| 00:08.02 | starseeker | and just for more fun: "House Votes To Expand National DNA Arrest Database" |
| 00:08.26 | starseeker | should have stayed off of slashdot today |
| 00:19.10 | starseeker | grumbles... tktable and tktreectrl aren't currently compatible with 8.6 aqua |
| 00:19.17 | starseeker | confound it |
| 01:32.59 | dtidrow | "German High Court Declares All Software Patentable" - morons |
| 01:33.25 | dtidrow | should be 'UNpatentable' |
| 01:33.42 | dtidrow | patenting math is just wrong.... |
| 02:02.39 | CIA-49 | BRL-CAD: 03starseeker * r39447 10/brlcad/trunk/src/isst/isst_tcltk.c: Hmm - freeing this isn't working, need to check on argv_from_string |
| 02:21.09 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) |
| 02:42.54 | *** join/#brlcad Nohla (~jesica@201.255.235.61) |
| 02:54.05 | *** join/#brlcad WhiteCalf (~Prez@2002:601f:5460::601f:5460) |
| 03:15.40 | starseeker | wonders why edit_com doesn't have issues with freeing the results of bu_argv_from_string... hmm... |
| 03:22.04 | brlcad | remember that there is potentially allocation for the array itself as well as for the indiv elements in the array |
| 03:22.37 | brlcad | if the array is on the stack, can't free it -- only free the members within it |
| 03:22.58 | *** join/#brlcad stevegt_ (~stevegt@cislunar.TerraLuna.Org) |
| 03:23.17 | brlcad | pair up your malloc calls with free calls (there's a bu_debug flag you can set that will print them out for you, helps with allocaiton problems) |
| 04:01.08 | *** join/#brlcad stevegt_ (~stevegt@c-69-181-134-76.hsd1.ca.comcast.net) |
| 05:07.21 | *** join/#brlcad IriX64 (~mario_dul@bas2-sudbury98-1177593046.dsl.bell.ca) |
| 05:07.36 | *** part/#brlcad IriX64 (~mario_dul@bas2-sudbury98-1177593046.dsl.bell.ca) |
| 06:29.58 | *** join/#brlcad CIA-40 (cia@208.69.182.149) |
| 07:24.00 | *** join/#brlcad DarkCalf (~Prez@96.31.84.96) |
| 08:18.56 | *** join/#brlcad Nohla (~jesica@201.255.253.132) |
| 10:41.17 | *** join/#brlcad d_rossberg (~rossberg@BZ.BZFLAG.BZ) |
| 10:51.58 | CIA-40 | BRL-CAD: 03erikgreenwald * r39448 10/brlcad/trunk/src/librt/primitives/metaball/metaball.c: make things even uglier |
| 13:22.03 | *** join/#brlcad stevegt_2 (~stevegt@c-69-181-134-76.hsd1.ca.comcast.net) |
| 13:33.24 | starseeker | Well, at least in the case of the -I statements being missing it doesn't look like Togl's fault |
| 13:33.48 | starseeker | it's using our cache, which is not defining ac_x_includes (or rather, is defining it as empty) |
| 13:35.19 | ``Erik | keeps seeing /BinaryCache/... show up :/ |
| 13:36.23 | starseeker | wonders if this is related to that configure.ac line we had to take out because of the Mac png issue |
| 13:39.01 | ``Erik | running configure in src/other/togl, giving it --with-tk=<path> --with-tcl=<path> and --with-x=<path>, it still fails, ignores my X path and uses the stuff it finds elsewhere O.o |
| 13:41.30 | starseeker | yeah, needs -x-includes=/usr/X11R6/include |
| 13:44.36 | ``Erik | same issue heh :D |
| 13:44.57 | starseeker | nods - that sets the header value, but it's not being included as an -I option |
| 13:49.10 | ``Erik | removed /opt/local/bin from the path, purged the caches, it's still not allowing /usr/X11R6/include in O.o |
| 13:49.21 | starseeker | nods |
| 13:49.32 | ``Erik | (and still finding the opt stuff somehow) |
| 13:50.53 | starseeker | ``Erik: I don't suppose symlinking X11R6 to X11 gets us anywhere? |
| 13:51.11 | ``Erik | ahhhhh, tkConfig.sh is overriding configure |
| 13:51.25 | ``Erik | have that symlink already |
| 13:52.00 | ``Erik | grumbles and installs mesa O.o |
| 13:55.50 | ``Erik | yup, that allows it to build |
| 13:58.52 | ``Erik | doesn't explain that other mac, though :/ unless togl/x11 requires mesa headers and pukes on apples version of X opengl stuff O.o |
| 13:59.07 | starseeker | that could be it |
| 14:01.03 | starseeker | braces himself and tries an aqua build |
| 14:01.25 | ``Erik | hah |
