| 00:11.45 | starseeker | Ouch http://www.wired.com/dangerroom/2010/08/hp-holds-navy-network-hostage |
| 00:16.14 | ``Erik | might be worth passing that around work |
| 00:23.41 | starseeker | nods - was thinking that |
| 00:25.01 | starseeker | government needs to retain control when hiring contractors to do work... |
| 00:27.45 | starseeker | brlcad: do you know of any environment we need to support that needs AC_HEADER_STDC or AC_HEADER_DIRENT? (or, for that matter, other macros identified as obsolete by the Autoconf manual?) |
| 00:28.15 | starseeker | I'd hate to go to the trouble of duplicating them if we don't need 'em |
| 00:51.18 | CIA-2 | BRL-CAD: 03starseeker * r40909 10/brlcad/branches/cmake/ (405 files in 91 dirs): Update cmake branch to r40906 |
| 01:18.44 | CIA-2 | BRL-CAD: 03starseeker * r40910 10/brlcad/branches/cmake/CMakeLists.txt: |
| 01:18.44 | CIA-2 | BRL-CAD: The LIBSTDCXX variable is used only in misc/pkgconfig/librt.pc.in (which already |
| 01:18.44 | CIA-2 | BRL-CAD: includes LIBM anyway) and src/other/step. This test is somewhat difficult to |
| 01:18.44 | CIA-2 | BRL-CAD: set up in CMake - the standard CHECK_LIBRARY routine won't do it - and since it |
| 01:18.44 | CIA-2 | BRL-CAD: shouldn't be an issue anywhere except src/other/step remove it - if step |
| 01:18.45 | CIA-2 | BRL-CAD: libraries need the LIBM option add it. Will have to set up an OpenBSD system to |
| 01:18.46 | CIA-2 | BRL-CAD: check this out. |
| 01:28.27 | CIA-2 | BRL-CAD: 03starseeker * r40911 10/brlcad/branches/cmake/ (CMakeLists.txt misc/CMake/BRLCAD_CheckFunctions.cmake): Don't duplicate obsolete Autoconf functionality unless there is a compelling reason |
| 01:29.28 | brlcad | starseeker: AC_HEADER_STDC is more a sanity test to make sure we're at least compiling with a c89 compliant environment |
| 01:31.17 | starseeker | brlcad: I can't find anywhere in the code we actually conditionalize on it (although I could be missing something...) - it's apparently so little used that no one has duplicated it as a CMake macro |
| 01:31.32 | brlcad | AC_HEADER_DIRENT is used, defines HAVE_DIRENT_H |
| 01:32.06 | brlcad | conditionalize on what? |
| 01:32.15 | starseeker | STDC_HEADERS |
| 01:33.22 | brlcad | that's the one direct symbol it defines, but there are various other secondary ones too |
| 01:33.27 | brlcad | for the various headers it tests |
| 01:39.18 | brlcad | that said, the check was before we required c89 and that's specifically what it tests for |
| 01:40.03 | brlcad | so it's really just a sanity check in configure.ac for the log .. not necessary for cmake |
| 01:41.11 | starseeker | well, I actually did duplicate most of it... |
| 01:41.17 | brlcad | starseeker: the -lstdc++ test possibly fails because we rely on a trick |
| 01:41.29 | starseeker | figures... the one one we need, DIRENT, is the one I don't have done yet... |
| 01:41.42 | brlcad | it's just a header test |
| 01:42.04 | brlcad | granted, a really nasty header to test for if we try to really make things backwards compatible to olden days |
| 01:42.10 | brlcad | but that's not necessary |
| 01:42.26 | brlcad | we can assume that if the header is there that it's good |
| 01:42.36 | starseeker | brlcad: phew :-) |
| 01:42.43 | starseeker | that'll save a day or so |
| 01:43.00 | brlcad | i mean, we need to test HAVE_DIRENT_H |
| 01:43.10 | brlcad | just not the complex testing that AC_HEADER_DIRENT performs |
| 01:43.21 | starseeker | oh, gotcha |
| 01:43.40 | starseeker | that would explain why it worked when I did the basic test a while back... |
| 01:44.33 | brlcad | AC_HEADER_STDC isn't needed (at least, STDC_HEADERS definitely isn't needed) |
| 01:45.10 | brlcad | might be useful to log whether the standard headers aren't available merely as a sanity test during configuration, but not really necessary |
| 01:45.11 | starseeker | brlcad: I think I actually did duplicate AC_HEADER_STDC more or less correctly, so if you like we can leave it in for logging here too |
