Binary Compatibility 80286

[BostonU SysMgr]

>From: timothym at tekigm.UUCP (Timothy D Margeson)
>I have not seen anything that demonstrates any form of compatibility between
>any of the Unix' available. You have so many various forms and versions of
>the ne'rdowell 'operating system' (I use the term loosely) that source written
>for one can not be expected to run on any other system.
>...
>The phalacy that Unix and C make things portable makes me sick whenever I
>hear or read someone tauting.

The first paragraph seems to contradict itself. If you can get over your
illness for a moment can you say the following three times fast:

	Vax-Sun-Celerity-3B-IS-VM370-CRAY-IBMPC-Alliant-Sequent-Encore-
	-PDP11-MACXL-Gould-Convex-Amdahl-PC7300-Zeus8000-Sperry1100-etc
	-etc-etc-etc

I knew you could.

>The software writer is what makes programs portable, nothing else can.

This is a necessary but not sufficient condition, try to port your MS/DOS
ASM programs, no matter how well written, to some of the above.

The original argument had something to do with suggestions about optimizing
operating systems to provide the lowest possible overhead for software
vendors.

I have a question, there seemed to be some temporary success of vendors in
the VMS market and, from experience, it becomes binary incompatible with
itself about every 6 months (if you doubt that, just ask anyone who did a
4.x upgrade, everything had to be re-released, a lot still doesn't work.)
We're not talking 'just re-compile it', we're talking major workovers of
code that was necessary. Yet, the vendors survive (why?)

Another question, aren't these numbers being thrown about (thousands of
MS/DOS programs) mostly games? And how many actually account for sales?  I
bet a small handful, and aren't equivalent programs (or better) available
for almost any but the lowest volume UNIX systems? Like spreadsheets,
word-processors and data-bases? (not to mention what comes bundled in any
UNIX system, about 200 common programs before you buy a thing, many of
which provide exactly what those MS/DOS users are buying, like editors.)
Or do we now take things like C, yacc, lex, nroff, troff for granted?

If games is the criterion, then I think MVS is in trouble also.

And how big is this 'amazing' PC market anyhow? I know that IBM does about
$2 billion off it a year, I suspect the entire rest of the market combined
doesn't add up to another $1 billion, probably much less.  Which all adds up
to, oh, about the yearly budget of Bell Labs or so (consider that IBM alone
is a $40+ Billion dollar company, DEC about $4 billion, ATTIS? I know they
just signed a $1 Billion contract with NSA for Unix systems, that's half the
whole IBM/PC market in one sale [that is, IBM's lion's share.])

Now, consider, IBM alone makes about 10-15X the entire PC market off of MVS
and VM. DEC makes about double the entire market off of VAXes and terminals
(I guess, and software and maintenance of course.) I assume AT&T is similar,
companies like CDC, DG, PRIME etc each have a market about the size of the
entire IBM/PC market, ok, let's say 1/2 each...the point stands.

Ok, something else to ponder, the operating budgets of the 4 major
universities I can see out my window (BU, Harvard, MIT, NorthEastern)
probably exceeds the entire IBM/PC market...(I exagerrate, I have to
go upstairs to see them all!)

Ahem, where exactly is this 'profitable volume market' that UNIX should be
rewritten for? Oh, right, but *those* people aren't buying PONG! Why...
THEY DONT EVEN HAVE FLOPPY DISK DRIVES!! Do you know how much more mag
tapes cost than floppies! And mag tape drives to make distribution tapes!
They don't even *sell* tape drives in Toys-R-Us!

Oops, sorry to drag the facts in...but the PC market is, if you are at all
serious about the computer market, a trivial glitch on the curve. (oh, I
would take that glitch! But I doubt very much that the success of UNIX
depends on it.) Though I do find this egalitarian capitalism amusing.

To ward off the net-boors, I am sure one fact here is inaccurate, delete or
correct that line and make sure the rest of the argument collapses before
you take the time to reply.

	-Barry Shein, Boston University