| 00:14.10 | CIA-40 | BRL-CAD: 03starseeker * r39481 10/brlcad/branches/dmtogl/src/mged/dm-togl.c: doing anything with the interp after doing dm_open with togl is causing memory issues - not clear why yet. |
| 00:22.03 | ``Erik | redrum O.o |
| 00:28.24 | starseeker | why does gdb refuse to linebreak??? |
| 00:31.41 | ``Erik | cuz you screwed it up with some term goobiness |
| 00:31.43 | ``Erik | ? |
| 00:32.02 | ``Erik | lemme guess, mged? |
| 00:32.06 | starseeker | yeah |
| 00:32.26 | ``Erik | it does some oddness on startup, makes it so'z ya can't even ^Z it |
| 00:32.51 | ``Erik | (-c still works sane iirc) |
| 00:33.46 | ``Erik | mmmm, pastrami and sauerkraut |
| 00:34.05 | starseeker | god this is strange |
| 00:34.23 | starseeker | I'm getting a memory error, but it's not associated with any variable access I can identify |
| 00:34.56 | ``Erik | heh, doing what? |
| 00:35.03 | poolio | ahoy |
| 00:35.05 | starseeker | Tcl_Eval |
| 00:35.15 | ``Erik | and; mac or linux (or bsd)? bus fault or segment fault? |
| 00:35.37 | starseeker | mac, EXC_BAD_ACCESS |
| 00:35.43 | starseeker | KERN_PROTECTION_FAILURE |
| 00:36.01 | ``Erik | not a debug build? is backtrace sane? |
| 00:36.08 | starseeker | Bus error |
| 00:36.09 | ``Erik | yo, hoolio |
| 00:36.16 | starseeker | debug build (or so I thought) |
| 00:36.34 | ``Erik | mac'll busfault on a stack smash |
| 00:36.37 | starseeker | I get a backtrace alright, but it does me no good |
| 00:36.52 | ``Erik | well, ... how many thousand calls deep is the bt? :D |
| 00:37.05 | starseeker | hang on... |
| 00:37.25 | starseeker | 36 deep |
| 00:37.33 | ``Erik | I think I got a stack smash bus fault at around 3k earlier today |
| 00:37.35 | ``Erik | huh |
| 00:38.03 | ``Erik | that's nothing O.o are the call variables filled in the bt? |
| 00:38.25 | starseeker | seem to be |
| 00:38.30 | CIA-40 | BRL-CAD: 03starseeker * r39482 10/brlcad/branches/dmtogl/src/mged/dm-togl.c: Break the offending line out into its own function to make sure it's isolated in a breakpoint sense - doesn't help functionally. |
| 00:38.31 | starseeker | hang on, I'll do a pastebin |
| 00:38.48 | ``Erik | wait until tomorrow, rest will make it obvious? ;) |
| 00:38.59 | starseeker | no, this has been a problem from the beginning |
| 00:39.14 | starseeker | I just commented it out, but right now it's the stopper for the non-classic mged load process |
| 00:39.23 | starseeker | so I have to uncomment it and get it working |
| 00:40.33 | ``Erik | <-- gonna assume tcl eval takes an environment and a string, which kinda narrows things down a bit O.o :D |
| 00:41.55 | starseeker | http://pastebin.org/286940 |
| 00:42.50 | ``Erik | doesn't really look like -ggstabs3 output |
| 00:43.06 | ``Erik | and it looks like ya optimized it |
| 00:43.16 | ``Erik | which'll throw the line #'s off |
| 00:43.22 | starseeker | shakes head - not intentionally |
| 00:43.40 | ``Erik | can ya pring the stirng passed to the togl cmd? |
| 00:44.16 | starseeker | sure, one sec... |
| 00:46.32 | starseeker | package require Togl; togl .topid_0.ul.togl -width 1280 -height 1280 -rgba true -double true |
| 00:46.59 | starseeker | followed by a pack command, then just the window name to allow capture of the togl TclObj |
| 00:47.45 | ``Erik | and only in the branch, I assume? O.o |
| 00:47.53 | starseeker | yes |
| 00:48.22 | starseeker | (this is where being able to experiment from a bwish command line would be nice...) |
| 00:48.29 | ``Erik | perhaps break the various tcl cmds into seperate tcl eval calls? |
| 00:48.53 | starseeker | maybe... I'm not having any problems in the togl_open function itself tough |
| 00:48.55 | starseeker | though |
| 00:49.36 | starseeker | the same command will fail in Togl_dm_init and succeed in togl_open |
| 00:50.10 | ``Erik | threading issue, mebbe? |
| 00:50.23 | starseeker | maybe, but this is an unthreaded tcl/tk build |
