Re: Apple Game Server: Blowing off cobwebs

[Michael J. Mahon <mjmahon@aol.com>]

BLuRry <brendan.robert@gmail.com> wrote:
> On Tuesday, January 8, 2013 11:32:59 PM UTC-6, Kevin wrote:
>> I assume you will be prompt with providing tidy source code (lol!) as I
>> would still like to squeeze this into to my virtual serial host
> 
> All the sources have been checked in to the sourceforge SVN so you can do
> what you want with the code.  Though I haven't released a new build yet,
> I can confirm that it is stable so far -- even much more so that what was
> there previously.  I need to swap out the //c on my desk with the //gs in
> the closet and do some more testing.  Then I'll know if it's working
> right for all possible configurations.
> 
> Also, I've taken 100 screenshots and have all the games A-F covered
> (well, most of them except for the ones written in Basic or disk-based) 
> The next major release of Apple Game Server will be a little richer.  The
> new image dithering routine seems to be doing an ok job, but it needs to
> be beefed up some time later.
> 
> Now that I have my head wrapped around Floyd-Steinberg dithering, I know
> how to adapt my HGR conversion to use error-dithering and hi/lo palette
> selection.  It would require rewriting the error diffusion to take two
> paths for each group of seven pixels, one for the low (green/purple) and
> one for the high (orange/blue).  Whichever path yields the least error
> wins, and so on.  I think if I do this before releasing it will make
> things a lot cleaner looking.

You'll also want to minimize random flipping between the two color sets to
minimize gross patterns. Maybe use some hysteresis in switching color sets
instead of just using the one with (slightly) lower errors. 

-michael - NadaNet 3.1 and AppleCrate II: http://home.comcast.net/~mjmahon