| 14:01.34 | ``Erik | Error in startup script: couldn't load file "/usr/brlcad/HEAD_32/lib/isst/issttcltk.plugin": dlopen(/usr/brlcad/HEAD_32/lib/isst/issttcltk.plugin, 10): Symbol not found: _Tcl_PkgInitStubsCheck |
| 14:01.41 | ``Erik | "load /usr/brlcad/HEAD_32/lib/isst/issttcltk.plugin isst" |
| 14:01.48 | ``Erik | $ ls /usr/brlcad/HEAD_32/lib/isst |
| 14:01.48 | ``Erik | issttcltk.0.plugin issttcltk.a issttcltk.la issttcltk.plugin pkgIndex.tcl |
| 14:01.57 | ``Erik | got an extra .0 in the filename |
| 14:02.06 | starseeker | O.o |
| 14:02.19 | ``Erik | apple makes libblah.0.dylib instead of libblah.so.0 |
| 14:03.12 | ``Erik | at this rate, you're gonna have the same appreciate of libtool that the tcl dudes do :> |
| 14:03.27 | starseeker | heh |
| 14:03.59 | starseeker | does your little sh script give a useful result? |
| 14:04.11 | ``Erik | hm, there's a symlink though |
| 14:04.14 | starseeker | (as opposed to forcing the plugin extension?) |
| 14:04.39 | ``Erik | ohhhhhhh, wait |
| 14:04.43 | ``Erik | n/m, that's not hte issue |
| 14:06.03 | ``Erik | it's trying to use system tcl84 |
| 14:06.10 | starseeker | Symbol not found: _Tcl_PkgInitStubsCheck almost seems like it's not getting something it needs from Tcl... |
| 14:06.14 | starseeker | ah |
| 14:06.18 | starseeker | phooey |
| 14:08.57 | *** join/#brlcad kil-9 (~user@cpe-065-191-162-160.nc.res.rr.com) |
| 14:35.05 | starseeker | glares at TEA |
| 14:35.45 | starseeker | it's days like today that make me want to finish that complete redo of the tcl/tk build logic |
| 15:00.49 | CIA-40 | BRL-CAD: 03brlcad * r39449 10/brlcad/trunk/TODO: collecting usage statistics. |
| 15:05.46 | CIA-40 | BRL-CAD: 03erikgreenwald * r39450 10/brlcad/trunk/ (7 files in 4 dirs): move the tcl/tk/togl isst from src/isst to src/adrt |
| 17:00.32 | CIA-40 | BRL-CAD: 03erikgreenwald * r39451 10/brlcad/trunk/m4/ (Makefile.am OpenGL.m4): add OpenGL searching shtuff from gamem4 |
| 17:03.51 | starseeker | ``Erik: yeah, here's the other one: |
| 17:03.54 | starseeker | http://ac-archive.sourceforge.net/ac-archive/ax_check_gl.html |
| 17:05.00 | starseeker | there's a third, but it's GPL |
| 17:07.30 | ``Erik | hm, the ax one seems to be GPL v2 |
| 17:07.48 | starseeker | hmm? |
| 17:07.54 | ``Erik | http://code.google.com/p/autoconf-gl-macros/ |
| 17:08.39 | starseeker | erum |
| 17:09.14 | starseeker | those are updated |
| 17:09.27 | ``Erik | <-- would assume that braden mcdaniel is 'more correct' about the license of braden mcdaniel's source than some aggregator O.o :) |
| 17:09.29 | starseeker | whch of course would be ideal... |
| 17:09.38 | starseeker | nods |
| 17:09.52 | starseeker | worth emailing him to see if LGPL would work? |
| 17:10.32 | starseeker | looks like he's put some thought into this |
| 17:10.35 | ``Erik | mebbe *shrug* |
| 17:10.44 | starseeker | ah well, email's cheap |
| 17:10.47 | starseeker | fires away |
| 17:10.58 | ``Erik | might wanna ask him if the version on that site is actually supposed to be mit |
| 17:11.30 | ``Erik | my gut feeling would be that the aggregator assumed the license they liked since braden didn't put it in the file itself... |
| 17:11.31 | starseeker | right - but even if it is, we'd want to either do it ourselves or be able to use current stuff from someone else |
| 17:11.53 | ``Erik | so letting him know it's out there might be a nice good faith gesture |
| 17:12.04 | starseeker | nods |
| 17:13.15 | ``Erik | the ax one looks like it requires some other m4 files, too |
| 17:26.26 | *** join/#brlcad stevegt_1 (~stevegt@cislunar.TerraLuna.Org) |
| 17:43.34 | starseeker | huh - same guy who does openvrml |
| 17:51.25 | ``Erik | damn those people who do several some-what related things O.o *duck* |
| 17:51.32 | starseeker | heh |
| 17:51.49 | starseeker | just amused - I had early stumbled onto openvrml |
| 17:56.51 | ``Erik | interesting, I try to load a model using the isst.tcl and get a buttload of db_dircheck duplicate warnings and 0 triangles loaded O.o (geometry works fine in the sdl version) |