| 00:50.33 | starseeker | (unless the aqua stuff turned it on) |
| 00:50.44 | starseeker | that could be |
| 00:50.45 | ``Erik | is the mged/libdm stuff trying to make calls into togl from different threads, though? |
| 00:50.49 | starseeker | tries it on linux |
| 00:51.01 | starseeker | possibly... |
| 00:51.04 | starseeker | dunno |
| 00:51.12 | ``Erik | <-- grasping, hasn't looked at the code and doesn't do dm... :D |
| 00:52.50 | starseeker | heh |
| 00:53.23 | starseeker | the tcl guys would say this is a really backwards way to approach the problem, most likely |
| 00:54.02 | starseeker | instead of doing it in tcl and binding things to c functions to do dirty work (like isst.tcl) we're doing it from the C side and trying to hook into the togl guts |
| 00:55.37 | ``Erik | heh, yeah... trying to treat it like a scripting system instead of the main driver of the program, for shame O.o swig has that issue, too, from what I've seen |
| 00:55.47 | ``Erik | thus my jabbering about cffi instead of swig :D |
| 00:57.31 | starseeker | eyes mged.c... |
| 00:58.01 | starseeker | ``Erik: problem is, we'd never be able to sell cffi as a mainstream solution |
| 00:59.05 | starseeker | god I'm tempted to see if I can get mged up inside bwish without libdm and libfb... |
| 00:59.16 | starseeker | or libdm anyway |
| 01:00.23 | starseeker | problem is i'd undoubtedly be throwing away effort :-( |
| 01:01.33 | starseeker | not to mention all the subtle little things I'm forgetting about |
| 01:05.10 | starseeker | ``Erik: 'course, if we grab the ecl code... |
| 01:05.17 | starseeker | that's in C, isn't it? |
| 01:06.37 | ``Erik | hehehe, think so, didn't look too hard at it |
| 01:07.22 | starseeker | that's designed to be embeddable |
| 01:09.47 | ``Erik | heh, thus the name O.o yeah, I started looking that direction, then figured save-lisp-and-die with a cffi loaded megalibrary would suit my needs *shrug* |
| 01:10.36 | ``Erik | a 28 meg executable might not be totally unreasonable for what I wanna do |
| 01:10.53 | starseeker | auuuugh - Linux fails differently |
| 01:11.18 | ``Erik | damn, that's one for the quote page O.o |
| 01:11.31 | ``Erik | without context, of course |
| 01:13.30 | starseeker | <snort> standard practice for the media - strip the context, it's more interesting that way |
| 01:13.58 | starseeker | concedes he is unlikely to solve this at 9:15pm at night, and starts wrapping it up for the day |
| 01:17.01 | *** join/#brlcad velociostrich (~nsd@c-68-37-119-2.hsd1.nj.comcast.net) | |
| 01:26.09 | poolio | starseeker: heh, this is prime time for coding! |
| 02:05.23 | ``Erik | *snrkt* http://www.collegehumor.com/picture:1938345 |
| 03:18.42 | starseeker | ggrrrr - src/adrt is still busted |
| 03:26.23 | starseeker | oh, nevermind - in dmtogl branch |
| 03:26.29 | starseeker | durh |
| 03:59.58 | starseeker | huh - on Linux, it's actually closer to working |
| 04:00.09 | starseeker | what in the heck is the Mac's problem? |
| 04:43.56 | *** join/#brlcad WhiteCalf (~Prez@96.31.84.96) | |
| 04:53.27 | *** join/#brlcad stevegt_ (~stevegt@c-69-181-134-76.hsd1.ca.comcast.net) | |
| 07:10.33 | *** join/#brlcad alex_joni (~alex_joni@emc/board-of-directors/alexjoni) | |
| 13:33.57 | *** join/#brlcad stevegt_ (~stevegt@c-69-181-134-76.hsd1.ca.comcast.net) | |
| 13:36.09 | ``Erik | http://brlcad.org/~erik/mosstoon.png |
| 15:12.30 | *** part/#brlcad cjdevlin (~devlin@99-74-181-148.lightspeed.cicril.sbcglobal.net) | |
| 16:48.25 | *** join/#brlcad Stattrav (~Stattrav@110.224.54.253) | |
| 17:24.20 | poolio | any of you guys know if there's a way to disable the option+key special characters in os x? |
| 17:28.13 | ``Erik | not off the top of my head... emacs issue? |
| 17:28.39 | ``Erik | <- has been using the escape meta approach in emacs use... "<esc> x" instead of "meta-x" |
| 17:28.57 | ``Erik | release the escape before hitting x, ... |
