ronald@robobar.co.uk (Ronald S H Khoo) (09/22/90)
cabbagehead@saturn.ucc.umass.edu (Chris Lloyd) writes: > Hey, anyone out there ever used NOS/VE? Ive been using and programming > on it for the last year and a half and have found it totally awesome. You mean *awesomely slow* surely ? > Any comments? Yeah, try comp.sys.cdc. NOS was *much* nicer. But that's a topic for alt.religion.computers, and I'm leaving that for now :-) -- my .signature is on holiday
steve@bat.lle.rochester.edu (Steve Swales) (09/23/90)
In article <1990Sep22.041629.4634@robobar.co.uk> ronald@robobar.co.uk (Ronald S H Khoo) writes: >cabbagehead@saturn.ucc.umass.edu (Chris Lloyd) writes: > >> Hey, anyone out there ever used NOS/VE? Ive been using and programming >> on it for the last year and a half and have found it totally awesome. > >You mean *awesomely slow* surely ? > >> Any comments? I'm afraid I have to agree with that reply. In fact, if there were a smiley face anywhere in the original post, I'd assume the person was making a joke. I've worked with a number of OS's over the years, and NOS/VE is, I'm afraid, near (if not at) the bottom of my list for good features. I'm pretty new to NOS/VE, and haven't beaten my brains out trying to fix any of this, but here is my list of BAD (and GOOD) points. BAD THINGS: 1) It is (as pointed out) 'awesomely' slow when used interactively. 2) The file naming scheme prevents using dots (e.g. for extensions) so virtually NOTHING is portable to/from other systems. 3) No wildcards... the local CYBERMEN tell me this is not strictly true, but I haven't seen evidence to substantiate their claim. I was able to toggle together reasonable copies of basic file maintenance commands quite quickly using Cybil (not sure about the spelling of that) and later using C, but these things should be build in. 4) All I/O seems to be geared exclusively for half duplex terminals or something. Specifically, nothing is accepted without a carriage return at the end. This makes writing things like 'more' impossible, and makes what would be a quite reasonable editor utterly bizarre from anything but a CDC terminal. 5) No 'SHOW USERS' or 'finger/who/w' type command that regular people have permission to run. I think this is build in, anyway. There is a command, but regular users cant run it. Maybe this is tunable, our CYBERMEN are rather conservative (they think MAIL is an unnecessary luxury!) That's enough for me... it's already worse than MS-DOS or RT11.... UGH! GOOD THINGS: 1) It DOES have a fairly straightforward hierarchical directory structure... better than VMS anyway...and (i'll really go out on a limb here...) as good as MS-DOS :-(. 2) It DOES have a really full command language... quite Algol-ish. But even this is a mixed blessing. Since you have to 'compile' it to do anything very interesting, you might as well have a 'dumber' command language, and use [insert favorite high- level language... C, in my case] for the more complex stuff, rather than inventing a whole new 'high-level' compiled language. 3) I rather like the editor, now that i've figured out how to make the thing work... those carriage returns I mentioned above had me really confused for a while. Still, I like EMACS better.... and that runs everywhere but on the Cyber. Now, I didn't use the Cyber until we switched to NOS/VE (from NOS), so I dont know which is better, or whether any or all of these bad/good points were there as well.... but, such as it is, this is my opinion of NOS/VE. -steve -- -------------------------------------------------------+"Come, Watson, come!" Steve Swales (716) 275-0265,-3857,-5101| he cried. "The game is steve@bat.lle.rochester.edu (128.151.32.111)| afoot. Not a word! {decvax,harvard,ames,rutgers}!rochester!ur-laser!steve| Into your clothes and University of Rochester 250 East River Road| come!" S.H. Laboratory for Laser Energetics Rochester, NY 14623| 'The Abbey Grange'
cabbagehead@saturn.ucc.umass.edu (09/24/90)
In article <9595@ur-cc.UUCP>, steve@bat.lle.rochester.edu (Steve Swales) writes: > In article <1990Sep22.041629.4634@robobar.co.uk> ronald@robobar.co.uk (Ronald S H Khoo) writes: > >cabbagehead@saturn.ucc.umass.edu (Chris Lloyd) writes: > > > >> Hey, anyone out there ever used NOS/VE? Ive been using and programming > >> on it for the last year and a half and have found it totally awesome. > > > >You mean *awesomely slow* surely ? > > > >> Any comments? > > BAD THINGS: > 1) It is (as pointed out) 'awesomely' slow when used interactively. I dunno, I use it on a Cyber 870 and can't complain a bit, and thats with 50 other people on.. The 830 gets pretty slow I have to admit, I've heard older versions of the OS were notoriously slow, but 1.5.1 is plenty quick. > 2) The file naming scheme prevents using dots (e.g. for extensions) > so virtually NOTHING is portable to/from other systems. ... Yea, well, all the worlds not a Unix system, but there is VX, a full Unix System V.3 subsystem that me and others have ported games (omega, hack etc..) to and can run them under NOS/VE with a little hacking. If you like porting stuff. > 3) No wildcards... the local CYBERMEN tell me this is not strictly .. The newer versions of SCL have wildcard support which I find very adequate. You just have to know how to use them. > 4) All I/O seems to be geared exclusively for half duplex terminals > or something. Specifically, nothing is accepted without ... Well, thats due to the fact all terminal IO is done by Device Interfaces (DI's) which offload the main cpu. Single character IO is Entirely possible, Ive written several programs that use it, you just have to know how. > 5) No 'SHOW USERS' or 