| 18:06.41 | starseeker | ``Erik: I'm betting it's related to argv somehow - try putting in just a straight string name |
| 18:10.13 | ``Erik | yes |
| 18:10.21 | starseeker | that worked? |
| 18:10.24 | ``Erik | bu_argv_from_string returns 0 and does not fill |
| 18:10.35 | starseeker | crap |
| 18:10.56 | ``Erik | looks like argv might need to be malloc'd before calling that func? |
| 18:11.04 | starseeker | I thought I was... |
| 18:11.26 | ``Erik | or pass &argv, lemme look |
| 18:12.38 | starseeker | hmm - I'm not mallocing, odd... |
| 18:13.01 | ``Erik | yup, that worked |
| 18:13.13 | starseeker | passing &argv? |
| 18:13.16 | ``Erik | no |
| 18:13.30 | starseeker | and I'll bet you can free it now too? |
| 18:13.31 | ``Erik | argv = (char **)malloc(sizeof(char *)*1000); before the argv_from_string |
| 18:13.35 | starseeker | smacks himself |
| 18:14.02 | starseeker | might want to malloc sizeof(char *)*strlen(Tcl_GetString(objv[3]) |
| 18:14.14 | ``Erik | yup |
| 18:14.26 | ``Erik | malloc() right before the load_g and free right after, works fine |
| 18:14.33 | starseeker | ``Erik: can you commit? I'm trying to rebuild |
| 18:14.46 | ``Erik | hrm |
| 18:15.07 | ``Erik | is it appropriate for taht function to fill existing memory, or should it try to allocate appropriately? |
| 18:15.08 | brlcad | "The input buffer is altered by this process. The argv[] array * points into the input buffer. The argv[] array needs to have at * least lim+1 pointers allocated for lim items plus a terminating * pointer to NULL. The input buffer should not be freed until argv * has been freed or passes out of scope. |
| 18:15.41 | starseeker | oh, oops |
| 18:15.53 | starseeker | so we need to make a local string, and use that |
| 18:16.09 | brlcad | and keep it allocated until you're done with the argv |
| 18:16.14 | ``Erik | sizeof(string) isn't right, though |
| 18:16.16 | starseeker | got it |
| 18:16.54 | starseeker | ``Erik: uh, right - that's why I said strlen |
| 18:17.06 | ``Erik | "a b c" would need 4 pointers, 16 or 32 bytes, but strlen is 5 |
| 18:17.15 | ``Erik | er, strlen I mean |
| 18:17.35 | brlcad | when I started implementing the encode/decode functions, I changed bu_argv_from_string to make a copy of your input string so it wasn't dependent, but that was all reverted when decode wasn't working right |
| 18:17.45 | starseeker | as long as we're slightly larger, it shouldn't matter should it? |
| 18:17.49 | ``Erik | sizeo(void *)*(strlen(string)+1) if all the names are single character, but that'd allocate way too much for 'normal' ones |
| 18:18.25 | starseeker | well, anything else is going to need to pre-process the string |
| 18:18.44 | ``Erik | yeh, that's why I mentioned something about allocating appropriately :D |
| 18:19.42 | starseeker | will over-allocation of that particular string ever be likely to cause problems? |
| 18:20.51 | ``Erik | probably not |
| 18:20.58 | CIA-40 | BRL-CAD: 03erikgreenwald * r39452 10/brlcad/trunk/src/adrt/isst_tcltk.c: allocate memory before filling |
| 18:21.42 | ``Erik | hrm, that's without the [split $args] though |
| 18:22.02 | starseeker | shrugs - try it quick and see if we need to pre-split it or not |
| 18:22.16 | ``Erik | imagines removing the bu_ func and using [split] is what we really want, no? |
| 18:22.45 | brlcad | is mildly intrigued by the tesla toyota partership .. hopefully the affordability and engineering of toyota with the look and power of tesla/lotus |
| 18:23.03 | starseeker | well, you can either pre-split and manually create an argv with the objv[n] entries, or do it this way |
| 18:23.09 | brlcad | and not the look and power of toyota with the tesla pricetag :) |
| 18:23.12 | starseeker | I'm guessing potato/potato |
| 18:24.00 | ``Erik | s/power/lackofweight/ :D *duck* |
| 18:26.05 | ``Erik | still looking at getting a turbo? |
| 18:26.23 | brlcad | as power == work / time == force * displacement / time == mass * acceleration * displacement / time ... sure, reducing mass and/or displacement, resulting in better acceleration and more power :P |
