[Cobra]

I have the latest drivers for my video card, so it can't be that. I even tried taking out everything related to SCP besides the .exes.

Assert: G3_count == 0
File: c:\fs2_open\code\render\3dsetup.cpp
Line: 304
[This filename points to the location of a file on the computer that built this executable]

Call stack:
------------------------------------------------------------------
    fs2_open_d-20061123.exe 00ab05cb()
    fs2_open_d-20061123.exe 00ab0d6e()
    fs2_open_d-20061123.exe 008a44ed()
    fs2_open_d-20061123.exe 006d9a76()
    fs2_open_d-20061123.exe 006da901()
    fs2_open_d-20061123.exe 006ddc10()
    fs2_open_d-20061123.exe 0088f2c1()
    fs2_open_d-20061123.exe 006df165()
    fs2_open_d-20061123.exe 006df3e9()
    fs2_open_d-20061123.exe 00b52f57()
    kernel32.dll 7c816d4f()
------------------------------------------------------------------

[StratComm]

To clarify, since Cobra's been talking to me about this for a good hour or more, he's trying to run one of the missions from the new Derelict revision and is getting this.  He's also getting something similar under D3D, but this one is from OGL.  Radeon 9250 with the latest Catalysts, etc.  Cobra's system isn't known as a beast by any means, but it should at least be able to process vanilla data.  What I didn't get out of him was which build this came from, though I suspect it is either a CVS build from the last 48 hours of the voice recognition build.  Not that it should matter, that file was last modified 9 months ago.

[WMCoolmon]

I may have just committed a fix to that error. I had it lying around but never committed it.

It'd be great if g3_start_frame and g3_end_frame could be gotten rid of altogether...

[taylor]

I may have just committed a fix to that error.

Technically that doesn't qualify as a "fix".  It qualifies more as a "<-- deleted, this was intended as funny but after re-reading it this seemed unintentionally harsh -->" type of thing.

There is debugging spew for both g3_start_frame and g3_end_frame, you need to figure out why it's actually out of frame there and fix that instead.

[Cobra]

the build i was using was the 20060123 build.

would the '24 have corrected this?

[WMCoolmon]

I may have just committed a fix to that error.

Technically that doesn't qualify as a "fix".  It qualifies more as a "<-- deleted, this was intended as funny but after re-reading it this seemed unintentionally harsh -->" type of thing.

There is debugging spew for both g3_start_frame and g3_end_frame, you need to figure out why it's actually out of frame there and fix that instead.

Because the radar was in another part of the render_frame function, and I moved it so that it would better work with scripting while still being rendered on top of stuff.

[Cobra]

well, that seems to have screwed up my rendering process.

[StratComm]

Cobra, do you use the Orb radar mode?  And have you tried switching it off?

[WMCoolmon]

It could be related to the toggling. What're the symptoms (And does it switch on and off with the feature?)

[Cobra]

yup, when i disabled the 3D Radar i was able to progress. the only thing that bothered me was the extreme slowness of the loading process, and the .ogg playback sounded terrible. lots of stuttering and a couple of times where the file sounded like it was being split in half, hearing two voices, one slightly higher than the other.

yesterday, i experienced a fighter06.pof IBX Cache error. something about it not having enough security or something. would this be the cause of the slowness of loading?

[EDIT] damn that redmenace, posting a new build that works better. but the .ogg playback still has some stuttering.

also, i got this:

Warning: EFF: No frame images were found.  EFF, fighter2t-01a-glow.eff, is invalid.

File:c:\fs2_open\code\bmpman\bmpman.cpp
Line: 1761
[This filename points to the location of a file on the computer that built this executable]

Call stack:
------------------------------------------------------------------
    fs2_open_d-20060126.exe 00877aaf()
    fs2_open_d-20060126.exe 00875ac7()
    fs2_open_d-20060126.exe 008781f4()
    fs2_open_d-20060126.exe 00800c12()
    fs2_open_d-20060126.exe 006e7165()
    fs2_open_d-20060126.exe 006e7018()
    fs2_open_d-20060126.exe 006eb25d()
    fs2_open_d-20060126.exe 006eba3a()
    fs2_open_d-20060126.exe 006f0291()
    fs2_open_d-20060126.exe 006f15a2()
    fs2_open_d-20060126.exe 008b40df()
    fs2_open_d-20060126.exe 006cfb4b()
    fs2_open_d-20060126.exe 006dd112()
    fs2_open_d-20060126.exe 0088ed3d()
    fs2_open_d-20060126.exe 006dba5c()
------------------------------------------------------------------

while i have mv_textures 'n junk.

[StratComm]

The amount of media is going to make mission load slow, period.  Derelict uses lots of ship types in missions so it's going to take even longer than you would expect looking at the retail campaign.  Don't know what the fighter06.ibx cache error was, but if it's more than a 1-time thing just make sure you delete fighter06.ibx from both Freespace2/data/cache and Freespace2/derelict/data/cache.

[Cobra]

check the edit.

i don't have the problem anymore, i just ran the debug.

[StratComm]

I've yet to trace the source of the Pegasus glowmap warning, but it should be harmless.  I got it but the pegasus and its glowmap still came through just fine (and that eff was only there to override the animated glowmap, if anyone had it installed).  It is almost certainly non-fatal, so just press ignore and it should be fine.

[WMCoolmon]

There's a .eff file in mv_models for the Pegasus, to override the .ANI glowmap in mv_adveffects.

[StratComm]

I know, I'm the one that made it, but it points to the static glowmap so it shouldn't be throwing an error.  It should be treating it as a single-frame animation.  Unless that's not allowed, of course.

EDIT: Speaking of which, the animated glowmap should probably be pulled from adveffects now, since it's sort of obsolete.  I'd like to get the effect back with the new glowmap, but I'm not good at the 2d aspect.

[Cobra]

but that means that it's giving me the error because i don't have mv_adveffects?

[StratComm]

No, I'm saying I get that error with the correct files.  I'm saying it's not important because it isn't actually affecting anything, so don't worry about it.  I'll figure out why it's happening and update WMC on it later, so the official release can get the warning gone (media VPs if it's media related, the code if that's the problem), but it's not something that will effect anything you're doing.  It won't show up outside of a debug build so it's not worth worrying about.

[Cobra]

well, you're right about the error not showing up outside of the debug build.

[WMCoolmon]

I still don't understand what was happening with the radar. I thought there was some sort of graphical oddity, besides the G3_count error, maybe in release builds?