| 00:38.00 | CIA-38 | BRL-CAD: 03starseeker * r37238 10/brlcad/trunk/src/libfb/if_tk.c: Still nothing like clean or correct, but at least window persists and closes with a close window event. |
| 00:50.38 | CIA-38 | BRL-CAD: 03starseeker * r37239 10/brlcad/trunk/src/libfb/if_tk.c: Ah, that's somewhat better - put Tcl's vwait to work. |
| 00:57.26 | CIA-38 | BRL-CAD: 03starseeker * r37240 10/brlcad/trunk/src/libfb/if_tk.c: throw in a call to destroy, which isn't called in the WM_DELETE_WINDOW binding after bwe overrode it |
| 00:59.22 | starseeker | suspects the tk framebuffer isn't reading memory quite right... wouldn't a pix be the easy case... |
| 01:11.20 | starseeker | interesting... X24 and tk have different pointer values for ifp and pixelp |
| 01:21.26 | starseeker | blinks |
| 01:21.29 | starseeker | uh... |
| 01:51.02 | starseeker | really hopes he doesn't have to go to the extreme of printing pixel values to distinguish between Tk_Photo problems and memory read issues, but may have to face the music... |
| 02:09.05 | starseeker | hmm, very interesting - no difference in the pixelp memory contents |
| 02:11.06 | starseeker | eyes the PhotoImageBlock... |
| 02:36.12 | *** join/#brlcad R0b0t1 (n=Enigma@unaffiliated/r0b0t1) |
| 02:40.48 | *** join/#brlcad akafubu (n=akafubu@unaffiliated/akafubu) |
| 03:08.53 | CIA-38 | BRL-CAD: 03starseeker * r37241 10/brlcad/trunk/src/libfb/if_tk.c: YES! Odd pointer behavior was behind it all - as of this check in first successful pix-fb display and raytrace with tk framebuffer. |
| 03:13.35 | CIA-38 | BRL-CAD: 03starseeker * r37242 10/brlcad/trunk/src/libfb/if_tk.c: Clear out some debugging code. |
| 03:14.01 | starseeker | now we'll see what aqua tk does... |
| 03:14.58 | starseeker | first working tk framebuffer screenshot: http://bzflag.bz/~starseeker/working_tk_framebuffer.png |
| 03:18.56 | starseeker | is 99% sure the in-MGED framebuffer won't be working at all... |
| 03:21.45 | starseeker | aaaand is apparently wrong, at least in the X11 case |
| 03:21.51 | starseeker | <blink> |
| 03:24.15 | starseeker | wonder if maybe it's using the X one there... |
| 03:25.19 | starseeker | must be... still have all those TJM to implement functions... |
| 03:50.25 | starseeker | well, aqua tk framebuffer ran: http://bzflag.bz/~starseeker/all_aqua_dm_fb.png |
| 03:50.47 | starseeker | (and crashed in-MGED, so lotta work to do there yet obviously) |
| 03:50.55 | starseeker | heads home - good note to end on |
| 05:53.05 | *** join/#brlcad louipc (n=louipc@archlinux/trusteduser/louipc) |
| 05:53.10 | brlcad | starseeker: nice progress!~ |
| 10:42.21 | *** join/#brlcad mafm (n=mafm@94.Red-83-49-87.dynamicIP.rima-tde.net) |
| 10:56.57 | *** join/#brlcad docelic (n=docelic@78-2-120-209.adsl.net.t-com.hr) |
| 11:49.28 | d-lo | Mernin all! |
| 12:27.00 | *** join/#brlcad Yoshi47 (n=jan@64.235.102.210) |
| 13:43.18 | *** join/#brlcad Phurl (n=mdupont@ip-81-210-245-60.unitymediagroup.de) |
| 14:12.04 | CIA-38 | BRL-CAD: 03bob1961 * r37243 10/brlcad/trunk/src/libged/bot_dump.c: Modified bot_dump to allow the user to specify combinations as well as bots. If a combination is specified, its tree is walked and any bots found get converted. Also noticed that the rt_db_internal wasn't getting freed. |
| 16:08.22 | starseeker | growl... it's not doing partial refresh in Aqua - just displaying at the end |
| 17:17.16 | brlcad | irpguardian needs to fix his sf.net e-mail |
| 17:19.47 | brlcad | starseeker: you need more than the dooneevent calls |
| 17:19.52 | brlcad | you have to tell the canvas to refersh |
| 17:22.41 | starseeker | brlcad: any idea how specifically to do so? I've been googling but I can't turn anything up so far |
| 17:23.57 | starseeker | i've tried update idletasks |
| 18:18.31 | d-lo | irpguardian wants to know the definition of 'fix his sf.net email' |
