/branches/bottie/ (include/tie.h src/librt/primitives/bot/tie.h): move tie.h into include/ |
| 19:40.29 | CIA-42 | BRL-CAD: 03erikgreenwald * r40110 10/brlcad/branches/bottie/src/librt/Makefile.am: compile in the tie stuff with float/double hack |
| 19:41.06 | CIA-42 | BRL-CAD: 03erikgreenwald * r40111 10/brlcad/branches/bottie/include/Makefile.am: add tie.h to the list |
| 19:42.03 | CIA-42 | BRL-CAD: 03erikgreenwald * r40112 10/brlcad/branches/bottie/src/adrt/ (7 files in 3 dirs): reflect the exodus of tie from this dark twisty maze |
| 19:46.13 | brlcad | you could make two noinst libraries with different CFLAGS to get the same effect without the need for a hack |
| 19:46.36 | ``Erik | yeah, that's how it is in trunk |
| 19:47.25 | brlcad | so it was actually more work to do the hack .. nice :) |
| 19:48.02 | ``Erik | in this case, I'm just calling $(LTCOMPILE) with immediate targets *shrug* good 'nuff for this experiment, should probably have the entire approachc ompletely changed at some point |
| 19:48.10 | ``Erik | nah, I use vim, not emacs, so it was like 4 keypresses :D |
| 19:48.41 | brlcad | not from the diff I'm seeing |
| 19:50.14 | brlcad | you removed the "clean" bits from src/adrt .. yet still had to go to another file (presumably Makefile) to come up with ltcompile lines that would work (which are only guaranteed on that platform btw) |
| 19:50.53 | brlcad | whatever, it's your playground branch .. |
| 19:51.03 | brlcad | it was just actually mergeable until that point |
| 19:53.44 | ``Erik | *shrug* if you wanna swap the noinst_LTLIBRARIES crap back in, knock yourself out, I'm getting ready to make a bit mess of the source during this tiny bits of time between iaso, dts, etc... |
| 19:54.05 | brlcad | like I said, your playground |
| 19:54.12 | starseeker | growls... source for flex is 2.1 megs with tests and such stripped out, and byacc is 270K |
| 19:54.18 | brlcad | I'm not touching it, especially after you pee in the sand :) |
| 19:55.38 | starseeker | considers a nastygram to the Visual Studio devs... come on guys, they're BSD licensed, why aren't they included??? |
| 19:56.50 | ``Erik | notes that LTCOMPILE is defined exactly the same way on 3 different os's O.o |
| 19:56.58 | brlcad | and worse when you don't own up to having done something lame, but again... |
| 19:57.00 | brlcad | me really no care -- whatever :) |
| 19:57.20 | ``Erik | I'm not sure it's all that lame... both ways are hacks, just different hacks *shrug* |
| 19:57.21 | brlcad | imagine that, they all are set up to run gcc |
| 19:57.30 | starseeker | you think that's lame - watch me (sooner or later) integrate flex and byacc with CMake into a Windows build of BRL-CAD :-P |
| 19:58.05 | ``Erik | they're hard defined before the notion of which compiler comes into it, that's why it's LTCOMPILE instead of COMPILE |
| 19:58.12 | brlcad | the noinst isn't a hack, that's how they say to do that |
| 20:00.25 | brlcad | there's post-configure assumptions on those lines in addition to the (-MT -MD, etc) compiler flags too, like whether dependency tracking is enabled |
| 20:00.34 | brlcad | turn off dependency tracking and the build is probably busted |
| 20:04.21 | brlcad | if you actually believe that the 16 repetitious lines are better than the 6 lines it took to describe the noinst libs, then I don't think there's any value in continuing that thread of discussion :) |
| 20:04.35 | brlcad | i'm all for quick hacks, especially in playgrounds |
| 20:06.02 | brlcad | starseeker: ... yeah ... why? :) |
| 20:06.45 | brlcad | those are nasty deps to pick up for such a tiny fraction of the codebase.. |
| 20:17.50 | starseeker | assumes a) were going to be using it for more of the codebase eventually (conv) and b) wants to try to do Windows building "right" if he's going to do it at all |
