| 00:14.33 | CIA-88 | BRL-CAD: 03starseeker * r40467 10/brlcad/branches/cmake/src/brlman/CMakeLists.txt: Use a variation of a trick from the CMake list to leave intact variables not using the @var@ style in brlman.sh.in. |
| 00:27.46 | CIA-88 | BRL-CAD: 03starseeker * r40468 10/brlcad/branches/cmake/ (CMakeLists.txt src/burst/CMakeLists.txt): Add CMake logic for burst. Not implementing the custom test routine as yet - that involves figuring out CTest and will come later. |
| 01:01.53 | CIA-88 | BRL-CAD: 03starseeker * r40469 10/brlcad/branches/cmake/ (5 files in 5 dirs): |
| 01:01.53 | CIA-88 | BRL-CAD: Add in some missed CMakeLists.txt files, and start working on bwish. FindX11 |
| 01:01.53 | CIA-88 | BRL-CAD: paths by default weren't finding Xi on gentoo, so now have local copy of FindX11 |
| 01:01.53 | CIA-88 | BRL-CAD: with expanded paths. btclsh and bwish appear to truly need the itcl/itk C |
| 01:01.53 | CIA-88 | BRL-CAD: libraries, which means the next step is to back up and add a C component to the |
| 01:01.53 | CIA-88 | BRL-CAD: Itcl/Itk definitions both for Find and for local compiling. |
| 02:54.42 | *** join/#brlcad kenagain (~kenagain@d149-67-221-118.try.wideopenwest.com) |
| 02:55.08 | *** part/#brlcad kenagain (~kenagain@d149-67-221-118.try.wideopenwest.com) |
| 03:43.59 | *** join/#brlcad Ralith (~ralith@S010600221561996a.vc.shawcable.net) |
| 04:17.16 | starseeker | brlcad: just curious - do you know why we're talking to itcl/itk via the C api instead of just using it via Tcl? |
| 04:20.41 | starseeker | is wondering why we can't just do Tcl_Eval(interp, "package require Itcl") instead of calling Itcl_Init and friends... |
| 05:05.12 | *** join/#brlcad Ralith_ (~ralith@S010600221561996a.vc.shawcable.net) |
| 07:11.46 | *** join/#brlcad louipc (~louipc@archlinux/trusteduser/louipc) |
| 07:28.07 | *** join/#brlcad SWPadnos_ (~Me@216.114.141.108) |
| 07:54.54 | *** join/#brlcad SWPadnos_ (~Me@dsl107.esjtvtli.sover.net) |
| 08:00.22 | *** join/#brlcad Led_Zeppelin (~user@ool-4573f43b.dyn.optonline.net) |
| 08:17.15 | *** join/#brlcad sub-vigilance (~raijin@pool-108-9-15-164.tampfl.fios.verizon.net) |
| 08:53.49 | *** join/#brlcad raininja (~raijin@unaffiliated/raijin) |
| 09:39.29 | *** join/#brlcad merzo (~merzo@82-146-132-95.pool.ukrtel.net) |
| 11:22.28 | *** part/#brlcad Led_Zeppelin (~user@ool-4573f43b.dyn.optonline.net) |
| 11:42.27 | *** join/#brlcad SWPadnos (~Me@emc/developer/SWPadnos) |
| 11:50.37 | *** join/#brlcad Ralith (~ralith@S010600221561996a.vc.shawcable.net) |
| 12:55.13 | *** join/#brlcad Nohla (~Nohla@201.255.233.105) |
| 13:11.52 | ``Erik | hm, a couple weeks ago, .bz's current uptime broke it's previous #2 O.o still quite a ways to break #1 though :) |
| 14:09.08 | CIA-88 | BRL-CAD: 03starseeker * r40470 10/brlcad/branches/cmake/ (6 files in 3 dirs): (log message trimmed) |
| 14:09.08 | CIA-88 | BRL-CAD: No idea if I'll get away with this, but rip out the C calls to Itcl/Itk init |
| 14:09.08 | CIA-88 | BRL-CAD: routines and replace them with package require calls. Also need to define |
| 14:09.08 | CIA-88 | BRL-CAD: ITCL_VERSION in brlcad_config.h since we're not directly using the itcl/itk |
| 14:09.08 | CIA-88 | BRL-CAD: headers any more. Also take out the while loop in bwish and just call the |
| 14:09.09 | CIA-88 | BRL-CAD: routine to add BRL-CAD path info up front - should be harmless if nothing there |
| 14:09.09 | CIA-88 | BRL-CAD: and odds are very good we'll need it. Much simpler, and also simplifies the |
| 15:30.06 | CIA-88 | BRL-CAD: 03starseeker * r40471 10/brlcad/branches/cmake/ (10 files in 10 dirs): Add in some more directories to CMake. Getting close to mged. |
| 16:05.39 | CIA-88 | BRL-CAD: 03starseeker * r40472 10/brlcad/branches/cmake/src/mged/ (attach.c setup.c): try the non-C itcl/itk thing in MGED too - can't tell if it works yet. |
