/
TODO.txt
26 lines (13 loc) · 1.59 KB
/
TODO.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
* Add a frame number indicator, useful for knowing what output file has been written
* If a drawScreen() happens, while single stepping for example, then pixels might not be rendered in precise granular sync until a digital logic change is emitted by VSM
Consider flushing the Display::simulate() since it might not have received any state update and black (or constant colour) pixels might not be rendered in sync with the single step
* Using the mode 7 test jig, it might be possible to render the distorted/scaled quad as a sprite by detecting when the edges need not be drawn. Then to render pixels to a frame full buffer for later display.
The frame rate then need not be the limiting factor for drawing frames. The distorted/scaled quad draw could read a list of sprites to draw into the frame buffer.
* Mode7 test: Before RXO and RYO on sheets 2/3 reach U9 on sheet 4, need to add in the 16 bits of "x org" and "y org" from Mode7Regs.ptn
Will also need the register loads
This can then be copied over to the other display board and RAM hooked up
Added a "map tile" display area, seems to work well.
If the U43 to U9 chain is too slow, then consider adding 2 or 4 paths based on the horizontal pixel position and longer address time latching for the output pixels
In the situation where they are RAMS, it would be possible to tie their data loads all together of course
* RXO and RYO would need to have an extra 4 bits to really make use of the tile map
* Interesting to note that the display is over scanning the screen, way beyond the pixels displayed by the original Bomb Jack 2D graphics board