| 01:45.17 | brlcad | k |
| 01:45.40 | starseeker | amusingly enough, I don't think it passes on the Mac... |
| 01:45.47 | brlcad | -lstdc++ is almost certainly going to be needed for portability |
| 01:46.01 | brlcad | hm, that doesn't sound right |
| 01:46.16 | starseeker | I believe that gets added in by CMake itself for CXX builds... |
| 01:46.26 | starseeker | let me check my vanilla brlcad build |
| 01:47.00 | starseeker | ah, wait - it did work |
| 01:47.08 | starseeker | oh, that's right - that one test wasn't behaving |
| 01:47.18 | brlcad | the reason is that we need (or at least needed) to compile everything with gcc, NOT some with gcc and some with g++ |
| 01:47.36 | brlcad | otherwise, symbols end up wrong |
| 01:47.41 | starseeker | hrm |
| 01:48.02 | starseeker | come to think of it, I haven't tried raytracing nurbs with a CMake build yet |
| 01:48.52 | brlcad | it was a really obscure issue when it happened, not easy to reproduce |
| 01:49.31 | brlcad | like loading a .so via java or jnilib maybe |
| 01:50.03 | brlcad | or linkage on a platform that requires resolved libraries |
| 01:51.34 | brlcad | fwiw, the "trick" in configure was that it tested for 'main' .. which isn't actually in the library, but of course is a symbol that can be referenced if you make and run a program from an autoconf test .. if cmake tests differently, looking for 'main' as a library symbol could easily fail |
| 01:51.39 | starseeker | hmm - apparently I lied - STDC_HEADERS is getting defined now... |
| 01:52.27 | starseeker | brlcad: the failure in the log is a conflict for "main" types - apparently the standard C file used for that test isn't happy in the CXX compiler |
| 01:53.36 | brlcad | sure, that makes sense |
| 01:54.10 | brlcad | they're apparently testing different, perhaps just by declaring a k&r style forward declaration and trying to reference it, and the conflict makes it unhappy |
| 01:54.37 | brlcad | testing for 'main' is wrong if that wasn't clear .. it was just simple and effective for an autoconf test |
| 01:54.59 | brlcad | it should test for a symbol actually in the library |
| 01:55.29 | CIA-2 | BRL-CAD: 03starseeker * r40912 10/brlcad/branches/cmake/ (CMakeLists.txt misc/CMake/BRLCAD_CheckFunctions.cmake): Go ahead and put back the working tests - a basic dirent.h test should be sufficient these days, so don't spend any more effort on it. |
| 01:56.19 | starseeker | nods - here's the C code they're using: http://paste.lisp.org/display/115289 |
| 01:58.34 | brlcad | yeah, char CHECK_FUNCTION_EXISTS() is bogus |
| 01:58.40 | brlcad | that's dumb |
| 01:58.52 | CIA-2 | BRL-CAD: 03starseeker * r40913 10/brlcad/branches/cmake/src/CMakeLists.txt: vas4 is no more |
| 01:59.08 | brlcad | autoconf's method is much better.. |
| 01:59.18 | starseeker | brlcad: when I traced that stdc++ test back in the svn logs, it appeared to have been added as a consequence of OpenBSD needing -lm when using -lstdc++ |
| 02:00.08 | starseeker | http://brlcad.svn.sourceforge.net/viewvc/brlcad/brlcad/trunk/configure.ac?revision=28672&view=markup |
| 02:00.10 | brlcad | that sounds secondary |
| 02:00.27 | ``Erik | watches new southpark O.o |
| 02:00.28 | starseeker | oh, wait, that's when the second conditional was added |
| 02:00.45 | starseeker | dig dig dig... |
| 02:01.13 | brlcad | r27473 is the add but doesn't say why |
| 02:01.49 | brlcad | adding openNURBS was definitely the timeframe, but only because it was the first c++ code to be added |
| 02:01.54 | starseeker | nods |
| 02:02.24 | brlcad | you could let cmake do it's thing and see if the problem resurfaces |
| 02:02.32 | starseeker | nods |
| 02:02.53 | starseeker | that would be my preference, but I need to stake out a defensible position that I'm not introducing regressions :-P |
| 02:02.58 | brlcad | several of the flags and tests are due to compiling c++ with gcc |
| 02:03.21 | brlcad | the -fexeceptions flag for _Unwind_Resume |
