Re: Ressurecting an IPX
[email protected] (Sandwich Maker) Mon, 14 Mar 2011 17:17:58 -0400 (EDT)
| Newsgroups | gmane.os.solaris.at-home |
|---|---|
| Message-ID | <[email protected]> |
" Date: Tue, 15 Mar 2011 06:59:33 +1100 (EST) " From: Craig Dewick <[email protected]> " " As a side-note, I still have a Solbourne 5/502 system which has a 2 gb HD " with OS/MP 1.something (derived from SunOS 4.1.3 or 4.1.4 I believe) yup. iirc 1b = 4.1.2, 1c = 4.1.3, there was no 1d. " on it. A true multi-processor/multi-tasking version of SunOS 4! Shame about " the company disappearing into the ether when Northrop-Grumman got hold of " it. And sourcing spare hardware, etc. is another one of those holy-grail " type searches. indeed! my only experience of solbourne systems came just at the time grumman was taking over, and i heard of a bunch o' good things that just got landfilled - like the extended backplane with enough slots for max ram -and- max cpus... i thought at the time that they were missing a bet in not porting the solbourne mp changes to sunos over to sun-std sun4m system arch, but perhaps there were demarcation or non-compete clauses... later at another place, i ran 4.1.4 on dual-supersparc ss20s and found it ran excellently, but its mp tech originated with ross for their hypersparc cpus and worked crudely by basically putting a spinlock around the whole kernel. [this was just as well for us, as we had one big honking single-threaded app that sucked up all the cpu cycles; giving it its own cpu meant engineers could continue working on other tasks - like setting up for the next run of the app] solbourne mp reworked the kernel more extensively with finer-grained spinlocks. our solbourne was something of a problem child and constantly had problems. i got to know the fe pretty well that summer... ________________________________________________________________________ Andrew Hay the genius nature internet rambler is to see what all have seen [email protected] and think what none thought