NetNews Usenet Archive 1992 #30/NN_1992_30.iso/spool/comp/arch/11759

Text File | 1992-12-21 | 3.8 KB | 86 lines

Newsgroups: comp.arch
Path: sparky!uunet!mcsun!dxcern!dscomsa!zeus02.desy.de!hallam
From: [e-mail] (Phill Hallam-Baker)
Subject: Re: <None> (Should be Open Systems, bloody NEWS system...)
Message-ID: <[e-mail]>
Sender: [e-mail] (usenet)
Reply-To: [e-mail]
Organization: Deutsches Elektronen Synchrotron, Experiment ZEUS bei HERA
References: <[e-mail]> <[e-mail]>  <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Fri, 18 Dec 1992 14:16:55 GMT
Lines: 73

In article <[e-mail]>, [e-mail] writes:

|>In article <[e-mail]>, [e-mail] (John DiMarco)
|>writes:

|>> My basic point is this: A system's "openness" is a measure of its
|>> replaceability by a system from another vendor. By this standard, VAX/VMS
|>is
|>> not very open, and random-workstation/UNIX-variant is more but not
|>completely
|>> open.
|>
|>I'll go along with that.


Complete codswallop. By this definition the most "open" system is a house brick.
It is trivial to port any application that runs on a house brick to any other
system.

Functionality also comes into it. If people write programs that require high
levels of functionality from an O/S then they are going to be less portable.
Does this mean that the O/S should as a result be regarded as being better if it
provides less features?

Consider any application written for a line mode terminal using FORTRAN I/O. The
portability of these programs was pretty much setttled a long time ago. On the
other hand any code using a windows system soon becomes a porting hazard. On
UNIX there are even difficulties in porting code from one UNIX system to another
of the model and from the same vendor. This is because sysops tend to want to
split up large libraries amongst disks and UNIX does not provide a mechanism
like logical names which allow a user to virtualize a disk system to make it
look like something else. Logical links don't count in this regard, they affect
the system environment.


|>> So far, the most "open" system I can think of is a 386 running DOS:
|>> - systems available from all sorts of people
|>> - CPU's available from (at least) AMD, Intel, IBM, and Cyrix
|>
|>Well, scrap Intel from the list :-), oh and IBM since they second source
|>Intel

Not IBM's choice of course, remember Microchannel and OS/2 and what IBM tried to
pull then?

|>> - DOS available from Microsoft and Digital Research (and IBM if you count
|>>   OS/2 as a DOS replacement), with pretty much complete binary and source
|>>   compatibility between them.
|>
|>Yes, I don't like Intel, simply, 680X0 are better, but they let the
|>investment
|>slide and domesDOS is.....and do I understand this correctly that Windows NT
|>is a multiprocessing, multitasking *SINGLE* user system? Is that right or
|>did something hit me over the head whilst I was sleeping?

UNIX is a single user system too. The only evidence of a multiuser capability is

I would guess that W/NTs single user functionality should outstrip that of UNIX
fairly easily. I bet that in '94 or whenever Microsoft will be comming out with
a "Multiuser" extension. After all their continued existence depends on always
being able to offer additional functionality.

W/NT on introduction will of course have to cope with the CP/M filestore which
it inherited from MSDOS. Once pure W/NT apps become avaliable it will be
possible to use some of the more high flutin' ones. Security in a multiuser
environment is provided in the main by the filestore.  Better to deliver a
system which can run existing apps for starters and move up to better systems
than have the initial OS/2 "throw it all away approach."


--

Phill Hallam-Baker