Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/5/84; site aicchi.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxr!mhuxt!houxm!ihnp4!aicchi!ignatz
From: ignatz@aicchi.UUCP (Ihnat)
Newsgroups: net.micro.att
Subject: Re: Open architecture of AT&T computers?
Message-ID: <538@aicchi.UUCP>
Date: Sat, 3-Aug-85 14:14:39 EDT
Article-I.D.: aicchi.538
Posted: Sat Aug  3 14:14:39 1985
Date-Received: Sun, 4-Aug-85 09:35:36 EDT
References: <1701@ecsvax.UUCP> <1116@ihuxn.UUCP> <204@mplvax.UUCP> <395@cuae2.UUCP>
Reply-To: ignatz@aicchi.UUCP (Ihnat)
Organization: Analysts International Corp; Chicago Branch
Lines: 62
Keywords: 3B2 AT&T

In article <385@cuae2.UUCP>, Chris Donahue of AT&T Info. Sys. Application
Engineering posts:

>> Are you speaking of things like "3b2/300 Computer Feature and Interface
>> Design Manual" Select Code 305-492?  I ordered one of these in March, and
>> it was back-ordered with no predicted delivery date.
>> 	carl lowenstein		marine physical lab	u.c. san diego
>> 	{ihnp4|decvax|akgua|dcdwest|ucbvax}	!sdcsvax!mplvax!cdl
>>
>If you ordered one of these, you haven't received it because of a backlog.
>It is restricted to Value Added Resellers of AT&T only. My group controls
>the distribution of that particular document (because we support what is
>done with it). I guess you didn't read my article defining "open architecture"
>in the eyes of AT&T. Sorry, we are unable to support anyone who wants to
>develop I/O feature cards. We have to restrict ourselves to VARs. The
>design of a 3B2 I/O feature card is MUCH more complex than designing a card
>for a PC, so it requires alot more support resources.

I'm afraid there's a really basic and potentially terminal problem here.
I agree that, particularly in these hectic early days, I wouldn't expect
AT&T to hand-hold anyone who dreamed of designing a 3B2 I/O feature card
and making their fortune.  BUT a more basic attitude, which has become
more prevalent with AT&T's entry into the competetive market, is displayed
by the attitude of restricting information:  "restricted to Value Added
Resellers of AT&T only..."  Also, they make the decision that, because
a 3B2 feature card is complex, and requires more support resources, the
information on how the card operates is restricted.

This, incidentally, is *not* just an attitude that Chris developed, or
even his group; it's reflected in any dealings you have with AT&T these
days.  A friend of mine bought a 3B2/300 recently; in the course of setting
it up, he discovered that unplugging one of the peripherals (I believe
it was the XM connector?) destroyed the format on the hard disk.  No problem,
normally; but... AT&T maintained that the format program was only available
to their own technicians.  Not even VAR's.  Now, he didn't even have to
go to excptional means to trash the disk; and he's supposed to pay, as
an individual, normal field support fees just to reformat a DISK??  Hell,
most small business would balk at that type of dependence!  (They did,
for both the IBM PC and the DEC Rainbow, remember...)  He can't even BUY
the damn thing!

More generally, if the software were totally bug-free, then I, as a person
who usually gets stuck with making a box get up and stay up, would never
care about how the system works.  Practically, Unix(Tm and etc.) is *not*
a mature operating system.  All of us have had to either scan through source
code--for the kernel, and utilities--to find out why something broke, if
we were lucky enough to be at a site that had a source license; or expend
hours of sweat in adb/sdb if not.  Wanting to know how the system works,
and what it's doing, is more than just a desire for abstract knowledge.
And if the bus is too complicated, people either won't be able to design
cards, or they won't be any good, right?

I *want* the 3B2 to make it.  It's got great possibilities.  But if AT&T
doesn't learn from the mistakes of other manufacturers--if they don't price
their machines and software competetively; if they don't decide reasonably
which information and utilities can be distributed--then they're, regrettably,
going to eat a lot of dust before they can elbow with the industry leaders.
-- 
	Dave Ihnat
	Analysts International Corporation
	(312) 882-4673
	ihnp4!aicchi!ignatz