Re: Apple ][gs VGA Card

[ntt@cleese.apana.org.au (Jonathan Pratt)]

jlee@clyde.ICS.UCI.EDU (Joseph) writes:
> Lim Thye Chean (ltchean@iss.nus.sg) wrote:
> : OK, so what have you figure out from here? I don't really know. From the text
> : alone, I would conclude that it is only using 640 x 200 and 320 x 200
> : resolution since you can still use your existing Apple monitor. The screen
> : shot does show that the VGA images are bigger than the Super Hires, so maybe
> : there is a difference. Not sure.
> 
> Unless you use overscan that you can get 640x400 on a GS monitor. It has
> been done.

It would be nice if Jawaid put us in the light, but the resolution
limitation (since it's s'posed to be able to use the GS monitor) would
have to be around 640x400 which you can get from the video overlay card. 

> :   - 256 colors on screen. You can choose from one of the 16 palettes in memory.

This is probably closest to the truth (but we'll find out someday soon ;^)

> : And seems like it includes only a image viewer. The only software that support
> : it is DiscQuest (or discQuest?). But all of the sudden, it mentions about
> : editing text in GS/OS. So what does it mean? ShadowWrite now support Second
> : Sight? Or all IIGS desktop images are adjusted so it looks 1:1 proportion?
> : Or what? Again I am confused.

This really does depend on the way in which the board works.  Please note
that the VOC can produce an image which is identical to the screen image
without getting a DSP and inputing the motherboard video output.  It does
so by intercepting writes to the screen portion of memory (and if the SS
board does this it will _have_ to be in slot 3)

If you know much about the VGA on a PC, you will understand that it has a
pallette colour loaded into a certain port (that is a value representing
the pallette colour to use)  Each byte from the video screen memory
actually allows the user to modify 8 pixels on the screen. (usually 8
consecutive horizontal pixels)  When you write to video memory, each bit
in the byte you write indicates whether or not the corresponding pixel is
modified to the colour in the port.  An 'on' bit means change the colour,
while 'off' bits make no changes.  This technique is called paged
graphics, and can speed things up quite considerably.

Now if Jawaid implemented a similar system, his SS card could potentially
intercept writes to the screen memory and modify pixels on the screen
accordingly, so thatit interprets 2 of four bits into a (possibly group
of) pixels on the screen allowing the card to be 100% invisible to the GS
and allow any program to make use of it (except ones that are dependant on
SLI's like 3200 veiwing/displaying s/w).  I expect that it would also have
a 'native' mode allowing access to a full 8 bits per pixel.

If this were the set up, in 'emulation mode' a single byte of screen
memory could refer to 4 or 8 (depending on 320 or 640 mode) pixels, half
on one scan-line, half on the next which would give a resolution of
640x400 which is close(r) to 1:1, although you would not recognise it in
current application.  But new ones could use the board to take advantage
of this resolution.

This is just hypothesizing, and might be as far off beam as some of the
other comments in this thread.  All I want to point out is that it is
possible for a new graphics engine to 'support' all (most) current
programs, and provide expanded features for new ones

cheers
ntt@cleese