Quote from: Zack on 17 Jul 2024, 01:26 PMNote: It's important to call vcsJmp3() a few times throughout the screen because the 6507 can only address 4KB of contiguous ROM space and allowing PC to go past $1fff would transfer control to TIA read registers, eventually crashing the 6507.

Took me a while to understand this, but to clarify for other readers - the kernel is pumping bytes onto the bus which are instructions executed by the 6507 -- various register updates (PF0/GRP0) etc. The kernel does this for 192 scanlines - and the 6507 happily executes these instructions. However, the 6507 program counter is "free running" - it thinks it is retrieving all of these instructions from memory linearly. So it's doing 192 scanlines of code execution without any looping or branching. Let's say (just roughly) 50 bytes per scanline.  So with just 4K of address space we would run out of runway within about 80 scanlines at the very best. Depends on where the PC is when we start. It seems to me that setting PC to $1000 at the start of a kernel, then every (say) 32 lines we do a vcsJmp3() to reset it... would be safe enough.

However, the kernel could keep a track of implicit PC address and only do the vcsJmp3() write when it becomes necessary/urgent. If I'm understanding all of this correctly, the current centered kernel uses 36 bytes without the pC rest. If the PC was set to $1000 just before the scanline loop, then that would suggest 113 scanlines would be possible before we're in trouble. That would require just one vcsJmp3() halfway down the screen.

Newest ship viewer binary. Nearly all the Elite ships now present. 29 in total.
fire to cycle to next ship
up/down to move ship closer/farther
left/right to change palette

I ended up automating pretty much all of the conversion process, so the last 10 or so ships were very quickly installed. I should have automated much earlier.

Latest ship viewer binary. A couple of extra ships added.
+ use left/right to select a new palette (random)

I thought I'd share the "workflow" of installing a new ship. First I grab the raw data, put it through the AI-generated tool to identify the clockwise faces, then do some manual editing of the data to place into structures in my engine. Finally I go through and change the face colours one by one until we have a finished ship. This particular ship has unusual faces so it does show up the fact that I'm not sorting faces at this stage.

Glorious...  this is a "ship test viewer" that lets you cycle between the installed 3D objects (joystick button press) and zoom in/out (joystick up/down).  This is using a newly-rewritten rendering pipeline that works on a poly/face level rather than an object level, giving finer granularity in timing. So, things can now get full-screen (although not fully-tuned so expect an occasional crash)...

This is an extended rendering test with random generation from the (current) 12 spaceship pool of shapes. Emulator (Gopher) was running this mostly full-speed, although I can get extra real speed by a bit more finesse in the rendering code. It's still pretty impressive, though.  Of particular interest to me is that some of the objects (the "death star-like" object) get pretty much full screen and the system is coping well with this.

Coriolis

Some of these are massive -- you don't get sizing information when displayed individually - think of this one as a mini death star, compared to the other spaceships. That yellow "window" is an entry for big spaceships.

Constrictor

Cobra MK1

Boulder

Capsule

Barrel

Asp

Asteroid

Anaconda