| 20:21.34 | CIA-42 | BRL-CAD: 03starseeker * r40113 10/brlcad/branches/cmake/CMakeLists.txt: Start putting together the logic needed for the 'summary' printout at the end of the build. |
| 20:22.13 | brlcad | starseeker: I'd argue that carrying around a fully developed lexer and parser aren't the "right" way :) |
| 20:22.40 | starseeker | well, if we require the installation of the gnuwin32 flex and bison that could work too |
| 20:22.51 | starseeker | I cringe at the idea of checking in generated files though |
| 20:22.57 | starseeker | only as a last resort |
| 20:23.24 | brlcad | I cringe too, though that pretty much is the most common practice for supporting windows |
| 20:23.42 | starseeker | common != good, and I would like to be good... |
| 20:23.44 | brlcad | requiring the installation of a lexer/parser combo is reasonable |
| 20:23.53 | starseeker | nods |
| 20:24.01 | starseeker | k, time for a test... |
| 20:24.07 | brlcad | just have to document what exactly needs to be installed |
| 20:24.25 | brlcad | right now it's a compiler and nsis if you're on windows |
| 20:24.39 | brlcad | so it'd be a compiler, nsis, and a lexer/parser |
| 20:24.47 | starseeker | nods - plus CMake :-) |
| 20:25.04 | brlcad | if starting from a checkout, yeah |
| 20:25.33 | brlcad | so there's checkout requirements, and tarball requirements -- not necessarily the same |
| 20:25.48 | starseeker | actually, I'm not sure if CMake builds function without CMake installed... |
| 20:26.03 | brlcad | that's where some projects opt for no lexer/parser requirement to compile from source tarball |
| 20:26.28 | brlcad | currently that's what we do (or at least how it's supposed to be set up) |
| 20:26.46 | brlcad | they should function -- that's one of the points |
| 20:27.19 | brlcad | spitting out msvc or makefile or xcode projects or whatever |
| 20:28.21 | starseeker | oh, sure - but I'm not sure if the makefiles would be portable system to system |
| 20:28.22 | brlcad | did I mention that one of the kitware devs was at mil-oss? |
| 20:28.32 | starseeker | was he? Sweet! |
| 20:29.15 | brlcad | yeah, patrick reynolds |
| 20:29.32 | brlcad | gave an ignite talk on continuous integration and build testing |
| 20:29.45 | starseeker | Once a cmake configure has run, it can build without CMake - but unlike autotools, you do need cmake to do the configure - there's no configure script generated from CMakeLists.txt, unless I'm misunderstanding how it works |
| 20:30.05 | starseeker | brlcad: was that the cdash/ctest stuff they have? |
| 20:30.19 | brlcad | yeah |
| 20:30.58 | brlcad | so yeah, you don't end up with a portable build |
| 20:31.13 | brlcad | but you end up with a self-contained build for a given platform environment |
| 20:31.22 | starseeker | right |
| 20:33.49 | brlcad | starseeker: curious, did you try running http://websvn.kde.org/trunk/KDE/kdesdk/cmake/scripts/ ? |
| 20:33.52 | brlcad | am2cmake |
| 20:34.56 | starseeker | whoops... no, I didn't |
| 20:35.01 | brlcad | think i've seen a php variant too |
| 20:35.15 | brlcad | might be good for an 80% bootstrap conversion |
| 20:35.27 | starseeker | nods |
| 20:35.34 | brlcad | no way it'll fully work, but might save a few days |
| 20:35.50 | starseeker | that could be very handy when I start hitting binaries - rt alone took forever |
| 20:35.52 | brlcad | http://emanuelgreisen.dk/stuff/kdevelop_am2cmake.php.tgz is another |
| 20:36.19 | brlcad | if anything, could take one of those two scripts and extend it to work with our layout |
| 20:36.44 | brlcad | throwaway tool to help get the job done |
| 20:36.59 | starseeker | nods |
| 20:37.01 | brlcad | ah ha! |
| 20:37.02 | brlcad | http://www.cmake.org/Wiki/CMake#Converters_from_other_buildsystems_to_CMake |
| 20:37.21 | brlcad | might even get some mileage out of the vstudio2cmake :) |
