fr@icdi10.uucp (Fred Rump from home) (12/07/88)
In article <102@epsilon.UUCP>, jlg@epsilon.UUCP (My account ) writes:
<
< A side note to anyone purchasing UNIX/XENIX for 286 and 386. The spec
< being used by NASA for the space shuttle computers is (are you ready?)
< S.C.O. XENIX 386. A company I consult with has purchased both the 286 and
< 386 versions for my project there, and I am quite satisfied with BOTH
< versions of the product. Both are at rev 2.2.3. The only complaint I have
< at all is the V.P.I.X. for the 386 (It is a serious performance drain even
< though it seems to do what it was advertised to do).
<
< -- John Grzesiak --
< @ Omega Dynamics
Well, now we'll know why if another shuttle crashes, blows up or otherwise
vanishes.
Seriously, I don't know why the surprise though. The Navy has SCO on it's
ships. (We're involved in that little effort) And the Army uses Xenix all
over the place. So what else is new?
VP/ix is slow as a task. Yes. You really need a 20 or 25 Mhz box to look like
an AT. And to run a bunch of DOS jobs is not the goods. It's really meant to
be the occasional task to prevent Xenix from having to be taken down as
before. But it works: Halleluja!
--
{allegra killer gatech!uflorida decvax!ucf-cs}!ki4pv!cdis-1!cdin-1!icdi10!fr
26 Warren St. or ...{bellcore,rutgers,cbmvax}!bpa!cdin-1!icdi10!fr
Beverly, NJ 08010 or INTERNET: fred@cdin-1.uu.net or fr@icdi10
609-386-6846 "Freude... Alle Menschen werden Brueder..." - Schillerdavidsen@steinmetz.ge.com (William E. Davidsen Jr) (12/10/88)
In article <7113@icdi10.uucp> fr@icdi10.uucp (Fred Rump from home) writes: | VP/ix is slow as a task. Yes. You really need a 20 or 25 Mhz box to look like | an AT. And to run a bunch of DOS jobs is not the goods. It's really meant to | be the occasional task to prevent Xenix from having to be taken down as | before. But it works: Halleluja! I think this comment may give people the wrong impression. VP/ix pays a penalty when updating the screen, without a doubt. Programs which do a lot of writes to the screen, particularly those which bypass the BIOS, will run a lot slower under VP/ix. However, programs just uning the CPU, such as compiles, database lookup, and data analysis will run at full speed. Programs which have both screen and CPU characteristics, such as large spreadsheet recalc, will slow in proportion to the screen usage. Taking this as a premise, a few measurements of CPU performance under VP/ix and raw DOS will show (by my measurements, anyway) that this assumption is correct. Disk i/o is may be slightly slower or faster depending on the buffer sizes, etc. The reason that BIOS driven screen writes are affected less than direct writes is that (a) they are slower to start with, and (b) the VP/ix BIOS is no worse than the DOS BIOS. I ran some CPU tests, from Dhrystone to the Microway benchmark (and of course Norton SI) and they seem to reflect the same performance in DOS and VP/ix, within the limits of the measurements. I actually find that the only thing which is really obnoxious under VP/ix is games. Does anyone else like PCPOOL? -- bill davidsen (wedu@ge-crd.arpa) {uunet | philabs}!steinmetz!crdos1!davidsen "Stupidity, like virtue, is its own reward" -me
hsu@santra.HUT.FI (Heikki Suonsivu) (12/16/88)
In article <12779@steinmetz.ge.com> davidsen@crdos1.UUCP (bill davidsen) writes: >will run a lot slower under VP/ix. However, programs just uning the CPU, >such as compiles, database lookup, and data analysis will run at full Unless everything is in the buffers, dos compiles and all other disk access seems to be something like 50% of speed in raw ms-dos. May be different if one is not using unix file system? Though, I'm running dosmerge 0.2 beta. Linking is even slower.