ronb@uunet.uu.net (Ron Baxter) (01/28/89)
We have a Sun3/160 fileserver with diskless nodes, VDU's and communication lines. It is overloaded and we are looking to upgrade it -- possibly to a Sun 4/260 that is well endowed with memory and disk capacity. However, we have just seem some specs for this Solbourne Computer (just announced, I believe) which has a Fujitsu SPARC CPU together with SunOS/SunView/NeWS etc licensed from Sun. So it looks like an alternative. Now we KNOW that the 4/260 would slot into our system with minimal upheaval, and we would expect to be able to deal with any problems that arise. My apprehension about the Solbourne is that maybe not everything works smoothly yet, or there may be some compatability problems with our existing equipment, or that some of the application software we take for granted may not go without a lot of pain, etc, etc. (For instance, the glossy doesn not mention availability of serial lines other than the 2 that are there, but we need to support several dialin/dialout modems.) So if anyone has investigated it, or had hands on experience, or stories to tell that may allay my fears (or reinforce them) I would be very grateful for the information. Thanks, and I will summarize anything I get. -- Ron Baxter, CSIRO Div Maths & Stats, PO Box 218, Lindfield, NSW, Australia. Phone: +61 2 467 6059 Email: ronb@natmlab.oz.au
chuck@trantor.harris-atd.com (Chuck Musciano) (02/03/89)
> However, we have just seem some specs for this Solbourne Computer (just > announced, I believe) which has a Fujitsu SPARC CPU together with > SunOS/SunView/NeWS etc licensed from Sun. The Solbourne is priced about 25% lower than the equivalent Sun machine, and is completely binary compatible. It runs a licensed copy of SunOS, right from Sun. Consider this, though: Sun is not stupid. They would not license their product to another company who is undercutting them by 25% unless they were ready to roll out significant price/performance improvements in the near future. Wait for the February product announcements that were hinted at at the Miami SUG. I predict you'll find a better, cheaper machine to meet your needs. Chuck Musciano Advanced Technology Department Harris Corporation (407) 727-6131 ARPA: chuck@trantor.harris-atd.com
stan@shell.UUCP (Stan Hanks) (02/09/89)
I am joining Solbourne as the corporate Science Advisor effective 7
Februrary 1989. I have a charter to figure out what's going on in the
research community that will be important in the future, what's missing
from current SunOS that would be most useful to the most people, and to
help preach the gospel of Sun Compatibility and show people the Way of
Distributed Computing.
I have been a long-time Sun user (since the Sun-1 running UNIPlus
V7/SysIII), heavily involved with many asspects of Sun usage (I started
the Sun OEM Users Group and chaired it for 2 years, help start the
regional Sun Users Group, etc), and a big proponent of Sun as the Right
Solution. Not to mention having written tons of code for these suckers!
I wouldn't have taken the job if I wasn't confident that the product was
on the money technically and that the company was on target business-plan
wise. The stated objective is "complete Sun-4 compatibility". If
something works on a Sun-4, you ought to be able to just take *THE BINARY*
and run it on a Solbourne, with no problems. If you can't, the factory
needs to know about it. To date, very few things that work on Sun-4's
haven't worked on Solbournes. Each time, it was due to a bug that has
since been fixed.
The issue of added peripherals (like SMD disks, 1/2" tapes, and serial
lines) has taken a back seat to getting product visibility and
availability. It will be addressed in the future. As there is a VME bus
with about a dozen available slots, you can add your own Dingbat 915
interface module if you just can't live without it, same as on a Sun.
I do know that the drivers will NOT be the same as Sun drivers, but don't
yet know what is different or how different they will be. But hey, my
first day isn't for another week, and I don't have the hardware yet!!!
I will be available by phone or e-mail from the time I join. I haven't
worked out some of the logistics yet, but I will post them when I have
them. In the mean time, all mail to "stan@rice.edu" or "stan@shell.uucp"
will be correctly forwarded.
Let me know what you think, if you've seen/used a Solbourne. If you'd like
to, let me know that too.