| 20:38.22 | starseeker | definitely, once I start actually trying to build on Windows :-P |
| 20:39.52 | starseeker | dreads that day - by the very nature of the beast I'll probably be trying to build a lot of things never before tried on Windows |
| 20:40.41 | starseeker | makes note of this - will be relevant http://www.cmake.org/Wiki/CMake:Multiple_versions |
| 20:43.23 | brlcad | starseeker: ah, found another confirmation that it's expected that one ship cmake with one's source code in order to bootstrap+compile |
| 20:44.28 | starseeker | brlcad: is that acceptable to us? |
| 20:44.55 | brlcad | it's not a show-stopper, just a tradeoff we don't presently have to put up with |
| 20:46.11 | starseeker | nods - I guess in principle all we need now is sh, m4 and make... |
| 20:47.24 | brlcad | it actually reminds me a bit of our old "cake" build |
| 20:47.36 | starseeker | winces |
| 20:47.38 | starseeker | ouch |
| 20:47.53 | brlcad | we bootstrap from checkout, build cake, then compile with it |
| 20:48.07 | brlcad | in theory, cmake wouldn't be much different of a setup |
| 20:48.49 | brlcad | except that you build cmake with cmake, heh |
| 20:49.18 | brlcad | ahh, they have a bootstrap |
| 20:49.53 | starseeker | oh, crud - looks like Visual Studio 2010 is now the "current" version |
| 20:51.40 | _psilva | heh |
| 20:52.14 | _psilva | we had to drop 2003 support to accomodate this 'current' version |
| 20:52.39 | _psilva | but the asians like their obsolete versions |
| 20:54.38 | brlcad | starseeker: can get it working in visual c++ express (free msvc8) |
| 20:55.01 | starseeker | er - that's what I mean - their express version is now 2010 |
| 20:55.20 | starseeker | lord knows what that breaks |
| 20:55.24 | brlcad | ah, right |
| 20:56.03 | brlcad | hm, I should try a build with express |
| 21:01.35 | _psilva | u can always try to use its 2k8->2k10 proj/sln conversion |
| 21:08.38 | CIA-42 | BRL-CAD: 03r_weiss * r40114 10/brlcad/trunk/src/librt/primitives/bspline/nurb_eval.c: added bu_bomb to catch possible invalid index returning from function rt_nurb_knot_index |
| 21:16.38 | *** join/#brlcad cosurgi (~cosurgi@atak.bl.pg.gda.pl) |
| 21:20.28 | CIA-42 | BRL-CAD: 03r_weiss * r40115 10/brlcad/trunk/src/librt/primitives/bspline/nurb_eval.c: added bu_bomb to catch possible invalid index returning from function rt_nurb_knot_index |
| 21:28.40 | CIA-42 | BRL-CAD: 03n_reed * r40116 10/brlcad/trunk/src/tclscripts/boteditor/ (botEditor.tcl botPropertyBox.tcl botTools.tcl): bot editor interface improvements; updating property info when commands are run |
| 21:31.22 | CIA-42 | BRL-CAD: 03erikgreenwald * r40117 10/brlcad/branches/bottie/include/ (bot.h rtgeom.h): stub the TIE instance into both bot representations. |
| 21:31.32 | CIA-42 | BRL-CAD: 03erikgreenwald * r40118 10/brlcad/branches/bottie/src/librt/primitives/bot/g_bot_include.c: bind the TIE instances |
| 21:31.48 | CIA-42 | BRL-CAD: 03erikgreenwald * r40119 10/brlcad/branches/bottie/src/librt/primitives/bot/bot.c: start wiring up conditional TIE execution |
| 22:05.56 | *** join/#brlcad CIA-42 (~CIA@208.69.182.149) |
| 22:11.41 | *** join/#brlcad CIA-42 (~CIA@208.69.182.149) |
| 23:00.34 | *** join/#brlcad IriX64 (~MarioDUli@bas2-sudbury98-1177879057.dsl.bell.ca) |
| 23:14.09 | *** join/#brlcad IriX64 (~MarioDUli@bas2-sudbury98-1178015089.dsl.bell.ca) |
| 23:24.55 | brlcad | any gov'y interested in a potential $500-1000 innovation prize? http://www.defense.gov/home/features/2010/0710_invest/ |
| 23:35.39 | CIA-42 | BRL-CAD: 03brlcad * r40120 10/brlcad/trunk/Makefile.am: need to capture all the numbers of the triplet, not just the first digit otherwise it'll just keep expanding the number out like 7.16.10000 |
| 23:54.16 | CIA-42 | BRL-CAD: 03brlcad * r40121 10/brlcad/trunk/src/librt/primitives/bspline/ (24 files): ws indent style cleanup along with doxygen comment conversion. |