| 02:03.24 | starseeker | there's this brlcad guy at work who keeps a close eye on things... :-P |
| 02:04.19 | brlcad | you very likely will introduce several regressions by the vary nature of the rebuild |
| 02:04.32 | brlcad | the hope is just to minimize as many as possible and hopefully they're not expensive |
| 02:04.46 | starseeker | wonders if we should add Keith's duck to the standard raytrace benchmark and/or regression tests... |
| 02:05.01 | brlcad | or hopefully, the investment pays for itself quickly in time and effort saved |
| 02:05.17 | starseeker | nods - well, only one way to find out... |
| 02:05.53 | brlcad | I wouldn't add any new benchmark images until we can baseline on a vgr=1 |
| 02:06.07 | brlcad | simh |
| 02:06.13 | starseeker | ah yes, ``Erik's missing original image :-P |
| 02:06.29 | brlcad | hm? |
| 02:06.58 | starseeker | he had put together an emulated environment of the original BSD4.4lite or whatever that was used back then, and then lost it |
| 02:07.06 | brlcad | ah |
| 02:07.10 | brlcad | I have one somewhere too |
| 02:07.15 | brlcad | in my archive |
| 02:07.19 | starseeker | IIRC, the missing simh piece was IO throttling - dunno if they've added it |
| 02:07.42 | brlcad | doubt it, but wouldn't be hard to add |
| 02:07.54 | starseeker | suppose we could set up some kind of qemu virutal machine with slow IO, install simh in that, then run :-P |
| 02:08.11 | starseeker | they did have a new release since ``Erik last tried it, iirc... |
| 02:08.12 | brlcad | literally, just a sleep() call tossed in with gettimeofday() to try and emulate a specific clock cycle |
| 02:08.19 | starseeker | ah |
| 02:08.53 | brlcad | I was already getting a tiny vgr count |
| 02:09.00 | brlcad | like single or double digits |
| 02:09.23 | starseeker | Oo, real close then |
| 02:09.38 | starseeker | shudders to think about a NURBS raytrace on vgr=1 hardware |
| 02:10.43 | ``Erik | did I tell ya that I found some simh stuff when prepping my mac for upgrade? |
| 02:10.55 | starseeker | oh, you did mention that - was it The Image though? |
| 02:11.22 | ``Erik | um, I don't know if it was the installed image, but it was all the bits to build one... I have one named ra81.img which I THINK is the 'disk' I installed to |
| 02:11.34 | starseeker | sweet |
| 02:11.56 | ``Erik | (the ra81 was a hard drive the size (and sound) of a washing machine, supposedly the drive would start to walk across the floor if it got a bit unbalanced |
| 02:12.00 | ``Erik | ) |
| 02:12.25 | starseeker | notes we also need an ascii NURBS format before any such file could become a "standard"... |
| 02:12.30 | starseeker | now that's hardware :-) |
| 02:14.24 | ``Erik | at one point, kermit was going to look for old manuals with timing information to help us 'fix' simh some |
| 02:15.05 | ``Erik | faking seek and scan times as well as bus delays might be ... interesting |
| 02:15.09 | CIA-2 | BRL-CAD: 03starseeker * r40914 10/brlcad/branches/cmake/src/proc-db/CMakeLists.txt: terrain is now wavy |
| 02:15.13 | starseeker | we need a PDP-11 vax? |
| 02:15.41 | starseeker | er PDP-11 or VAX |
| 02:16.25 | ``Erik | vax 11/780, with mods |
| 02:16.43 | brlcad | vgr was a vax 11/780 |
| 02:16.59 | starseeker | where do we get documentation on the mods? |
| 02:17.07 | ``Erik | from kermit... he did them |
| 02:17.59 | starseeker | ah |
| 02:18.12 | ``Erik | iirc, he and mike turned vgr into the only dual core vax ... ever... |
| 02:20.20 | starseeker | if you found your image, this might be a good time to remind him :-) |
| 02:20.41 | ``Erik | um, he's probably busy shuffling money right now |
| 02:20.56 | starseeker | heh |
| 02:21.11 | ``Erik | 'real soon now', I'll drift by to chat to him about that and isst |
| 02:21.22 | starseeker | nods |
| 02:22.09 | ``Erik | (amusingly, lee's trying to claim he's busy shuffling money, too) |
| 02:22.57 | ``Erik | I think he got funded to try doing some, uh, upstairs type code using nvidia's "optix" |
| 02:23.21 | ``Erik | might be interesting |
