Return to BSD News archive
Xref: sserve comp.unix.bsd:8500 comp.unix.solaris:667
Newsgroups: comp.unix.bsd,comp.unix.solaris
Path: sserve!manuel.anu.edu.au!munnari.oz.au!news.hawaii.edu!ames!elroy.jpl.nasa.gov!usc!sol.ctr.columbia.edu!hamblin.math.byu.edu!news.byu.edu!ux1!fcom.cc.utah.edu!cs.weber.edu!terry
From: terry@cs.weber.edu (A Wizard of Earth C)
Subject: Re: Solaris 1.1 vs. Solaris 2.0 (BSD vs AT&T)
Message-ID: <1992Dec1.184115.8699@fcom.cc.utah.edu>
Sender: news@fcom.cc.utah.edu
Organization: Weber State University (Ogden, UT)
References: <id.U_0V.SJ3@ferranti.com> <22947@venera.isi.edu> <id.FOCV.Q52@ferranti.com>
Date: Tue, 1 Dec 92 18:41:15 GMT
Lines: 109
In article <id.FOCV.Q52@ferranti.com> peter@ferranti.com (peter da silva) writes:
>In article <22947@venera.isi.edu> allard@isi.edu (Dennis Allard) writes:
>> Unix has the following going for it.
>
>> 1. Millions of current users (NT has no users yet)
>
>In this context you have to consider the millions of Windows and even DOS
>users.
This isn't true. Most DOS applications won't run under NT without some
extensive modifications, or running under a virtual (8086) machine. Most
modern applications require more balls than an 8086, and thus require a
"port" before they will work, quoted because DOS people tend to be unable
to write portable code to save their lives -- this is NOT applicable to
all DOS types, just most of them. Fully 40% of UNIX people can write
portable code, on the other hand. ;-).
>> 2. Has been ported to more platforms than DOS or Windows (will NT
>> seek to rival Unix in that regard. If so, how???)
>
>DOS and Windows run on more computers than the total of all UNIX shipments
>on all platforms.
This has very little to do with anything. Even Intel is well aware of the
limitations of the 80x86 architecture. This is why the 80586 (renamed
"pentium") is mostly a RISC chip with a backward compatability mode. The
initial NT marketing blurbs called for NT to run on a lot of hardware it
will probably never run on.
You can't lump DOS and Windows users into the same boat. Windows is
basically for people who would otherwise by Macintosh. DOS is for people
who want applications.
The MS Windows API is easily supported on UNIX with X, and I will argue that
the trouble one has to go to to port a "DOS specific" Windows app to NT will
be about equal to the trouble it would take to port to UNIX, givwn that the
interface issues are now irrelevant.
>> 3. Supports a network based windowing system. Will NT? (The Windows
>> API is not designed to work over a network, to my knowledge)
>
>Does the target market care?
Yes, very much so (at least this is what I believe). The inability to
distribute applications in this way basically cripples NT's use of "compute
servers" without addition hardware and software. This is because the idea
of a "compute server" requires a more powerful box to run the computation
than the end user has on their desk (ie: not an NT box) and SMB is ill
suited to large network/mainframe/supermini environments. Tell me you aren't
using TCP/IP, DECNet, XNS, or IPX to connect to the news service you are
reading this from.
>> 4. In particular, supports X windows. Will NT support X windows?
>
>Yes, though I can't imagine many people will care.
People care because X alleviates the same drawbacks that have kept NeXT from
taking over the computer market entirely:
1) The inability to run applications non-locally.
2) The inability to adequately divorce an application from a platform
running compute intensive display services (thus drastically
reducing the throughput of the applications).
3) The ability to have decent display services without paying Adobe
lots of money (for all the benefit Adobe has brought to printing,
I believe this has hurt NeXT more than any other single thing).
NT is just another NeXTStep, but crippled by having to grandfather the DOS
Windows Interface.
>> 5. Freeware versions being worked on by quite compentant people. Will
>> NT have freeware versions?
>
>Does the target market care?
I agree here; the target market doesn't give a damn if it's free; if it did,
there wouldn't be any margin in pursuing it.
>> The typical criticisms I here about Unix are about things such as
>> cryptic commands and lack of a User Guide. I make those criticisms
>> myself. However, such things are easy to fix. Rarely does one hear
>> more substantive criticism about the OS architecture.
>
>Naturally, The problems of UNIX are easy to fix. But nobody has actaully
>gone and fixed them. I doubt anyone will.
I think that NeXTStep, Destiny, and Univel's UNIXWare have all done a lot
in this direction. UNIXWare, for instance, comes up as a graphical
interface on initial installation, and is never text unless you choose some
tty application (like an xterm). This, and the ability to be either a
NetWare client (like DOS) *or* a NetWare Server makes UNIXWAre a strong
contender in the (as yet unreleased NT) market.
I think there will be a shakedown, but I don't think NT is ready to compete
head-on in the UNIX marketplace, and UNIX is very close, if not already
able to compete in NT's marketplace. NT will probably compete head on
with MS Windows, and only somewhat (say 30%) with DOS.
Terry Lambert
terry@icarus.weber.edu
---
Any opinions in this posting are my own and not those of my present
or previous employers.
--
-------------------------------------------------------------------------------
"I have an 8 user poetic license" - me
Get the 386bsd FAQ from agate.berkeley.edu:/pub/386BSD/386bsd-0.1/unofficial
-------------------------------------------------------------------------------