TSE "Milestone 3.5" Update
by Kyle Carter · in Torque Game Engine Advanced · 06/30/2006 (11:06 am) · 246 replies
Hi guys,
We've checked the next TSE update into CVS. It's NOT Milestone 4 - so no fancy lighting - but it does include many bugfixes, a new vastly more efficient rendering infrastructure (ported from Marble Blast Ultra), way faster skinning, and the first drop of the new Atlas codebase.
Brian has more complete info on the delta in this update, and should be posting it sometime in the not too distant future - I just wanted to let you guys know what was up ASAP. :)
Please note that some parts of this are still a little rough, and the new Atlas codebase does not yet have the full complement of art tools, but everything is functional and ready to use.
This thread is for posting general feedback on the release and collaborating on any bugs that might come up. Please be patient, as here in the States we're just about to go into a 4 day weekend - so responses might be a little slow. But we are committed to making everything work 100%, so definitely let us know what your experiences are!
Regards,
Ben Garney
Torque Technologies Director
We've checked the next TSE update into CVS. It's NOT Milestone 4 - so no fancy lighting - but it does include many bugfixes, a new vastly more efficient rendering infrastructure (ported from Marble Blast Ultra), way faster skinning, and the first drop of the new Atlas codebase.
Brian has more complete info on the delta in this update, and should be posting it sometime in the not too distant future - I just wanted to let you guys know what was up ASAP. :)
Please note that some parts of this are still a little rough, and the new Atlas codebase does not yet have the full complement of art tools, but everything is functional and ready to use.
This thread is for posting general feedback on the release and collaborating on any bugs that might come up. Please be patient, as here in the States we're just about to go into a 4 day weekend - so responses might be a little slow. But we are committed to making everything work 100%, so definitely let us know what your experiences are!
Regards,
Ben Garney
Torque Technologies Director
#222
07/12/2006 (7:12 pm)
Hey, its great to know that my fps will be going up sometime, :-D
#223
We both work on the same project. I will try to get the latest build today and tell you how it goes for me. I have a 6600 GT as well though not mobil. There were known issues with 6600 GT with previous TSE milestones. I don't know if any were addressed yet. I'll let you know.
07/13/2006 (8:57 am)
@ PauliverWe both work on the same project. I will try to get the latest build today and tell you how it goes for me. I have a 6600 GT as well though not mobil. There were known issues with 6600 GT with previous TSE milestones. I don't know if any were addressed yet. I'll let you know.
#224
Looks like we have to look into the problem with our project Paul.
07/13/2006 (12:46 pm)
Well I downloaded the new update for our project and i'm getting some low fps. Do our interiors have LOD? The framerates only seem to drop near interiors and some of the water. I also Downloaded 3.5 from CVS and I see a marked improvement. I'm consistently getting 180 + and 150 + with 4 of the large bases in view that are included in the TSE Demo. *Note that interior previously caused low fps in previous milestones on my card. The interiors in our project caused low fps before too but still the same after our merge. It's either we missed something in the merge and/or need to do some optimizations on interiors.Looks like we have to look into the problem with our project Paul.
#225
I wouldn't be suprised to find that the problem is halfway between, i guess i'll continue to test things out on both and see what problems exist in just our build and in 3.5 .
07/13/2006 (12:57 pm)
I was happier when i thought Ben was going to fix it for me :( I wouldn't be suprised to find that the problem is halfway between, i guess i'll continue to test things out on both and see what problems exist in just our build and in 3.5 .
#226
This caused Con::active to be set to true before it should have been which caused Con::init to assert. I just commented out the two Con::active assignments to get around it for now.
07/13/2006 (12:59 pm)
It should be noted that this particular hack caused me quite a headache while using the Dynamic Library resource.// This prevents infinite recursion if the console itself needs to // re-allocate memory to accommodate the new console log entry, and // LOG_PAGE_ALLOCS is defined. It is kind of a dirty hack, but the // uses for LOG_PAGE_ALLOCS are limited, and it is not worth writing // a lot of special case code to support this situation. -patw Con::active = false; consoleLog.push_back(entry); Con::active = true;
This caused Con::active to be set to true before it should have been which caused Con::init to assert. I just commented out the two Con::active assignments to get around it for now.