| 02:23.35 | starseeker | well, if you have 2 dollar bills you can "shuffle", right? :-P |
| 02:23.55 | ``Erik | isn't there actually a valid legal tender $2 bill? |
| 02:24.03 | starseeker | yep :-) |
| 02:24.23 | starseeker | if you want to mess with a cashier's brain, find a few and use 'em :-) |
| 02:24.28 | ``Erik | yeh |
| 02:24.51 | ``Erik | woz likes to buy sheets of money and have a printer perf them to mess with people... he's gotten 'detained' for it :) |
| 02:26.12 | starseeker | blinks - who's been working with randmt.c in libbn? |
| 02:27.41 | ``Erik | me, why? |
| 02:28.03 | ``Erik | yes, it's ugly, it almost looks like c++ or java, shup |
| 02:28.04 | starseeker | /src/libbn/randmt.c:59: warning: integer constant is too large for âlongâ type |
| 02:28.14 | starseeker | and a few other such errors |
| 02:29.17 | ``Erik | oh, hah, magic is too wide |
| 02:30.51 | ``Erik | testing a build now |
| 02:31.25 | CIA-2 | BRL-CAD: 03starseeker * r40915 10/brlcad/branches/cmake/src/libbn/randmt.c: Comment out unused parameters so strict flags don't complain about these lines - this reduces the errors to complaints about integer constants being too large and missing initializer |
| 02:31.45 | ``Erik | heh. |
| 02:31.50 | ``Erik | so much for that testing pass |
| 02:32.02 | starseeker | heh, sorry |
| 02:32.07 | starseeker | figured I'd swat the one I could |
| 02:32.19 | ``Erik | auto* stuff got confused, so I had to start an autogen.sh |
| 02:32.28 | CIA-2 | BRL-CAD: 03starseeker * r40916 10/brlcad/trunk/src/libbn/randmt.c: Do the commenting in the trunk too. |
| 02:33.02 | ``Erik | my program to convert a word into hex added the newline character at the end, that's the issue I think |
| 02:33.20 | ``Erik | the 0A at the end of the magic is wrong |
| 02:33.52 | starseeker | ah |
| 02:36.14 | ``Erik | heh, 'ghougle' |
| 02:37.02 | CIA-2 | BRL-CAD: 03erikgreenwald * r40917 10/brlcad/trunk/src/libbn/randmt.c: fix magic overflow |
| 02:37.31 | CIA-2 | BRL-CAD: 03brlcad * r40918 10/brlcad/trunk/src/other/tkhtml/Makefile.am: let the build work even if we're configured to use our own tcl/tk because we need to be able to include built files in the source distribution (before tcl is built). |
| 02:39.27 | brlcad | starseeker: instead of commenting params out, you can also wrap the variable name with UNUSED() |
| 02:39.46 | brlcad | e.g.: int main(int UNUSED(argc), char *UNUSED(argv)[]) |
| 02:39.51 | starseeker | oh, OK - that's cleaner? |
| 02:39.51 | CIA-2 | BRL-CAD: 03starseeker * r40919 10/brlcad/branches/cmake/src/libbn/randmt.c: grab the MERSENNE_MAGIC fix from trunk |
| 02:40.05 | ``Erik | usually does self assignment in the body for 'stuff in progress' |
| 02:40.24 | brlcad | that implies, "hey, this is the API and it's supposed to be this way, but we're not yet using these parameters) |
| 02:41.44 | starseeker | ``Erik: still getting a missing initializer warning |
| 02:42.00 | brlcad | self-assignment only works because gcc is stupid or we have some flag that says obey us even when we're stupid |
| 02:42.10 | starseeker | randmt.c:59: warning: missing initializer |
| 02:42.18 | starseeker | randmt.c:59: warning: (near initialization for âglobal_state_static.mtâ) |
| 02:42.19 | brlcad | UNUSED should work cross-platform |
| 02:42.24 | ``Erik | oh, hrmmmm |
| 02:42.28 | ``Erik | I kinda cheated on that one |
| 02:42.46 | ``Erik | stop using that bustedassed gentoo crap |
| 02:42.47 | ``Erik | ? |
| 02:42.50 | ``Erik | :D *duck* |
| 02:42.58 | starseeker | nope, on Mac at work :-P |
| 02:43.15 | ``Erik | huh, worked on my mac with the strict flags all on |
| 02:43.25 | ``Erik | I'm still 10.5, though |
| 02:43.46 | brlcad | maybe has --enable-warnings on, adds more |
| 02:43.46 | starseeker | 10.5.8 <shrug> |
| 02:43.57 | starseeker | no, it's the cmake build |
| 02:44.05 | starseeker | hang on, here's the line... |
| 02:44.21 | ``Erik | cheated with a partial fill of a static struct |
