| 00:53.36 | Ralith | if you have any ideas on other approaches, I'd love to hear them |
| 01:20.28 | Ralith | brlcad: should I be committing things that amount to "let's see what happens when I comment *this* bit out!" |
| 01:21.40 | ``Erik | if it's just thrash, probably not... if you find a handful that actually make it closer to what you want, probably... |
| 01:22.39 | Ralith | yeah, if it actually makes a noticable improvement I'll certainly be checking that in. |
| 02:21.00 | brlcad | Ralith: if it does anything at all interesting (i.e. worth commenting on even if just mentally to yourself) beyond "well, shit, that didn't do anything" .. then yeah, probably should commit it |
| 02:35.15 | Ralith | it's mostly the "well, shit, that didn't do anything" case I was wondering about. |
| 02:35.18 | Ralith | thanks |
| 03:35.03 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-7.sbndin.btas.verizon.net) |
| 05:10.25 | Ralith | okay, something weird is going on. |
| 05:11.02 | Ralith | I comment out the loadIdentity for the GL texture matrix, and the corner of white goes to a full window of white. |
| 05:11.11 | Ralith | I uncomment it, and it stays that way. |
| 05:11.34 | Ralith | (also, it appears that the white isn't from Qt as hoped; the presence or abscence of the Qt call doesn't effect it. |
| 05:13.53 | Ralith | ) |
| 05:28.38 | *** join/#brlcad stevegt_ (n=stevegt@c-24-130-122-25.hsd1.ca.comcast.net) |
| 06:52.34 | Ralith | and now it's started being a corner again, and I haven't even changed the code. |
| 06:52.35 | Ralith | wth. |
| 06:53.30 | Ralith | it seems to change at random. |
| 06:53.38 | Ralith | (not during a run, though) |
| 07:03.13 | Ralith | contacts the author of the Ogre raw OpenGL rendering code he's been working off of |
| 08:00.16 | CIA-32 | BRL-CAD: 03Ralith 07http://brlcad.org * r1564 10/wiki/User:Ralith: Log for 2009-07-11 |
| 10:48.49 | ``Erik | OpenGL is a state machine, could it be that you were seeing uninitialized state left over on the card (or in the driver)? |
| 11:45.50 | *** join/#brlcad docelic (n=docelic@78.134.197.189) |
| 12:54.00 | *** join/#brlcad docelic (n=docelic@78.134.196.188) |
| 14:09.46 | *** join/#brlcad samrose (n=samrose@24.11.214.181) |
| 14:43.32 | *** join/#brlcad _sushi_ (n=_sushi_@84-72-8-250.dclient.hispeed.ch) |
| 14:54.45 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-7.sbndin.btas.verizon.net) |
| 15:53.18 | *** join/#brlcad Patmcc19 (n=chatzill@71-223-53-70.phnx.qwest.net) |
| 16:35.47 | *** join/#brlcad Axman6_ (n=Axman6@pdpc/supporter/student/Axman6) |
| 17:09.48 | *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6) |
| 18:00.16 | *** join/#brlcad PrezKennedy (i=Matthew@whitecalf.net) |
| 18:40.15 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-7.sbndin.btas.verizon.net) |
| 21:31.18 | Ralith | ``Erik: uninitialized? I don't entirely follow. |
| 21:31.39 | Ralith | I mean, Ogre's doing its rendering, and then Qt's doing its own, and both are meant to operate on a raw context; I would imagine any uninitialized state would be taken care of. |
| 21:31.55 | Ralith | I'm pretty sure the white background is actually a side effect of a 'clear pass' that the reset code does |
| 21:33.36 | *** join/#brlcad docelic (n=docelic@78.134.196.188) |
| 21:33.59 | ``Erik | I d'no how qt and ogre work, I know back in the day, doing straight OGL, if you attached a context and viewed it without drawing to it, you'd get the results of the last render, if you didn't set the matrices, they'd still have the old values (even though the old program stopped and the new one started) |
| 21:34.54 | ``Erik | just like some os's would have arbitrary junk in memory you just malloc'd from some other program that had used that memory before |
| 21:34.58 | Ralith | yeah, there's quite a bit of code in there that goes out of its way to tidy up after Ogre so other rendering can be done, and restore to the state Ogre expects afterwards. |
| 21:35.18 | Ralith | so assuming that code is complete, that shouldn't be an issue. |
| 21:35.19 | ``Erik | *shrug* then that's probably not it, was just an idea :) |
| 21:35.23 | Ralith | yeah |
| 21:35.27 | Ralith | good one |
| 21:35.41 | Ralith | considering that it's such a big issue that it needs explicit handling in the first place |
| 21:43.11 | ``Erik | <-- used to have something of a clue, way back in the day O:-) |
| 21:43.27 | ``Erik | y'know, with the punch cards and magentic drums and audiocouple tty's and all that |
| 21:43.40 | Ralith | hehe |
| 21:43.45 | Ralith | back when you could see the bits with the naked eye? |
| 21:49.50 | ``Erik | shit, we used two pieces of wire and a small magnet to do input |