[comp.sys.att] Signature of ATT 6300 graphics

rds95@leah.Albany.Edu (Robert Seals) (04/12/88)

Yo,
    I'm using Turbo C 1.5 to write some purportedly portable (!?)
programs that include some graphics. My version of TC does not
autodetect a PC6300Plus adapter/screen as being a 6300, but a cga, even
though it does support drawing in 6300 mode when initialized
explicitly as being 6300.

    Zo, does anyone know what kind of signature to look for to try
to identify an ATT as being ATT?

    Murky buckets, etc.

rob
Robert Seals @ ASRC
rds95@leah.albany.edu or
rob@asrcmv.albany.edu

jeffm@mmintl.UUCP (Jeffrey Miller) (04/15/88)

In article <682@leah.Albany.Edu> rds95@leah.Albany.Edu (Robert Seals) writes:
>    I'm using Turbo C 1.5 to write some purportedly portable (!?)
>programs that include some graphics. My version of TC does not
>autodetect a PC6300Plus adapter/screen as being a 6300, but a cga, even
>    Zo, does anyone know what kind of signature to look for to try
>to identify an ATT as being ATT?

Try looking for "Olivetti" or the machine name somewhere.  I just can't
remember where it's supposed to be although I've run across it before.
It's the same kind of thing Norton, etc. use to attempt to identify the
computer name, such as in Norton's SI.  But you still prpbably need to
make sure the user has not removed his/her ATT video card.  Perhaps you
can just search the BIOS area of memory (>=F000:0) for the string.

Jeff
-- 
* Jeff Miller:  Ashton-Tate  52 Oakland Ave., E. Hartford, CT  06108-9911  *
* (203) 522-2116 x257  UUCP: ...!seismo!utah-cs!utah-gr!pwa-b!mmintl!jeffm *
*                         or ...!ihnp4!philabs!pwa-b!mmintl!jeffm          *

pechter@dasys1.UUCP (Bill Pechter) (04/16/88)

In article <682@leah.Albany.Edu> rds95@leah.Albany.Edu (Robert Seals) writes:
>Yo,
>    I'm using Turbo C 1.5 to write some purportedly portable (!?)
>programs that include some graphics. My version of TC does not
>autodetect a PC6300Plus adapter/screen as being a 6300, but a cga, even
>though it does support drawing in 6300 mode when initialized

Unfortunately, not even Borland can figure out a way to autodetect the 6300
display.  I asked them the same question last month.  One way to do it
(a real kludge -- but it should work is to look for the copyright notice from
AT&T &/or Olivetti in the Bios Roms).  I may try to do this myself in the future
but I haven't tried it yet.

Before complaining about the problem -- the reason we've got the autodetect
difficulty is the extremely good cga emulation that Olivetti designed into the
M24/M25 machines.  Everything I've tried on the video installs for a CGA and
runs beautifully -- so I guess we should be glad for the small problem. :->
 
-- 
Bill Pechter          {sun!hoptoad,cmcl2!phri}!dasys1!pechter
Lakewood Microsystems, 103 Governors Road, Lakewood NJ 08701
			(201)370-0709 Evenings
Big Electric Cat Public Access Unix, New York, NY

mcripps@mtuxo.UUCP (XMP12-M.CRIPPS) (04/22/88)

In article <682@leah.Albany.Edu>, rds95@leah.Albany.Edu (Robert Seals) writes:
> Yo,
>     I'm using Turbo C 1.5 to write some purportedly portable (!?)
> programs that include some graphics. My version of TC does not
> autodetect a PC6300Plus adapter/screen as being a 6300, but a cga, even
> though it does support drawing in 6300 mode when initialized
> explicitly as being 6300.
> 
>     Zo, does anyone know what kind of signature to look for to try
> to identify an ATT as being ATT?
> 

If all you're worried about is video, that's easy.  Change the video mode
to 40h (the AT&T Hi-Res mode) and do a get video mode call
(int 10h, function 0fh) to see if it returns 40h.

Alternately, write something to 0b800:6000 and see if you can
read it back (the AT&T board has 32K of video memory, the CGA has 16K).
Using these methods, your routines can support *ALL* AT&T machines with the
non-standard video boards, without being ROM dependent.

The get video mode trick may return a false positive if other manufacturers
use 40h as a mode, but I'm not aware of any.  The memory test may be true
on non-AT&T boards, but functionally the boards are probably the same, so
it wouldn't matter (The Sigma Color 400 is mapped the same as the AT&T, for
instance).

------------------------------------------------------------------------------
       ^         Michael J. Cripps
     / I \       AT&T Lincroft, NJ	These opinions reflect none other
    -------      mtuxo!mcripps		than myself.
   /_|___|_\     (201) 576-2428
  /|___|___|\
 /___|___|___\   Have a nice day fnord.
-------------------------------------------------------------------------------