| 17:33.47 | poolio | yeah...it's actually not an emacs issue. I'm using matlab and it crashes whenever I hit option+any key :( |
| 17:33.56 | ``Erik | ah, wow, lame |
| 17:34.27 | poolio | And I've got synergy set up so I use option+keys all the time on one monitor, but if I hit it when in the other it goes poof |
| 17:34.28 | ``Erik | they set up to by utf-7 only? O.o |
| 17:36.31 | poolio | Well, it looks like that may be the case for the "use inside terminal mode" but when using their desktop it works fine...looks like bugreport time |
| 18:03.38 | CIA-40 | BRL-CAD: 03starseeker * r39483 10/brlcad/branches/dmtogl/ (125 files in 20 dirs): Update dmtogl branch to trunk r39482 |
| 18:36.41 | *** join/#brlcad Yoshi47 (~jan@64.235.102.210) | |
| 18:43.34 | *** join/#brlcad Ralith (~ralith@216.162.199.202) | |
| 19:23.21 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) | |
| 19:32.08 | *** join/#brlcad packrat (~packrator@99-67-225-40.lightspeed.livnmi.sbcglobal.net) | |
| 20:07.03 | starseeker | hrm... might trace back to curr_dm_list |
| 20:07.31 | CIA-40 | BRL-CAD: 03erikgreenwald * r39484 10/brlcad/trunk/src/liboptical/ (Makefile.am init.c sh_toon.c): beginning of a "toon" shader. |
| 20:34.42 | starseeker | Woooo-hoooooo! http://www.stixfonts.org |
| 20:34.51 | starseeker | They released 1.0! |
| 20:41.27 | *** join/#brlcad Nohla (~jesica@201.255.242.92) | |
| 21:01.43 | Ralith | starseeker: what exactly is that? The next computer modern? |
| 21:15.55 | starseeker | Ralith: it's a very large font designed to properly support mathematical display |
| 21:16.07 | starseeker | so MathML might actually go somewhere now |
| 21:25.35 | Ralith | starseeker: I wasn't aware that current LaTeX fonts were omissive in that respect. |
| 21:35.30 | starseeker | they don't work for web browsers |
| 21:35.56 | starseeker | metafont is rather different from the TrueType/OpenType/*Type systems |
| 21:35.59 | CIA-40 | BRL-CAD: 03r_weiss * r39485 10/brlcad/trunk/src/conv/obj-g_new.c: adding ability to process command line options from user |
| 21:48.36 | *** join/#brlcad stevegt_ (~stevegt@cislunar.TerraLuna.Org) | |
| 22:06.53 | Ralith | fair enough |
| 22:07.06 | Ralith | I didn't realize that opentype was capable of doing the things that metafont does for math, though. |
| 22:07.39 | Ralith | it does some impressive language stuff, yes, but I haven't seen anything that screws around vertically like math notation tends to. |
| 22:35.55 | starseeker | Ralith: try installing the stix fonts and viewing this page: https://www.eyeasme.com/Joe/MathML/MathML_browser_test |
| 22:43.23 | Ralith | updates font cache |
| 22:46.17 | Ralith | starseeker: um, those are images |
| 22:46.44 | starseeker | not all of them |
| 22:46.54 | starseeker | there should be three columns |
| 22:46.57 | Ralith | oh, far right |
| 22:47.02 | Ralith | yeah that's not rendering right at all |
| 22:47.18 | Ralith | fittingly, it's the vertical stuff that's most screwed up. |
| 22:47.57 | Ralith | perhaps if I restart chrome |
| 22:48.50 | starseeker | the beta fonts work for me - just installed the 1.0 and things got odd - may have to restart myself |
| 22:48.50 | Ralith | nnnnope. |
| 22:48.55 | starseeker | odd |
| 22:48.58 | Ralith | what browser? |
| 22:49.02 | starseeker | firefox |
| 22:49.14 | starseeker | sighs |
| 22:49.23 | Ralith | all the characters render here, but the typesetting and complex expressions are screwed |
| 22:50.19 | starseeker | weird. Of course, the font might have changed since the beta... |
| 22:51.40 | starseeker | that would be highly ironic... |
| 22:51.42 | Ralith | firefox renders it beautifully |
| 22:51.49 | Ralith | I suspect it's chrome's fault. |
| 22:51.58 | starseeker | could be |
| 22:59.09 | Ralith | WebKit-based browsers such as Chrome and Safari do not currently support MathML.[9] |
| 22:59.13 | Ralith | aw. |
| 23:00.25 | starseeker | https://trac.webkit.org/wiki/MathML |
| 23:01.35 | Ralith | cool |
