NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/comp/os/coherent/6491

Text File | 1992-12-29 | 5.2 KB | 109 lines

Newsgroups: comp.os.coherent
Path: sparky!uunet!spool.mu.edu!cass.ma02.bull.com!mips2!bubba!sje
From: [e-mail] (Steven J. Edwards)
Subject: I predict ...
Reply-To: [e-mail]
Organization: Bull HN, Worldwide Information Systems, Billerica, Mass., USA
Distribution: comp
Date: 28 Dec 92 17:08:04
Message-ID: <[e-mail]>
Sender: [e-mail] (Usenet News Manager)
Lines: 96

Eenie, meenie,
    Chili beanie,
        The spirits are about to speak!

For 1993, I predict:

1) Coherent will not be sold extensively in the dealer retail channel,
particularly in the United States.  This is because of the added costs
associated with dealer distribution (e.g., mark up) along with the
fact that many dealers dislike direct competition with the factory
supplier.  It could appear as bundled software in some cases, however,
particularly with mail order systems.

2) Both GCC and CFront (via Comeau C++) will appear prior to the end
of the first quarter.  Both will be roughtly equally well received as
GCC users discover the dreaded "GNU Bloat" will cost them in extra
disk and memory resources as much as the additional price for CFront.
The old MWC C compiler will continue to be bundled if only to make for
good ad copy.

3) Hardware FPU support will arrive, but will not be fully integrated
into the system executables so as to minimize distribution cost and
complexity.  Compilers and assemblers will generate either FPU or
alternative software simulation depending on environment variable
values.  Sadly, the extra time needed to save FPU registers during
context switching will slow Coherent interrupt processing to be closer
to regular SCO rates.

4) The MWC BBS support staff will correct an interesting security hole
in the mwcbbs that allows spoofing by any Coherent user of any other
Coherent user on the Maillist file.  (MWC ONLY, feel free to contact
me for details.)

5) Virtual memory will come sometime before the end of the third
quarter.  By then, the roar of discontent from GNU users with less
than 4 Mbyte RAM will become impossible to ignore.

6) Shared libraries will appear shortly after virtual memory because
disk space will then become the limiting factor for the victims of
software obesity.

7) Sockets and IP network functionality will be slow to arrive because
of the great variablility of third party network cards.  Indeed, the
real driving force is not socket/IP programming per se, but the needs
of the X Window system.  Some clever company will make a good chunk of
money by selling plug-and-play hardware/software packages for Coherent
users to connect to the Internet.  Also concerning networking, and
much to my personal dismay, LocalTalk will not be officially supported.

8) The X Window system will be ported, but only after the above listed
prerequisites appear.  The loud noise head shortly thereafter will be
the sound of thousands of Coherent users tossing their IDE cards and
drives into the junkpile and lining up to buy 500 Mbyte SCSI systems.
Coherent will then be accused of just being a front for the HD
manufacturing cartel, much like MS Windows was seen as a scheme of the
RAM manufacturers.

9) The 24 line limit on the console driver will remain for no good
reason.

10) Microsoft will continue to ignore the MS-DOS users that also run
Coherent; MS-DOS will still need special treatment with respect to
disk and partition positioning.

11) Gains in sales for MWC will be offset somewhat by a greater load
on technical support resources as the general buying audience
demographics change.  While most of the first 50 thousand or so
Coherent users may be technically savvy to some extent, I doubt that
this will be generally true of users to come.  Most of the real "Unix
hacker" crowd has already tried Coherent and the future customers will
need more hand holding than those of the past.

12) Some sharp entrepreneurs will start a Coherent magazine and will
begin to make money after a year or two; it will take that long for
a sufficient number of third party advertisers to appear.  Possible
titles: _Adherent_, _We're Not USL_, and _Poor Man's Multitasking_.

13) An "official" graphics standard will be provided and supported.
However, it will only get commercial recognition if it is done in a
general, yet comprehensive fashion so as to motivate third party
commecial application software ports.  A close look at VGA graphics
accelerators may be of considerable benefit.

14) And so not to annoy the tridekaphobics, one last prediction: One
year from now, it will be clear that Coherent will be either a big
winner or a big blowout.  The deciding factor will not be the low
buy-in price, nor will it be the offical MWC support, nor will it be
the unofficial community support.  The deciding factor will simply be
the number and magnitude of well known third party commecial
applications.

 [The above opinions expressed are my own; not necessarily held by others.]
      == Steven J. Edwards           Bull HN Information Systems Inc. ==
      == [phone]              300 Concord Road         MS 820A ==
      == [e-mail]     Billerica, MA 01821          USA ==
"That Government which Governs the Least, Governs Best." -- Thomas Jefferson