| 16:07.06 | CIA-88 | BRL-CAD: 03starseeker * r40473 10/brlcad/branches/cmake/ (3 files in 3 dirs): OK, this builds mged but does not yet produce a working mged. |
| 16:30.55 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) |
| 17:34.21 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) |
| 19:46.29 | *** join/#brlcad Nohla (~Nohla@201.255.233.105) |
| 20:10.23 | *** join/#brlcad raijin (~raijin@pool-108-9-15-164.tampfl.fios.verizon.net) |
| 20:10.23 | *** join/#brlcad raijin (~raijin@unaffiliated/raijin) |
| 20:32.40 | *** join/#brlcad ibot (~ibot@rikers.org) |
| 20:32.40 | *** topic/#brlcad is BRL-CAD Open Source Solid Modeling || http://brlcad.org || http://sf.net/projects/brlcad || #brlcad logs: http://ibot.rikers.org/%23brlcad/ || BRL-CAD release 7.16.10 is posted! (20100805) |
| 20:34.06 | *** join/#brlcad R0b0t1 (~Enigma@64.126.35.185) |
| 20:34.07 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) |
| 20:44.45 | CIA-88 | BRL-CAD: 03starseeker * r40474 10/brlcad/branches/cmake/ (9 files in 9 dirs): |
| 20:44.45 | CIA-88 | BRL-CAD: Add in some more directories, update the nirt CMakeLists.txt. For now, turn off |
| 20:44.45 | CIA-88 | BRL-CAD: the CONF/DATE/etc updating in the topelevel CMakeLists.txt, as the constant |
| 20:44.45 | CIA-88 | BRL-CAD: addition of make logic keeps trigging configures - need to remember to re-enable |
| 20:44.46 | CIA-88 | BRL-CAD: it at the end |
| 21:23.48 | CIA-88 | BRL-CAD: 03starseeker * r40475 10/brlcad/branches/cmake/ (3 files in 3 dirs): Add in util and vdeck. The remaining dirs in src will require solving a couple different problems - (Java, installing .tcl files properly, etc.) |
| 22:03.52 | CIA-88 | BRL-CAD: 03starseeker * r40476 10/brlcad/branches/cmake/ (CMakeLists.txt db/CMakeLists.txt): Commit a CMake file for using asc2g to build and install the db models. Appears to work in parallel, but I only have 2 CPUs here so needs more testing. |
| 22:05.28 | starseeker | pants mentally |
| 22:05.35 | starseeker | hmm, tclscripts or docbook next... |
| 22:05.43 | starseeker | or man pages I suppose... |
| 22:10.01 | CIA-88 | BRL-CAD: 03starseeker * r40477 10/brlcad/branches/cmake/src/fbed/CMakeLists.txt: Whoops, got ahead of myself - need to straighten out the src/other logic before I can ditch the BRL-CAD prefix |
| 22:10.17 | brlcad | shouldn't (ever?) reference /usr/brlcad in code |
| 22:10.43 | brlcad | bu_brlcad_root() and bu_brlcad_data() are the API mechanisms to consistently find ourselves |
| 22:11.59 | brlcad | at a glance, the problem I see with the bwish changes is that it might not work uninstalled any longer (and would have to test relocation) |
| 22:12.35 | CIA-88 | BRL-CAD: 03starseeker * r40478 10/brlcad/branches/cmake/CMakeLists.txt: Turn back on the generated files. |
| 22:17.11 | CIA-88 | BRL-CAD: 03starseeker * r40479 10/brlcad/branches/cmake/src/libbu/CMakeLists.txt: Let bu know about thread stuff |
| 22:54.25 | starseeker | OK, obviously I still don't have everything in src hooked in, and there are a few stray things like bench and enigma, but a no-docs build comparison on my system between autotools and cmake, make -j2 + configure time for both systems (no autogen.sh for the autotools build) was 6.2 minutes for CMake, 10.2 minutes for autotools |
| 22:54.42 | starseeker | doesn't include install time either |
| 22:56.19 | starseeker | so not spectacularly faster, but (as expected) a little gain due (probably) to fewer IO operations in the CMake build |
| 22:57.35 | starseeker | will be curious to see how the Mac does |
| 22:58.05 | starseeker | gentoo's pretty good at running build tools fast |
| 23:00.56 | starseeker | actually, now that I think about it a fair chunk of that gain is also gonna be doing the tcl/tk configures in parallel with other builds |
| 23:01.09 | starseeker | hmm - have to try it again using local libs |
| 23:26.07 | *** join/#brlcad Nohla_ (~Nohla@201.255.233.16) |