Only 998 to go!



Another box of cornflakes.

Not trying to be an ass or hijack the thread. I simply want to remind John and all of the texture performance hit.

When the video card must change the texture, it must stop streaming 3D triangles and calculate what to render. The double edged sword to this object improvement is the performance hit of texture switch.

10 Tmods within the same gird, where each Tmod shares one texture , will have less performance hit then 3 Tmods that use 3 textures.

EAW is quickly aproaching the relm of texture atlas. It may already be in the code, but I have not looked.

You are perportionaly aproaching where your memory enhancements will start to negitivly effect the graphics pipeline.

Yesterday Pobs ane I flew online at GamzRanger with V20 and the "G V20TestPak" loaded.
It seemed fine

Here is Pobs flying over TMod256


Hovever, there was a problem with the German side with groundstarts as the bombers were destroyed on takeoff.
When we switched to fly Me109s from those airfields it was obvious that something was badly wrong.
It took several hours of searching, and the writing of a small program to help me, but I eventually found and fixed the problem

I do not know if it is due to my using a faster PC, but in the routine that rewrites the target file data when you add/move/delete a TMod an extra zero byte was added to the data by the editor. The data is in a grid with 32 columns and a row for every TMod. So if there are 10 TMods there are 10 rows and therefore 320 bytes in the file. So the file should be rewritten with 320 bytes, and it was clearly programmed to do that. Nowhere was it instructed to add a 321st byte, but my PC did that every time, even if I saved a file that had not been modified.
So I added a routine to check the file length and rewrite the file if one or more extra bytes had been added. It works.

The problem with the bombers was caused by this, because when the new scenario files are written the individual target data files are assembled to make "tardata.dat". Just one file with an extra byte pushes all the following data back by one byte so the rest of the tardata.dat file is corrupt.



The "999TMod.exe" has been deleted and replaced by a "999TModV2.exe", which had a new set of game files in the "G V20TestPak" folder, and a later version of the target editor (V11) in the "999TED" folder.

Jel

Hi Tony.

When are you going to start using the finished Terminus model I posted at the GEN?

Hi Allen.

I'm aware of the texture problem, and we are looking out for it, but as Tony mentioned, there seems to be no problems yet.

The 1000 limit is not going to be a real maximum, but it will be good to have a few more spare slots so we can have a proper selection of one off models like the Eiffel Tower.

John
I am just testing code with what I have, and trying to add the final pieces to the editor.
Currently I am working on the frontline.dat routines, and with a bit of luck the editor will be finished very soon.
Then I can get back to playing online- the tools are for others to use

Allen
Understood, but I do not think modders will go overboard and select 100s of TMods. What we are doing is to make a wider range of TMods available for modders to select from.

Jel

Another box of cornflakes.