| 02:44.22 | CIA-2 | BRL-CAD: 03starseeker * r40920 10/brlcad/branches/cmake/src/libbn/randmt.c: Use UNUSED to wrap variables instead of commenting out |
| 02:45.16 | starseeker | cd /Users/user/brlcad/cmake-build/src/libbn && /usr/bin/gcc -Dlibbn_EXPORTS -DHAVE_CONFIG_H -DBRLCADBUILD=1 -g -ggdb3 -D_FORTIFY_SOURCE=2 -fPIC -I/Users/user/brlcad/cmake-build/include -I/Users/user/brlcad/cmake/include -I/Users/user/brlcad/cmake/../brlcad-install/include -pedantic -W -Wall -Werror -Wno-long-long -o CMakeFiles/libbn.dir/randmt.c.o -c /Users/user/brlcad/cmake/src/libbn/randmt.c |
| 02:46.11 | CIA-2 | BRL-CAD: 03erikgreenwald * r40921 10/brlcad/trunk/src/libbn/randmt.c: set the first element of the array, hopefully this is good enough to silence the warning |
| 02:47.40 | brlcad | curious, those are just the strict warning flags |
| 02:47.55 | brlcad | different versions of gcc perhaps |
| 02:47.58 | ``Erik | yeah, looks like about the same that my macS are fine with |
| 02:48.11 | starseeker | huh. Anyway, that got it |
| 02:48.15 | starseeker | thanks ``Erik |
| 02:48.30 | CIA-2 | BRL-CAD: 03starseeker * r40922 10/brlcad/branches/cmake/src/libbn/randmt.c: Import fix for initializer from trunk. |
| 02:48.30 | ``Erik | np, my mess anyways |
| 02:48.41 | ``Erik | (even though I'm doing that for s2) |
| 02:49.15 | starseeker | is just making sure CMake is still building after syncing to latest trunk |
| 02:49.28 | CIA-2 | BRL-CAD: 03starseeker * r40923 10/brlcad/trunk/src/libbn/randmt.c: Switch trunk to UNUSED as well. |
| 02:56.32 | brlcad | UNUSED(*is) wrong, *UNUSED(is) right |
| 02:56.46 | starseeker | oh, sorry |
| 02:56.52 | brlcad | just the name -- it still needs the right type |
| 02:59.14 | starseeker | will fix in a sec - defining the HAVE_CARBON_CARBON_H variable revealed a problem |
| 02:59.35 | starseeker | is forced to wonder if focus.c is still necessary, since it was apparently off in his build and he didn't notice it... |
| 02:59.41 | starseeker | probably just didn't use things enough |
| 03:02.22 | CIA-2 | BRL-CAD: 03starseeker * r40924 10/brlcad/branches/cmake/src/libbn/randmt.c: Fix use of UNUSED |
| 03:03.41 | CIA-2 | BRL-CAD: 03starseeker * r40925 10/brlcad/branches/cmake/src/libbn/randmt.c: whoops, get both |
| 03:03.44 | CIA-2 | BRL-CAD: 03starseeker * r40926 10/brlcad/trunk/src/libbn/randmt.c: fix use of UNUSED in trunk too. |
| 03:03.49 | brlcad | focus brings mged to focus if you invoke within Terminal |
| 03:05.18 | starseeker | ah |
| 03:12.26 | CIA-2 | BRL-CAD: 03starseeker * r40927 10/brlcad/branches/cmake/ (3 files in 3 dirs): |
| 03:12.27 | CIA-2 | BRL-CAD: Now that the Carbon header flag is on, we need the results of FindCarbon for |
| 03:12.27 | CIA-2 | BRL-CAD: libdm - in the process, discovered that we need to special case framework |
| 03:12.27 | CIA-2 | BRL-CAD: arguments when passed in as part of a lib list to the BRLCAD macros - they (and |
| 03:12.27 | CIA-2 | BRL-CAD: only they, so far, in that they are multi-word non-list returns from a find |
| 03:12.27 | CIA-2 | BRL-CAD: macro) need a space between arguments to be preserved. We're now building |
| 03:12.27 | CIA-2 | BRL-CAD: successfully on Mac again, although functionality testing is not done yet. |
| 03:18.31 | starseeker | ``Erik: you're familiar with the QTIME and XTIME parameters described here? http://simh.trailing-edge.com/pdf/vax780_doc.pdf |
| 03:20.41 | starseeker | or I suppose those timing parameters is what we need historical notes on... |
| 06:42.11 | *** join/#brlcad IriX64 (~kvirc@bas2-sudbury98-1177593749.dsl.bell.ca) |
| 07:04.37 | *** join/#brlcad merzo (~merzo@smartbussiness.mobicom.net.ua) |
| 13:06.31 | starseeker | wonders if the "EXPRESS" file for IFC is the express the NIST tools work with... http://www.iai-tech.org/products/ifc_specification/ifc-releases/ifc2x3-release/ifc2x3-release-summary |
| 13:08.51 | ``Erik | don't recall seeing those parameters, but yeh, need the info on what exactly we had to figure out the statistical delays |