| 18:25.36 | ``Erik | unfuck it? |
| 18:25.52 | ``Erik | (is it not forwarding to a valid email? |
| 18:40.32 | *** join/#brlcad manoj (n=manojkmo@117.204.114.52) |
| 18:43.46 | *** part/#brlcad manoj (n=manojkmo@117.204.114.52) |
| 18:47.31 | d-lo | irpguardian reports "email is fixeded" |
| 19:57.18 | brlcad | cool |
| 20:05.44 | brlcad | starseeker: perhaps Tk_CanvasEventuallyRedraw() |
| 20:06.37 | starseeker | will give that a try (currently in the middle of trying 8.5.8 - tcl mac dev said there were some changes made since 8.5.6) |
| 20:06.46 | brlcad | and then call Tcl_DoOneEvent(TCL_ALL_EVENTS); |
| 20:06.50 | brlcad | till there are no more |
| 20:07.02 | brlcad | nods |
| 20:07.37 | starseeker | just about ready to merge in the update, actually - couple files to add to Makefile.am and I can go ahead and upgrade... |
| 20:08.20 | brlcad | you have our build mods? |
| 20:08.31 | starseeker | that's what's been taking all the time :-) |
| 20:08.44 | brlcad | there are a handful made to their unix/Makefile.in |
| 20:08.47 | starseeker | if I was gonna build it, figured to build with our changes |
| 20:08.54 | starseeker | yeah, that was the most complex merge |
| 20:09.25 | brlcad | they applied a couple of our fixes, so some things might not be necessary |
| 20:09.42 | brlcad | would have to dig through their tracker to figure out what, though |
| 20:09.48 | starseeker | nods - yeah, looks like your Haiku fixes made it in too :-) |
| 20:10.21 | starseeker | has slogged through it - should be just about ready aside from adding new files |
| 20:10.32 | brlcad | the patch that extends Tcl_Eval() processing to the extent of 32-bit should be there |
| 20:10.41 | starseeker | yep |
| 20:10.49 | starseeker | you and bob have a changelog credit :-) |
| 20:10.53 | brlcad | they modified it a little, but supports the same end-result |
| 20:10.58 | brlcad | ah |
| 20:11.01 | brlcad | cool |
| 20:12.18 | starseeker | per advice of #tcl guys, emailed Daniel A. Steffen - he is the one who said to be sure I was working with the newest code first |
| 20:13.03 | starseeker | they won't put much/any effort into the old Carbon backend, so I need to check the Cocoa one if I want to really report bugs |
| 20:23.30 | starseeker | isn't quite clear yet if this means he'll have to try 8.6b1... that would be entertaining but might also force the shelving of the Aqua efforts until 8.6 is stable |
| 20:44.05 | CIA-38 | BRL-CAD: 03erikgreenwald * r37244 10/brlcad/trunk/src/librt/ (4 files in 3 dirs): shift marching cube tables out of metaball and into NMG |
| 20:45.27 | starseeker | well... that worked great |
| 20:45.33 | starseeker | now it won't run at all |
| 20:48.09 | ``Erik | "bull in chinashop" debugging? :D |
| 20:48.21 | starseeker | heh |
| 20:49.38 | starseeker | now the Tcl_DoOneEvent call just says Tcl_WaitForEvent: CFRunLoop finished |
| 20:49.43 | starseeker | Abort trap |
| 20:50.02 | starseeker | commenting out the DoOneEvent call results in the same behavior previously observed |
| 20:50.26 | brlcad | starseeker: if 8.6b1 makes it all work and doesn't have noticaeable bugs on linux, that's reason enough to use it |
| 20:50.34 | brlcad | we've used an alpha release before for that same reason |
| 20:50.49 | starseeker | brlcad: won't that involve trying the new itcl/itk stuff too? |
| 20:51.30 | brlcad | didn't say it would be simple .. basically said iff it all works ;) |
| 20:51.35 | starseeker | hehe |
| 20:51.52 | starseeker | ok, that would at least make the devs more likely to help with bugs... |
| 20:52.30 | brlcad | could put it all in a branch |
| 20:52.35 | brlcad | follow it there |
| 20:52.35 | starseeker | nods |
| 20:53.03 | brlcad | that way won't make life hard for our ARL dev-users until it's all working |
| 20:53.11 | brlcad | or whenever we decide to sync |
