🚀 go-pugleaf

In article <1993Jan22.175833.5348@netcom.com>, steveth@netcom.com (Steve Thomas)
writes:

|>In article <1993Jan19.122854.22863@jyu.fi> HOUGHTON@jylk.jyu.fi (DAVID
|>HOUGHTON, LK) writes:
|
|>>	c) SUNs are not the way to go
|>
|>Well, there's DEC, IBM, HP, and Silicon Graphics/MIPS versus Sun.  Sun is
|>out in left field, as far as standards go.  Sun is behind the curve in
|>the technology department--their processor architecture is getting old.  IBM
|>and HP look like their going to play tag with DEC's benchmarks, but I don't
|>see Sun anywhere in sight.  Sun's got a lot of market share right now--but
|>I look for it to erode, becuase they really have nothing going for them
|>EXCEPT that market share...

Its the old bandwaggon argument. Suns bandwaggon is one that is starting to slow
down. If you asked what to buy a year or tow years ago I would have said Sun.
Today they not only haven't got the beef they seem to have been taken over by
the type of suits that made DEC a bad buy in the mid eighties.

Unless Sun can deliver a new SPARC architecture Intel are going to eat their
market share for lunch. The high end of the PC market is competitive with the
SPARC and will get more so with the Pentium. Windows NT will give them a far
better operating system and there will be all that old MSDOS software avaliable
still.


|>>	d) OpenVMS is better than OSF
|>
|>Well, this is going to be difficult.  I say this knowing both operating
|>systems very well (especially VMS).  VMS, by standards of quality, is way
|>better than OSF.  VMS does what it does really well--without a hitch in
|>mission-critical environments.  VMS also tends to be overkill for simple
|>workstation environments.  Software is more expensive (although it tends to
|>be of higher quality) under VMS--and there's less public domain stuff out
|>there for it.

Not really relevant, every significant package is avaliable for VMS. Even on
alpha the vast majority of the GAWKs, XVs, XRNs and other freebies are already
there. The only significant missing piece at the mo is GNU C and thats just a
minor linker problem, since most VMS alphas are shiping with field test DEC C
its not surprising that nobody has got round to doing a port yet.

|>  Personally, I think VMS is great, and I use many of its
|>advanced features.  For the uninitiated, though, VMS is a hard sell.  If you
|>have a shop with lots of VMS in it, then I think you'll find that you can
|>sell VMS (because the expectation of quality is very high--and only VMS can
|>deliver it). 

My experience is that VMS sells itself with the users. The real selling point
though is the ease of system management. It is possible to manage a very large
VAX system with almost no system management beyond adding and twiddling with
user quotas. The system managers can operate at the level of improving the
system and helping the users rather than struggling to keep the system going at
any level at all.

|> OSF, quite simply, breaks a lot (compared to VMS).  I've talked
|>to lots of people that had never known VMS deny this up and down and tell me
|>that they have no problems with thier operating environment--it's all about
|>what you're used to.  I once talked to a developer from Microsoft.  I
|>mentioned
|>to him the their "Quick-C" development package just plain didn't work.  He
|>quickly retorted: "But I ONLY need to reboot about once an HOUR!".  Different
|>planet... (My VAXStation needed to be rebooted a MONTH ago--hit the power
|>switch by accident...).

We have had VAXes run for months and even years without problems in a real time
environment. If you don't give users privs then VMS crashes are extreemly rare.
I have never used a UNIX system of which the same could be said.

Also VMS gives a better upgrade path to Windows NT which will clobber UNIX.


--

Phill Hallam-Baker