| 18:27.23 | brlcad | I'd like to, it's not any higher on my priority list at the moment .. still have major house projects to attend to first |
| 18:27.49 | ``Erik | hrm |
| 18:30.51 | ``Erik | interesting |
| 18:31.15 | starseeker | libtool: link: cannot find the library `../../src/adrt/libtie.la' or unhandled argument `../../src/adrt/libtie.la' |
| 18:32.12 | starseeker | may have been caused by parallel make... |
| 18:32.41 | brlcad | find . -name \*.la -exec rm {} \; && make |
| 18:33.13 | starseeker | brlcad: I just typed make again and it contined... |
| 18:33.32 | brlcad | k |
| 18:33.40 | brlcad | I'd assumed you'd already tried that |
| 18:33.51 | starseeker | no - just noting that something needs fixing in the make logic |
| 18:34.14 | starseeker | ``Erik moved isst back into src/adrt |
| 18:40.27 | starseeker | ``Erik: why not bu_malloc/bu_free? |
| 18:40.46 | ``Erik | cuz my fingers aren't used to bu_ there |
| 18:41.06 | starseeker | ah |
| 18:41.08 | ``Erik | is tracking that dup thing, will change once he gets this file back to normalish |
| 18:41.11 | starseeker | no particular reason? |
| 18:41.12 | ``Erik | *shrug* |
| 18:41.23 | starseeker | np |
| 18:41.40 | starseeker | every once in a while you need raw malloc/free, just wasn't sure if this was one of those cases |
| 18:45.24 | ``Erik | does tcl have (map) ? |
| 18:45.34 | starseeker | urm |
| 18:45.37 | starseeker | dunno |
| 18:45.49 | ``Erik | cuz that's what you're faking with that foreach :/ |
| 18:46.56 | starseeker | looks |
| 18:47.40 | ``Erik | (load_g (lambda (x) (get listwidget x)) (curselection listwidget)) or something |
| 18:47.52 | ``Erik | knowwhutahmean,vern? |
| 18:48.17 | ``Erik | isst.asd ftw! *cough* :D |
| 18:48.45 | ``Erik | and if the ::isst::treeview stuff is going to replace it, is it worth effort at the moment? hrmmm |
| 18:50.34 | starseeker | tree building will be a bit more complicated |
| 18:51.27 | starseeker | don't see a proper map as such |
| 18:52.14 | starseeker | just some proc definitions using foreach |
| 18:52.37 | starseeker | you might be able to do one using the C api... |
| 18:53.16 | ``Erik | OH! |
| 18:53.18 | ``Erik | hah |
| 18:53.43 | ``Erik | you open the file to extract the names for the dialog, don't close it, then open it with load_g |
| 18:54.06 | starseeker | oooops |
| 18:54.10 | starseeker | sorry |
| 18:54.20 | starseeker | that's the problem with quick hacks |
| 18:56.00 | CIA-40 | BRL-CAD: 03erikgreenwald * r39453 10/brlcad/trunk/src/adrt/isst_tcltk.c: build char** argv list from tcl objv |
| 18:57.20 | CIA-40 | BRL-CAD: 03erikgreenwald * r39454 10/brlcad/trunk/src/adrt/isst_tcltk.c: close db after reading the names |
| 18:57.49 | ``Erik | hrm, split may not be right O.o |
| 18:58.18 | starseeker | HAH! sweet - it's working with aqua |
| 18:58.33 | starseeker | including real fullscreen |
| 18:58.54 | starseeker | and mouse button 2 is now mouse button 3 |
| 18:58.56 | starseeker | growl |
| 18:59.25 | ``Erik | yeah, everyone has their own notion of which button is which # :/ |
| 19:51.38 | CIA-40 | BRL-CAD: 03starseeker * r39455 10/brlcad/trunk/src/adrt/isst_tcltk.c: Put the working argv build back in for now. |
| 20:00.10 | *** join/#brlcad akafubu (~akafubu@unaffiliated/akafubu) |
| 20:11.05 | *** join/#brlcad akafubu (~akafubu@unaffiliated/akafubu) |
| 20:36.27 | ``Erik | heh |
| 20:38.11 | ``Erik | mebbe something like, uh, "load_g $oglwin $filename [llength [split $selectedobjs]] $selectedobjs" ? |
| 20:39.15 | ``Erik | or we could break up load_g into several functions, then foreach to the 'add this name' bit, might be useful for adding/removing geometry from the engine in the long run O.o |
| 21:06.00 | *** join/#brlcad akafubu (~akafubu@unaffiliated/akafubu) |
| 21:27.05 | *** join/#brlcad akafubu (~akafubu@unaffiliated/akafubu) |
| 21:33.23 | *** join/#brlcad akafubu (~akafubu@unaffiliated/akafubu) |
| 21:38.09 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) |