New DOS Interrupt List Posted

[Ross M. Greenberg]

What you've all been waiting for: the new and improved DOS List
of Interrupt Vectors. Ta-Ta-Ta-Tah!!!

Considerable work was put into this by both John Ruschmeyer
(ihnp4!vax135!petsd!moncol!john) and Bill Frolik (ihnp4!hplabs!hp-pcd!bill).
Your humble editor did a little, too.

There was a little grumbling last time that a lot of this stuff is
found in books. Right. Then buy the book.

I would appreciate any additions and corrections be sent to me, and
if diff format, if possible.

And it would be nice if everybody pestered the folks on the net from
Microsoft to fill in some of the UNKNOWN's below. I should
hope that they know what they're for.

Then again, maybe not :-).

--
------------------------------------------------------------------
Ross M. Greenberg @ Time Inc, New York
--------->{vax135 | ihnp4}!timeinc!greenber<---------

I highly doubt that Time Inc. would make me their spokesperson.
----
"I saw _Lassie_. It took me four shows to figure out why the hairy kid
never spoke. I mean, he could roll over and all that, but did that
deserve a series?"

[Rick Sellens - Mech. Eng.]

In article <3...@timeinc.UUCP> gree...@timeinc.UUCP (Ross M. Greenberg) writes:
>
>What you've all been waiting for: the new and improved DOS List
>of Interrupt Vectors. Ta-Ta-Ta-Tah!!!
>

>And it would be nice if everybody pestered the folks on the net from
>Microsoft to fill in some of the UNKNOWN's below. I should
>hope that they know what they're for.

It strikes me that "used internally by DOS" is a perfectly legitimate
description. As soon as you document the function of an interrupt you
make an intrinsic commitment that all future releases of DOS will
support that interrupt. This could make it very difficult to upgrade
the system in a reasonably neat way.

The use of undocumented "features" makes your code dependent on a
particular version of the operating system. I would suggest that you
use undocumented interrupts *only* when there is no other way to
accomplish what you want to do. When this occurs you should pester
Microsoft to *add* a feature to the next release of the operating
system.

I hope the software I'm using now wasn't written with dirty code,
because eventually I might like to upgrade to DOS 3.0 or 4.0 or 5.0....

Rick Sellens
UUCP: watmath!watdcsu!rsellens
CSNET: rsellens%wat...@waterloo.csnet
ARPA: rsellens%watdcsu%waterlo...@csnet-relay.arpa

[Ross M. Greenberg]

Making code that depends on some of the little known features of
DOS isn't always bad:

If I'm writing code that is specifically designed to get around one of the
short-comings of DOS 2.x, then I've just got to make sure that 3.x
people don't use the code. That's why my code always checks the
version number (at one time an undocumented feature, BTW!).

I would like to see Microsoft put something in like:

INT X, AH = yy - Turn on multi-user, multi-tasking (Not supported)

then I could use it, and know it will go away (maybe) in the next
release.

--
------------------------------------------------------------------
Ross M. Greenberg @ Time Inc, New York
--------->{vax135 | ihnp4}!timeinc!greenber<---------

I highly doubt that Time Inc. would make me their spokesperson.
----

"I had a cat. She died. Had a goldfish. Died. Guppies. Died.
Gerbils. Died. Tippy. Died." - little girl
"Alright! So I don't like small animals!" - Mr. Death

[Bill Crews]

> The use of undocumented "features" makes your code dependent on a
> particular version of the operating system. I would suggest that you
> use undocumented interrupts *only* when there is no other way to
> accomplish what you want to do. When this occurs you should pester
> Microsoft to *add* a feature to the next release of the operating
> system.
>
> I hope the software I'm using now wasn't written with dirty code,
> because eventually I might like to upgrade to DOS 3.0 or 4.0 or 5.0....

For the majority of DOS applications, the above is certainly good advice.
However, DOS applications that are really ambitious do typically have to
use version-specific features of DOS. I suggest this is quite OK if:

there is no way to avoid it,

the payoff is worth it to the end user,

and it is documented to be version-specific.

I imagine that the vast majority of those interested in the list are involved
with software of such sophistication.
--

/ \ Bill Crews
( bc ) Cyb Systems, Inc
\__/ Austin, Texas

[ gatech | ihnp4 | nbires | seismo | ucb-vax ] ! ut-sally ! cyb-eng ! bc