Path: utzoo!attcan!uunet!zephyr.ens.tek.com!tektronix!nosun!techbook!fzsitvay
From: fzsitvay@techbook.com (Frank Zsitvay)
Newsgroups: comp.sys.dec.micro
Subject: Re: Rainbow EchoMail Digest
Message-ID: <1990Sep21.172327.16426@techbook.com>
Date: 21 Sep 90 17:23:27 GMT
References: <9009161251.AA14995@remote.dccs.upenn.edu> <1990Sep18.181936.4321@Solbourne.COM>
Organization: TECHbooks of Beaverton Oregon - Public Access Unix
Lines: 61

In article <1990Sep18.181936.4321@Solbourne.COM> imp@marvin.Solbourne.COM (Warner Losh) writes:
>In article <9009161251.AA14995@remote.dccs.upenn.edu>
>GTHEALL@penndrls.upenn.edu (George A. Theall) writes: 
>
>>Date:     09-12-90 (08:27)
>>To:       ALL
>>Subject:  BASCOM 7.0
>>From:     FRANK ZSITVAY
>>
>>does anyone know if you need code blue to run the microsoft languages
>>on a rainbow??
>
>I think that you do, but you may also need one other thing.
>
>>it appears that the compiler (and all code it produces) needs to run on
>>a machine with a similar interrupt structure to the ibm.  does code blue
>>remap all interrupts to their ibm counterparts in such way that the
>>microsoft languages can function??
>
>Some compilers (I don't know exactly which ones) rely on using
>interrupt 18H to do some function (don't have a clue what).  Code Blue
>is very good about emulating PC interrupts, however, it doesn't do a
>thing with INT 18 (since that is a ROM basic interrupt that has long
>since moved elsewhere).  However, INT 18 is basically used by the
>rainbow to talk to its hardware.  So, if the comiler establishes an
>INT 18 handler for whatever reason, the machine will instantly crash
>(or will crash the next time the screen or the keyboard is touched by
>the program.  I can't tell for sure since I stepped through the stuff
>using DEBUG).  However, all hope is not lost.
>
>There is a program that I wrote that you can install for those
>compilers that don't *NEED* code blue to work.  Turbo C and Turbo C++
>(and I think Turbo Pascal 5.x) are some examples.  However, there are
>some comilers that do require code blue to run (eg TC++ IDE).  It has
>been reported to me that the way that I fool the program breaks these
>programs because it also fools code blue (a very anti-social thing to
>do).
>
   is there anything i can do to make int 18 work like it does in
the pc clones??   i'd really like to use bascom on the rainbow, although
if i can't i could live with the restriction.

   what of the possibility of an adaptation kit, ala windows??  is
there someone out there that markets such an item??

   right off, i know of several things that won't work on the rainbow
when programmed in bascom.   the interrupt driven i/o for the serial port
would have no way of working unless it knew about the rainbow in the first
place, the various graphics libraries wouldn't work, all of the machine
specific stuff would die.   but i'd still like to use it even with these
restrictions if possible.

   the qbx environment uses mda mode 0 for output, and the scrolling is 
slow enough on my 16 mhz 286 to lead me to believe that it uses bdos
functions to write to the screen.


-- 
fzsitvay@techbook.COM - one of these days i'll get it right...

Version 2 of anything is usually the version that works.