david@torsqnt.UUCP (David Haynes) (03/10/91)
tim@ohday.sybase.com (Tim Wood) writes: >Here's my vote that Sequent designate Mach (or OSF/2) as their standard >operating system, and layer BSD and System V guests onto it. Would this >be doable in, say, three years? >Just trying to broaden the discussion, >-TW This is interesting.... A couple of points come to mind. 1. You can get Mach for the Sequent systems now and have been able to get it for at least 3 years that I know of. 2. I have never been asked by a university for it. 3. Right now, I wouldn't say that Mach was a commercially viable alternative to UNIX. While systems such as the NeXt are running it successfully, they have the advantage of a tightly defined hardware environment. In a more general environment, there are more choices and therefore more places for errors to occur. 4. Discussions I have had with some folks from CMU indicate that Mach needs a lot of work under the hood, before even they will feel comfortable with it. 5. I spend a good portion of my day hearing people say things like, "We've decided to standardize on COBOL. We are now converting all our FORTRAN code to the standard." Do you really think these folks are ready for a new O/S? 6. Please notice that I have not said that Mach is not a technologically viable O/S or that Mach does not have its place in the universities, and eventually the commerical market place, but I do question the timeframe outlined above. These opinions are definitely my own and in no way, shape or form should be thought to reflect Sequent philosophy, policy or corporate direction. I have enough troubles... ;-) -david- -- David Haynes Sequent Computer Systems (Canada) Ltd. david@torsqnt.UUCP ---------------------------------------------------------------------------- Sometimes I think the world is filled with mindless sheep constantly bleating "MIIIIPPPPPSSSS, MMMMIIIIIPPPPPSSSS" until they are lead to slaughter.
zoo@aps1.spa.umn.edu (david d [zoo] zuhn) (03/11/91)
>>>>> On 10 Mar 91 13:22:06 GMT, david@torsqnt.UUCP (David Haynes) said:
David> 1. You can get Mach for the Sequent systems now and have been able to
David> get it for at least 3 years that I know of.
David> 2. I have never been asked by a university for it.
This doesn't surprise me at all. We heard about Mach for the Symmetry
when I was working on the systems staff of the local CS dept, which
runs an S27 as the departmental backbone machine. We started to look
into Sequent's Mach, and what we saw scared us:
1) Sequent said that they would not support anything on that OS. We
could not afford to have an unsupported machine as the backbone.
2) No NFS at all. Another *VERY* negative point for our site.
3) A fairly high cost for unsupported software (the number $15,000
sticks in my mind.
Things may have changed much in the couple of years since these first
looks into Mach. I somehow doubt it, given the rate of change in
Sequent OS releases.
NOTE: I no longer work for the above mentioned CS dept, and am in no
way, shape or form speaking for them or expressing an official opinion
of the CS dept. I don't even think we formally brought up the idea
with the head honchos, given the lack of support and NFS.
david d [zoo] zuhn Univ. of Minnesota Dept. of Astronomy
zoo@aps1.spa.umn.edu Automated Plate Scanner Projectgustav@arp.anu.edu.au (Zdzislaw Meglicki) (03/13/91)
In article <ZOO.91Mar11005948@grumpy.spa.umn.edu>, zoo@aps1.spa.umn.edu (david d [zoo] zuhn) writes: |> >>>>> On 10 Mar 91 13:22:06 GMT, david@torsqnt.UUCP (David Haynes) said: |> |> David> 1. You can get Mach for the Sequent systems now and have been able to |> David> get it for at least 3 years that I know of. |> David> 2. I have never been asked by a university for it. That's strange, because I represent the University (the Australian National University) and I have asked for Mach on the Sequent. What more, I have gone through all the trouble and expense of getting AT&T UNIX V R4 license that was required by the Sequent, and in spite of numerous e-mail messages I still didn't get the tapes. In the meantime, I've been told by David Golub from Carnegie Mellon, that they actually have Mach 3.0 single-server BSD 4.3 system running on their Sequent. Sequent itself provides only Mach 2.5 with DYNIX