home *** CD-ROM | disk | FTP | other *** search
-
-
-
- I'm writing my own bob and sprite handling routines
- for my 1200 under Amospro. These routines can use full AGA capacities.
- The 256 colors bobs works fine, but are too slow....:-(. I discover that
- recoding the screen using the copper would speed up the blitting. Amos is
- really good for coding games , but when you have to access the hardware
- directly, the job can be done faster and better in C. Also the compiler
- doesn't produce code for 020,030,040 and 68882 copros, doesn't optimize
- the code (see GCC) and the amos degugger (Monitor) really sucks!!!
- (see SAS cpr, REAL debugger). So I decided to do all the job using C.
- It's why i am very interest in MARP... I have just one question about
- MARP : will MARP support AGA using highly-optimized-direct-hardware-access?
- If not, I think I go to do my own ARP during the christmas holidays...
-
- I have a big question about sprites : in the sprite bank, bobs are
- stored plane by plane and this is good for blittering (but slow). Sprites
- lists needs a line by line coding(i.e. 1st line of the 1st plane until 1st
- line of the last plane, then 2nd line ...). Therefore sprite memory
- storing isn't compatible with the bob one. So I thougth that when the
- 'sprite' command is called, amos recoded the image datas in another place
- in memory, but I watch the memory just before and after a sprite call, and
- I noticied no change. So, does amos use a special copperlist to handle
- sprites and bobs using the same memmory data, or are they another way ?
-
- PS: I hope you don't suffer too much reading my really bad english... :-)
-
- /---------------------------------------------------------------\
- | Nenad Rakocevic Email: p6ip007@cicrp.jussieu.fr |
- | Amiga User : A1200 020/28MHz 882/33MHz 6Mo |
- | internal: HD 80Mo external: OverDrive 250Mo |
- \---------------------------------------------------------------/
-
-