| 13:11.06 | starseeker | pulls his brain out of the winding twisty passages of CAD standards and gets ready to head in... |
| 13:36.38 | brlcad | starseeker: yes, it is |
| 13:37.01 | brlcad | http://www.iai-tech.org/products/ifc-overview |
| 13:37.36 | brlcad | remember that step covers end-to-end product lifecycle data management |
| 13:38.19 | brlcad | interesting .. they have some real numbers on the bloat incurred with step vs step-xml |
| 13:38.32 | brlcad | 3x to 4x larger |
| 13:41.01 | starseeker | so in principle we could grab that ifc file (if its license permits) and pipe ifc files through the step convertor? (at least in principle?) |
| 13:41.14 | brlcad | sure |
| 13:41.25 | brlcad | the converter wouldn't know what to do with it, but it could read it |
| 13:41.30 | starseeker | schweet |
| 13:42.13 | brlcad | they're just part 21 files (i.e., step files) for both |
| 13:42.27 | brlcad | but then that's basically just a container format |
| 13:42.54 | brlcad | what's in that container is the AP that it conforms to, which we read AP203 .. but those building files are certainly not 203 |
| 13:43.07 | starseeker | nods - I guess the real work would be translating ifc concepts into geometric primitives or attribute data |
| 13:43.40 | starseeker | might be fun, but certainly not a priority |
| 13:46.03 | brlcad | yeah, it looks like they're using their own "AP" set of entity classes, just using the step container format (express) |
| 13:46.27 | brlcad | doesn't look like they map to any AP in 10303 |
| 13:46.27 | starseeker | hah - http://www.iai-tech.org/developers/ifc-implementation/ifc-impl-agreements/cv-2x3-112 |
| 13:46.31 | brlcad | they even cover geometry |
| 13:47.00 | brlcad | yeah |
| 13:47.15 | brlcad | http://en.wikipedia.org/wiki/Industry_Foundation_Classes <-- search for Body |
| 13:47.42 | starseeker | might be a useful format for importing/exporting building descriptions, but probably not much beyond that |
| 13:48.59 | brlcad | wonders why he received five copies of cliff's brlcad-devel mail |
| 13:49.05 | starseeker | O.o |
| 13:49.14 | starseeker | erm. sorry |
| 13:49.21 | brlcad | two of the first, five of the second, seven in all |
| 13:50.41 | starseeker | not sure what happened - did the list get spammed too? |
| 13:52.13 | brlcad | if I got that many copies, probably |
| 13:52.53 | starseeker | growl... sorry |
| 13:53.29 | starseeker | ``Erik: you seeing multiple copies of the emails? |
| 13:53.43 | brlcad | might just have been me or some mixup with sf.net |
| 13:54.05 | brlcad | or maybe just me, who knows.. doesn't matter unless it keeps happening :) |
| 13:54.31 | starseeker | I don't see any extra addys in the gmail header... |
| 13:55.11 | starseeker | heh - well, we'll see if I scared off Ganesh |
| 13:55.15 | brlcad | forum archives only list the messages once (though gets the threading wrong), so maybe just sourceforge accidentally sending out multiple times due to some failure |
| 13:55.54 | starseeker | is rather intrigued by that army.mil document |
| 13:57.07 | starseeker | and their download files for that matter, although I'm not sure how to convert them into somthing usable... |
| 13:58.50 | starseeker | OK, really on the road this time |
| 14:01.10 | brlcad | could get in touch with those guys |
| 14:03.06 | ``Erik | is not seeing duplicates, but entourage may be searching for those |
| 14:57.07 | *** join/#brlcad Zaebos (~irc@pd95b7f5e.dip0.t-ipconnect.de) |
| 15:08.28 | brlcad | finds david's draft and forwards |
| 15:10.17 | brlcad | wonders when tkhtml will finally get put to bed |
| 15:10.27 | CIA-2 | BRL-CAD: 03brlcad * r40928 10/brlcad/trunk/src/other/tkhtml/Makefile.am: missing line-continuation slash causing MISSING FROM DIST errors |
| 15:11.13 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) |
| 15:44.32 | CIA-2 | BRL-CAD: 03starseeker * r40929 10/brlcad/branches/cmake/ (CMakeLists.txt src/other/openNURBS/CMakeLists.txt): Add some more checks for header files. |
