Huh, this is the first time I have seen this post (or sub-board?). Anyway I think I can contribute... at least in terms of VICE. I've made several hacks of it, and a few of them are now official parts of VICE (I give feedback to the VICE Team, although they don't always agree/adopt my contributions).
Well, I submitted 2MHz for VICE/C128. My implementation was very simple... they eventually made it official with "enhancements" (like clock-stretching).
My original contribution allowed the Z80 to run at "official" 2MHz (i.e, 4MHz / 2). But VICE 2.1, 2.2 (and later?) are crippled! The Z80 still runs at UNREALISTIC speed of 1MHz. I have not investigated if 2.3 or 2.4 allows "full speed" Z80 like my original submission allowed.
I submitted "2-bank Color RAM" (for C128... shock!) which greatly improved split-screen modes, like GRAPHIC 4. This has been made official, but I can't find a reference to my contribution (the official release followed my submission).
I submitted various DISK enhancements. Most of them were rejected, but a few were accepted...
In earlier versions of VICE (2.1 or less?) it was easy to find me with menu item "HELP/ABOUT", but in more recent versions it only lists the main contributors... so you would have to search the source code to confirm my contributions.
... I use to just submit DIF files via email (in the past)... I haven't submitted anything (recently) that has been accepted via SourceForge.
Off topic, I want to post an FD-X000 drive experiment on this forum, and then forward the results to the VICE Team... because I believe their current implementation is *seriously* flawed... Specifically, I don't think any of those CMD drives, using Fast-Serial, can respond in 16 microseconds... which is what the emulation currently uses (and causes immediate failure of my Media-Player).
I have been trying to think of a *general* interlace algorithm that would work with VIC-20 (hardware... but only NTSC) and C128 (both 40 or 80 columns with NTSC or PAL). Hmmm... I think I am most qualified to hack x128 for 40 columns (80 columns also seems possible, but different). I am really not qualified to hack VIC-20... although it was my first CBM, I have not actually used one in decades!
Last Edit: Jan 31, 2015 10:48:00 GMT by hydrophilic: Thought of something else...
Z80 speed has been working a lot better since VICE 2.4, although now it is a little too fast, at least it's not half-speed anymore.
Sadly, I really don't know C and have no real understanding of the VICE-codebase at all. I think between you and me we know the most about interlace-modes on the C128 and VIC20 - you with your VIC-IIe-interlace, me with all those VDC-modes and my two VIC-20 demos for NTSC-machines that use interlace. Especially VIC-20 NTSC-interlace is pretty straightforward, although there are still some timing issued with the NTSC-VIC-emulation in general that I would have to investigate further.
For the VDC it would be good start if emulation allowed for larger screen areas first, and interlace later.
Thanks for report... I guess I need to try your "TRON" with VICE 2.4 and see how it compares to older version and my NTSC machine!
Yes, x128 emulator should really support wider images (like 720) and also interlace...
Sadly I don't have time to hack x128 now, but it seems not very much trouble (with double-raster enabled) to support true interlace...
Well if somebody can make VDC interlace work, it should not be much effort to port that effect to VIC-II... not sure about VIC-20, but if VIC-II interlace works, then just a little more work for original VIC chip...