instead of BSD 4.3 put on top of it. Their, i.e., Carnegie's version "doesn't fit Sequent's list of bugs, but then many of those listed involve Dynix/Mach compatibility, which CMU is not interested in." (D.G.) Since our S27 is configured mostly as a research machine, and we don't have any DYNIX based commercial software installed, a system such as that from Carnegie Mellon would suit me just fine. I've written to D.G. asking for more details and a possibility of getting the installation from them (19th Feb), but I didn't get any information back yet. I am not exactly paranoid, but it smacks of a conspiracy... Am I ever going to see Mach on my S27, I wonder? (In the meantime it's been turned into a GNU machine - at least that I can get via ftp). -- Gustav Meglicki, gustav@arp.anu.edu.au, Automated Reasoning Project, RSSS, and Plasma Theory Group, RSPhysS, The Australian National University, G.P.O. Box 4, Canberra, A.C.T., 2601, Australia, fax: (Australia)-6-249-0747, tel: (Australia)-6-249-0158
gustav@arp.anu.edu.au (Zdzislaw Meglicki) (03/13/91)
Aha, one more thing... Someone suggested here in comp.sys.sequent that Sequent should embrace OSF/2. What Sequent could do is to provide two alternative OSs - much like DIGITAL (VMS and Ultrix) or CRAY (once upon a time). A system such as OSF/2 may turn to be highly portable and with much of the work already done by OSF maintenance on the side of the Sequent could be much reduced. Also, since OSF/1 is based on Mach 2.5, and Sequent has already Mach 2.5/DYNIX combination working (although not without problems), to offer OSF/1 or 2 should be that much easier. -- Gustav Meglicki, gustav@arp.anu.edu.au, Automated Reasoning Project, RSSS, and Plasma Theory Group, RSPhysS, The Australian National University, G.P.O. Box 4, Canberra, A.C.T., 2601, Australia, fax: (Australia)-6-249-0747, tel: (Australia)-6-249-0158
csg@pyramid.pyramid.com (Carl S. Gutekunst) (03/14/91)
>A system such as OSF/2 may turn to be highly portable and with much of the >work already done by OSF maintenance on the side of the Sequent could be >much reduced. Haw! Haw! Haw! Haw!!!! *Guffaw* *Gasp* *Giggle* Oops, I better restrain myself. Never worked for a system vendor, I see. :-) <csg>
shwake@raysnec.UUCP (Ray Shwake) (03/16/91)
gustav@arp.anu.edu.au (Zdzislaw Meglicki) writes: >Aha, one more thing... Someone suggested here in comp.sys.sequent >that Sequent should embrace OSF/2. What Sequent could do is to >provide two alternative OSs - much like DIGITAL (VMS and Ultrix) >or CRAY (once upon a time). A system such as OSF/2 may turn to >be highly portable and with much of the work already done by OSF >maintenance on the side of the Sequent could be much reduced. Might one be so blunt as to point out that OSF/2 does not exist! Heck, OSF only released OSF/1 last November. Realistically, don't expect OSF/2 for another 18-24 months, perhaps another year beyond that for it to appear as a robust, commercial product. ----------- uunet!media!ka3ovk!raysnec!shwake shwake@rsxtech
jonathan@cs.pitt.edu (Jonathan Eunice) (03/25/91)
In article <266@raysnec.UUCP> shwake@raysnec.UUCP (Ray Shwake) writes:
gustav> Aha, one more thing... Someone suggested here in comp.sys.sequent
gustav> that Sequent should embrace OSF/2. What Sequent could do is to
gustav> provide two alternative OSs - much like DIGITAL (VMS and Ultrix)
gustav> or CRAY (once upon a time). A system such as OSF/2 may turn to
gustav> be highly portable and with much of the work already done by OSF
gustav> maintenance on the side of the Sequent could be much reduced.
shwake> Might one be so blunt as to point out that OSF/2 does not exist!
shwake> Heck, OSF only released OSF/1 last November. Realistically, don't
shwake> expect OSF/2 for another 18-24 months, perhaps another year beyond that
shwake> for it to appear as a robust, commercial product.
I'd be *real surprised* to see OSF/2 in 18-24 months. This mythical
beast is nowhere close. OSF is much more likely to release
incremental systems, some of which may partially adopt the true
micro-kernel approach. Specialized server implementations, for
example. At least, this is what Ira Goldstein et al were saying at
Uniforum this Janurary. At this point, OSF/2 is identical to Microsoft's
NT -- micro-kernel-vaporware.