| 16:47.06 | CIA-2 | BRL-CAD: 03starseeker * r40930 10/brlcad/branches/cmake/ (6 files in 4 dirs): Cleanup, add in custom check for ALLOCA that mimics the autoconf test, check for dtrace header if that option is enabled. |
| 16:56.19 | *** join/#brlcad merzo (~merzo@11-19-132-95.pool.ukrtel.net) |
| 17:02.58 | CIA-2 | BRL-CAD: 03bob1961 * r40931 10/brlcad/trunk/src/libged/rt.c: Keith found a breakage while using preview (the gd_rt_cmd struct was referring to memory locations that were no longer valid due to a call to zap). This fixes it. |
| 17:17.10 | CIA-2 | BRL-CAD: 03starseeker * r40932 10/brlcad/branches/cmake/ (CMakeLists.txt misc/CMake/BRLCAD_CheckFunctions.cmake): |
| 17:17.10 | CIA-2 | BRL-CAD: add a few defines present in the autotools build - getting closer. Lot of |
| 17:17.10 | CIA-2 | BRL-CAD: specialized AC macros to duplicate for types, although AC_C_CHAR_UNSIGNED is |
| 17:17.10 | CIA-2 | BRL-CAD: considered obsolete and won't be ported - only tcl/tk seem to use it in our |
| 17:17.11 | CIA-2 | BRL-CAD: code, and their build systems can handle it. |
| 17:34.33 | CIA-2 | BRL-CAD: 03starseeker * r40933 10/brlcad/branches/cmake/ (CMakeLists.txt misc/CMake/BRLCAD_CheckFunctions.cmake): Check for members in structures |
| 17:56.42 | CIA-2 | BRL-CAD: 03brlcad * r40934 10/brlcad/trunk/src/libged/rt.c: if the else is wrong, there's no point in keeping it around. |
| 17:57.59 | CIA-2 | BRL-CAD: 03brlcad * r40935 10/brlcad/trunk/src/libged/rt.c: ws |
| 18:02.33 | CIA-2 | BRL-CAD: 03brlcad * r40936 10/brlcad/trunk/NEWS: keith and bob fixed a bug in mged where it was crashing while doing a preview. |
| 18:11.59 | CIA-2 | BRL-CAD: 03indianlarry * r40937 10/brlcad/trunk/src/rt/view.c: Set variable 'scanline' to NULL after free. If not NULL 'rt' would try and release a second time. Problematic when raytracing multiple frames. |
| 18:29.53 | CIA-2 | BRL-CAD: 03starseeker * r40938 10/brlcad/branches/cmake/ (5 files in 4 dirs): Add some more type checks, try to perform the YYTEXT_POINTER test for lex. |
| 18:45.42 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) |
| 19:12.37 | CIA-2 | BRL-CAD: 03brlcad * r40939 10/brlcad/trunk/NEWS: |
| 19:12.37 | CIA-2 | BRL-CAD: keith fixed a bug in the raytracers where they were crashing if you tried to |
| 19:12.37 | CIA-2 | BRL-CAD: render multiple frames (e.g., via -M script). the problem was freeing a |
| 19:12.37 | CIA-2 | BRL-CAD: scanline but not setting it to null afterwards, causing it to free erroneously |
| 19:12.37 | CIA-2 | BRL-CAD: later. |
| 19:33.28 | *** join/#brlcad Ralith (~ralith@d142-058-092-067.wireless.sfu.ca) |
| 19:55.57 | *** join/#brlcad R0b0t1 (~Enigma@64-136-219-55.dyn.everestkc.net) |
| 19:55.57 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) |
| 20:00.31 | *** join/#brlcad Zaebos_ (~irc@pd95b7f5e.dip0.t-ipconnect.de) |
| 20:01.50 | starseeker | ``Erik: aw, great - it seems to be looking for the Tcl source dir to get the private headers, and of course that doesn't exist since it's a /usr/tmp/ports/... path |
| 20:03.58 | starseeker | ``Erik: why does the BSD portage install assign that variable if it won't exist after the portage process is finished? |
| 20:05.07 | ``Erik | it defines it? O.o |
| 20:05.21 | ``Erik | (and it's called ports, portage is what ricer weenies call their bad clone) |
| 20:05.30 | starseeker | less /usr/local/lib/tcl8.5/tclConfig.sh |
| 20:06.02 | ``Erik | ah, probably just what tcl does and it's never broken for anyone (or other people worked around the breakage) |
| 20:06.34 | starseeker | nods - well, a vanilla itcl build chokes I can tell you that |
| 20:07.33 | *** join/#brlcad cosurgi (~cosurgi@atak.bl.pg.gda.pl) |
| 20:07.44 | ``Erik | huh, there's a /usr/ports/lang/itcl |
| 20:07.48 | starseeker | 'course, it's itcl's fault for using private headers in the first place, but it looks like I'm gonna have to validate that directory in FindTCL and turn on the local tcl/tk build if it's not there |