#227
If you were on coronada getting those frame rates its because of the fix we had to implement to get around the broken Vertex handling code constantly overflowing its buffer. Matt Fairfax said a few months agothat eventually GG would get around to fixing this by implimenting a per zone multi buffer system but it has not been released yet. Right now that interior is cut it up into quite a few pieces to keep it from overflowing, that combined with the solid glass roof makes that map lag really bad as it cant be portaled. Dont worry though, that map wont be in the final cut I was using it to play with shaders :p. We can talk more in our forums about this as I dont want to hijack this thread anymore :x
07/13/2006 (1:06 pm)
Let me get the interiors finished before I worry about lod's :pIf you were on coronada getting those frame rates its because of the fix we had to implement to get around the broken Vertex handling code constantly overflowing its buffer. Matt Fairfax said a few months agothat eventually GG would get around to fixing this by implimenting a per zone multi buffer system but it has not been released yet. Right now that interior is cut it up into quite a few pieces to keep it from overflowing, that combined with the solid glass roof makes that map lag really bad as it cant be portaled. Dont worry though, that map wont be in the final cut I was using it to play with shaders :p. We can talk more in our forums about this as I dont want to hijack this thread anymore :x
#228
07/13/2006 (4:29 pm)
I'm going to make 1 more comment and then give Ben back his thread. i overwrote a stock TSE & Starter.fps's Data folder, withour data folder and conducted some tests (not done the best way but it got me the answers i wanted). The missions that would run without crashing the engine (I'll do more thourough tests later), one of them (which had nothing but a terrain) demonstraited the same direct X warning for a short period of time (not the entire time running like all missions in our custom code base) but again this mission had just a terrain and sky. thats it. So its not ENTIRELY our code base.
#229
The fix for that is to replace with this code:
This will prevent the console from inadvertantly activating. (I'm sure glad I don't do space rocket control software. ;)
07/13/2006 (9:02 pm)
@Scott Burns:The fix for that is to replace with this code:
const bool save = Con::active;
Con::active = false;
consoleLog.push_back(entry);
Con::active = save;This will prevent the console from inadvertantly activating. (I'm sure glad I don't do space rocket control software. ;)
#230
Thanks, that did the trick. I'm sure NASA programmers use just as much duct tape and bubblegum on their code as the rest of us.
07/14/2006 (8:17 am)
@BenThanks, that did the trick. I'm sure NASA programmers use just as much duct tape and bubblegum on their code as the rest of us.
#231
NASA does indeed use duct tape (a special kind that can survive extreme cold): www.cnn.com/2006/TECH/space/07/11/space.shuttle.ap/
07/16/2006 (8:37 am)
Warning: Useless postNASA does indeed use duct tape (a special kind that can survive extreme cold): www.cnn.com/2006/TECH/space/07/11/space.shuttle.ap/
#232
07/20/2006 (5:53 pm)
Ok, thats just damned random....
#233
07/20/2006 (5:57 pm)
Not COMPLETELY random... it was directly related to the post immediately above it.
#234
07/20/2006 (6:02 pm)
I meant in terms of astronaut's wanting duct-tape to hold their jetpacks on.....
#235
07/20/2006 (6:12 pm)
I think its a desire not to float off into space and die.
#236
For myself the lagging gittering feelings is linked with mouse movement, seems like there's no mouse smoothing going on. Haven't fussed with looking to see if that's it yet.
Can anyone confirm or is it just system related (sli bug)?
My specs:
P4 3.5ghz
1gb ram
2 sli-7900GTX 256mb
07/24/2006 (4:01 pm)
@T Squared,For myself the lagging gittering feelings is linked with mouse movement, seems like there's no mouse smoothing going on. Haven't fussed with looking to see if that's it yet.
Can anyone confirm or is it just system related (sli bug)?
My specs:
P4 3.5ghz
1gb ram
2 sli-7900GTX 256mb
#237
its not just mouse. its even keyboard motion. its mainly when i'm looking towards a interior object. with normal maps on it. i notice the slow down in fps. "metrics(video);"
07/24/2006 (4:46 pm)
@Jonathan Sanduskyits not just mouse. its even keyboard motion. its mainly when i'm looking towards a interior object. with normal maps on it. i notice the slow down in fps. "metrics(video);"
#238
07/25/2006 (12:29 pm)
Camera interpolation is currently off (got broken with TGE merge - TGE has same camera problem) , and on slow video cards you can experience lag and jerkiness because the video card can get too many frames ahead of the input. These issues are on my bug list.
#239
07/25/2006 (12:52 pm)
Is there any major differences in wheeled vehicles between TGE 1.4 and TSEMS 3.5? I'm up to the porting of vehicles now and the tires don't seem to be showing up when I add the vehicle.
#240
07/25/2006 (2:23 pm)
I noticed tires don't show up either, for me.
Associate Kyle Carter
We're working on fixes but it may be a little bit before they're out (ie, not gonna happen today :).