| 20:53.20 | starseeker | will use the dmtogl branch - that's related in any case |
| 20:54.46 | starseeker | tars up the 8.5.8 work just in case it's ever useful for something... |
| 20:57.54 | *** join/#brlcad talcite (n=matthew@dhcp-143-147.mcme-students.carleton.ca) |
| 20:59.36 | starseeker | raises eyebrow - the current 8.6b1 files are from a year ago? |
| 20:59.45 | starseeker | riiight |
| 21:00.13 | starseeker | hunts dev repositories |
| 21:02.13 | starseeker | ``Erik: why the heck do all my crit files belong to someone called bugQ |
| 21:02.14 | brlcad | sounds about right |
| 21:02.20 | brlcad | hah |
| 21:02.28 | ``Erik | cuz you make lots of bugs |
| 21:02.30 | ``Erik | seemed fitting |
| 21:02.35 | starseeker | heh |
| 21:02.35 | ``Erik | (means a uid conflict) |
| 21:02.55 | ``Erik | I copied the passwd list the other day for d-lo |
| 21:03.06 | brlcad | must have manually made your account |
| 21:03.11 | brlcad | pre syncings |
| 21:03.32 | starseeker | ``Erik made it a couple weeks back, iirc |
| 21:03.39 | brlcad | that'd be why |
| 21:03.56 | brlcad | you got "the next available", which was far behind the current one on .bz |
| 21:03.56 | ``Erik | chownchownchown |
| 21:04.16 | brlcad | now if someone(tm) would just finish up the migration, that wouldn't be a problem |
| 21:04.23 | starseeker | has discovered that large branch syncs should be done on crit or his home box... |
| 21:05.27 | ``Erik | yes, someone should commit a day or 2 and just hammer out the migration... O.o :D |
| 21:05.31 | starseeker | ``Erik: ah, thanks :-) |
| 21:05.36 | brlcad | think networking on .bz has some problem |
| 21:05.54 | brlcad | connections getting throttled or maybe the firewall is just too big and needs flushing |
| 21:06.04 | brlcad | gets a cound |
| 21:06.23 | brlcad | 3k entries.. not "too" badd |
| 21:06.30 | brlcad | but that's a lot of searching per-packet, heh |
| 21:07.33 | brlcad | yeah, it's had 9k before the last time it was flushed |
| 21:14.10 | ``Erik | I'm sure it's a linear search, too |
| 21:25.34 | starseeker | eyes another large sync... |
| 21:25.37 | starseeker | arrgh |
| 21:31.02 | starseeker | syncing STABLE is gonna be epic |
| 21:37.03 | starseeker | Hmm - this might be a decent test for the obj importer: http://www.open3dproject.org/ <evilgrin> |
| 21:40.07 | CIA-38 | BRL-CAD: 03irpguardian * r37245 10/brlcad/trunk/src/rt/view.c: |
| 21:40.07 | CIA-38 | BRL-CAD: Removed some unneeded comments, and started to add a new function that will store |
| 21:40.08 | CIA-38 | BRL-CAD: the value of the times taken to complete each pixel. |
| 21:43.08 | *** join/#brlcad Ralith (n=ralith@69.90.48.97) |
| 21:44.39 | CIA-38 | BRL-CAD: 03starseeker * r37246 10/brlcad/branches/rel8/ (209 files in 45 dirs): Syncing rel8 to trunk r32744 |
| 21:47.20 | CIA-38 | BRL-CAD: 03starseeker * r37247 10/brlcad/branches/dmtogl/ (228 files in 55 dirs): Syncing dmtogl to trunk r32744 |
| 22:08.26 | brlcad | hm, mail still not getting through to irpguardian |
| 22:08.42 | brlcad | or maybe just hasn't had time to sync |
| 22:21.31 | CIA-38 | BRL-CAD: 03erikgreenwald * r37248 10/brlcad/trunk/src/librt/primitives/ (metaball/metaball_tri.c nmg/nmg_tri_mc.c): migrate cube+edgeset into nmg_tri_mc.c, tweak function signature |
| 23:18.55 | *** join/#brlcad Ralith (n=ralith@69.90.48.97) |
| 23:37.45 | starseeker | grrr |
| 23:38.00 | starseeker | 8.6 is going to be a pain |
| 23:54.13 | starseeker | they're doing what appear to be Apple frameworks, which is good, and triggering their own configure stuff with a make call, which is bad |
| 23:54.40 | *** join/#brlcad alex_joni (n=alex_jon@emc/board-of-directors/alexjoni) |
| 23:54.57 | starseeker | gonna have to take a step back and look at the right way to do the tcl/tk build, I'm afraid... |
| 23:55.15 | starseeker | check out this framework thing... |
| 23:55.24 | starseeker | heads home |