| 20:07.53 | ``Erik | wonder who mai... aw feck, me again. damnit. |
| 20:08.12 | starseeker | ``Erik: you could install that of course - that would avoid my building it, but it wouldn't fix the problem |
| 20:08.51 | ``Erik | like like there was stuff to work around that someone did before I adopted it |
| 20:29.34 | *** join/#brlcad merzo (~merzo@11-19-132-95.pool.ukrtel.net) |
| 21:23.17 | *** join/#brlcad Ralith (~ralith@d142-058-092-067.wireless.sfu.ca) |
| 21:24.30 | *** join/#brlcad printbot (~Paulh@cpc1-roch3-0-0-cust785.10-1.cable.virginmedia.com) |
| 21:27.32 | *** part/#brlcad printbot (~Paulh@cpc1-roch3-0-0-cust785.10-1.cable.virginmedia.com) |
| 22:33.13 | CIA-2 | BRL-CAD: 03starseeker * r40940 10/brlcad/branches/cmake/src/other/ (3 files in 3 dirs): (log message trimmed) |
| 22:33.13 | CIA-2 | BRL-CAD: This appears to be about the minimal change to get this working - IncrTcl's use |
| 22:33.13 | CIA-2 | BRL-CAD: of the private Tcl headers is a severe problem when trying to build using a |
| 22:33.13 | CIA-2 | BRL-CAD: system Tcl/Tk - in practice, it shouldn't work at all. The workaround is to use |
| 22:33.14 | CIA-2 | BRL-CAD: the headers from the local Tcl/Tk when building - this works if the system |
| 22:33.14 | CIA-2 | BRL-CAD: version and local version are compatible. We need 8.5 from the system version |
| 22:33.15 | CIA-2 | BRL-CAD: now too so this should work without needing to add local copies of the 8.4 |
| 22:33.35 | *** join/#brlcad Ralith (~ralith@d142-058-092-067.wireless.sfu.ca) |
| 22:38.59 | CIA-2 | BRL-CAD: 03starseeker * r40941 10/brlcad/branches/cmake/CMakeLists.txt: FreeBSD wants newlines at the end of these files, so go ahead and add them - doesn't break at least on the Mac, need to check Linux |
| 23:11.29 | *** join/#brlcad Ralith (~ralith@d142-058-092-067.wireless.sfu.ca) |
| 23:12.12 | CIA-2 | BRL-CAD: 03starseeker * r40942 10/brlcad/branches/cmake/ (CMakeLists.txt misc/CMake/BRLCAD_CheckFunctions.cmake): We want to check /usr/local - try to define the variables needed |
| 23:30.57 | CIA-2 | BRL-CAD: 03starseeker * r40943 10/brlcad/branches/cmake/src/ (4 files in 4 dirs): Need some more instances of TCL_INCLUDE_PATH. With this change, we have now build successfully on FreeBSD. |
| 23:31.46 | starseeker | ``Erik: you BSD guys are annoying, you know that? :-P |
| 23:31.56 | starseeker | good shakedown though |
| 23:36.34 | starseeker | hmm - not seeing the pix files in srcdir/pix/... |
| 23:37.40 | ``Erik | :D |
| 23:37.45 | ``Erik | you linux guys are wusses |
| 23:37.58 | starseeker | hey, I made it in the end |
| 23:38.12 | starseeker | itcl/itk is the real annoyance |
| 23:39.20 | starseeker | will check on why those scripts are wiping out later... |
| 23:39.35 | ``Erik | yeh, that's the issue... the levels of kluge are staggering...wonder when 86 will be final |
| 23:40.24 | starseeker | dredges scientific notation from his memory to handle the big numbers... |
| 23:41.00 | ``Erik | numbers beyond taking off your shoes to count to? :D |
| 23:41.16 | starseeker | heh - it's Tcl/Tk, whadya you think? |
| 23:41.44 | ``Erik | oh, hah, thought you were off on a tangent :D |
| 23:41.50 | starseeker | first define units of course - years or decades |
| 23:42.03 | starseeker | ah :-P |
| 23:42.29 | ``Erik | at some point, you just stop using number and give it names |
| 23:42.42 | ``Erik | mesazoic, cenezoic, tclazoic |
| 23:42.45 | starseeker | hehe |
| 23:43.20 | starseeker | was actually studying a geologic time chart recently as a way to organize software history - that might be able to handle Tcl/Tk, come to think of it |
| 23:46.04 | *** join/#brlcad Ralith (~ralith@d142-058-092-067.wireless.sfu.ca) |
| 23:46.21 | *** join/#brlcad R0b0t1 (~Enigma@64-136-219-55.dyn.everestkc.net) |
| 23:46.21 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) |
| 23:49.51 | *** join/#brlcad R0b0t1 (~Enigma@64-136-219-55.dyn.everestkc.net) |
| 23:49.51 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) |