Stan Hanks
Research Computer Scientist, (and Postmaster!)
Shell Development Company, Bellaire Research Center (713) 663-2385
...!{sun,psuvax1,bcm,rice,decwrl,cs.utexas.edu}!shell!stan stan@rice.edulayer@ucbarpa.berkeley.edu (Kevin Layer) (02/09/89)
munnari!natmlab.dms.oz.au!ronb@uunet.uu.net (Ron Baxter): >... > So if anyone has investigated it, or had hands on experience, or stories > to tell that may allay my fears (or reinforce them) I would be very > grateful for the information. I recently verified that Allegro CL works (the Sun4 version), without modification, on Solbourne's machine(s). Over the years, Allegro CL (and Franz Lisp, before it) has had the pleasure of `highlighting' problems in software said to be "upward" compatible (i.e., the infamous "OS upgrade"). Nothing exercised by Allegro CL (which is quite a lot) showed any non-conformance to the SPARC architecture or SunOS 4.0. Kevin Layer Franz Inc.
siedelbe@stout.ucar.edu (Mike Siedelberg) (02/11/89)
chuck@trantor.harris-atd.com (Chuck Musciano) writes: >... >Consider this, though: Sun is not stupid. They would not license their >product to another company who is undercutting them by 25% unless they >were ready to roll out significant price/performance improvements in the >near future. Wait for the February product announcements that were hinted >at at the Miami SUG. I predict you'll find a better, cheaper machine to >meet your needs. Since my direct reply to the original message bounced I will try this route. I recently had a tour of the Solbourne facility in Longmont which is ten minutes away from Boulder. The Solbourne machine is not a direct clone of the standard Suns. Solbourne uses their own proprietary internal bus for their major boards ie. CPU, memory, ethernet, SCSI, frame buffer, and a couple of serial ports. There are also other optional things like color, which will be available. Included too is a seven slot VME standard chassis, which is their way to outer world upon which they have already put on an SMD controller (not part of their current line yet, however), a second ethernet controller (same as SMD, but supposedly much faster I/O wise that the standard ethernet), and who knows what other gadgets. Solburne has a seven year agreement with Sun for future support and transfer of updated SPARC technology. Another company that has a Sun SPARC license is Cypress, who is developing a 20MIPS SPARC chip set. Both Solbourne and Sun (I believe) will be using these chips to offer a 20MIPS SPARC machine, which is part of the new product announcements referred to above. According to everthing I can find out, Sun is granting licenses to other Companies, so that they can do exactly that; sell at a cheaper price. Sun gets royalties and at the very least they get more of their type of machines out in the world. The Solbourne machine is enough different though, that it may not be the machine for you. For instance, Solbourne does not, nor is there any current plan to, support Sun format of boards. So if you intend to use the ALM, MCP, or TAAC (all of which I currently use) plus any others that might come from sun, you are SOL. Also right now they do not have any SMD support. And so on. What they do offer is the ability to operate a 2nd CPU on their K-bus as a slave only to the master CPU. If anyone is interested in firsthand info, Solbourne has hired at least 2 salespersons to answer questions (and sell you things :-) ). The one I know is: Don Geiser (303)741-0020 ps. the CURRENT rumor about the release of impending Sun products is that nothing will be announced until around Mar-Apr and then no products out until 60 days after that.
jim%applix@harvard.harvard.edu (02/23/89)
In article <8901302316.AA02200@snooze>, franz!snooze!layer@ucbarpa.berkeley.edu (Kevin Layer) writes: > > So if anyone has investigated it, or had hands on experience, or stories > > to tell that may allay my fears (or reinforce them) I would be very > > grateful for the information. We just recently took our office automation package, Alis, up to Solbourne to test its compatibility and our Sun-4 code ran like a champ. There was one issue with their new style keyboard not using the same key station codes for the keyboard ioctl's, but they put a Sun keyboard compatibility module in the kernel and everything ran fine. Nice box... Jim Morton, APPLiX Inc., Westboro, MA UUCP: ...harvard!m2c!applix!jim jim@applix.m2c.org