'finger/who/w' type command that regular .. It's pretty easy to write such things, and have those who want to use it use it. I am constantly being pestered by bozo's on our Vax system because they see me on, I wish I could take myself out of the show users A LOT of the time. I have peace and quiet on the Cyber. > GOOD THINGS: > 1) It DOES have a fairly straightforward hierarchical directory > structure... > 2) It DOES have a really full command language... quite Algol-ish. > But even this is a mixed blessing. Since you have to 'compile' > it to do anything very interesting, you might as well have > a 'dumber' command language, and use [insert favorite high- > level language... C, in my case] for the more complex > stuff, rather than inventing a whole new 'high-level' > compiled language. Compile??? what the hell are you talking about..obviously you dont know what your talking about, there are several ways to run procs, attach them, put the catalog your using in you command list (Path), put them in object libraries (they get tokenized so they run faster), exetcute them as a task...all pretty easy. > 3) I rather like the editor, now that i've figured out how to make > the thing work... those carriage returns I mentioned above > had me really confused for a while. Still, I like EMACS > better.... and that runs everywhere but on the Cyber. The editor is really nice, I don't like emacs better. Utilities under NOS/VE are extremely nice. The common command language (SCL) is extremely integrated with utilities and the OS in general, the consistency in everything is *extremely*. Ive found that with new users the learning curve goes straight up then levels flat. Once you know how to get around it's a dream to use. > Steve Swales (716) 275-0265,-3857,-5101| he cried. "The game is If learn NOS/VE with an open mind you'll be a lot better off. Chris Lloyd - cabbagehead@saturn.ucc.umass.edu
g_ahrendt@fennel.cc.uwa.oz.au (Gunter Ahrendt) (09/26/90)
>From: ronald@robobar.co.uk (Ronald S H Khoo) >> Hey, anyone out there ever used NOS/VE? Ive been using and programming >> on it for the last year and a half and have found it totally awesome. > >You mean *awesomely slow* surely ? > >> Any comments? > >Yeah, try comp.sys.cdc. NOS was *much* nicer. But that's a topic >for alt.religion.computers, and I'm leaving that for now :-) I agree with the original posting, NOS/VE is the finest Operating System, and the fastest i've ever worked with. The Speed of the Full Screen Editor is incredible. As to the statement that NOS/VE is slow.....the Operating System is the most customisable you'll find anywhere, our system was sluggish until we changed the some parameters to suit our site and viola! zip.....zoom.....! We have a (NOS)/VE Dual State System too. I found the changeover very easy and user friendly.....conversion aids for everything....a very worthy successor.
rpeglar@csinc.UUCP (Rob Peglar) (09/26/90)
In article <1990Sep26.135622.2331@fennel.cc.uwa.oz.au>, g_ahrendt@fennel.cc.uwa.oz.au (Gunter Ahrendt) writes: (bunches of previous postings deleted) > > I agree with the original posting, NOS/VE is the finest Operating System, and > the fastest i've ever worked with. The Speed of the Full Screen Editor is > incredible. > > As to the statement that NOS/VE is slow.....the Operating System is the most > customisable you'll find anywhere, our system was sluggish until we changed the > some parameters to suit our site and viola! zip.....zoom.....! > > We have a (NOS)/VE Dual State System too. I found the changeover very easy and > user friendly.....conversion aids for everything....a very worthy successor. Indeed. NOS/VE is a remarkable operating system. The only, truly insurmountable problem is: (two parts) a) VE only runs on Cybers b) Cybers (present-day) only run VE Almost everything else is religion. These two points are the only reason that Cybers and VE will never, ever, nohow, noway make a significant dent on the way people - in the widest sense - compute. And it's a shame. Rob -- Rob Peglar Comtrol Corp. 2675 Patton Rd., St. Paul MN 55113 A Control Systems Company (800) 926-6876 ...uunet!csinc!rpeglar
pyue@ria.ccs.uwo.ca (Paul Yue, CCS) (09/27/90)
I've talked to the the Cyber support dude here (who also happens to know
a thing or two about emacs), and according to him, emacs should work
with the new C compiler from CDC. Unfortunately, this new compiler is
priced around CDN$15000, so we have been unable to try it out.
Locally, we can get a display of all jobs running on the system (batch, system,
interactive, etc) by using the command display_all_jobs. I guess this is a
local command? We also have an InterNet finger server that does this
disaj command at a finger request (there isn't an /etc/passwd file that it
can read names from though, and this server won't show last login times).
NOS/VE has a lot of tricks up its sleeve, that casual users never know about,
eg. job scheduling. We have our Cyber set up here to incrementally demote
a job's priority with increasing continual CPU usage. Helps keep things
moving when there's a lot of jobs going. The virtual DAS is very nice too.
We're going to be getting a Cyber 2000 soon, and our Cyber guru can't wait
to install X windows on it to give it a go.
--
-------------------------------------------------------------------------------
Paul K. Yue pyue@{uwo.ca|nve.uwo.ca|uwovax.bitnet} uunet!watmath!ria!pyue
-------------------------------------------------------------------------------
Computing and Communications Services Natural Sciences Centre, UWO
London, Ontario N6A 5B7 Canada Voice:(519) 661-3525 FAX:(519) 661-3486steve@bat.lle.rochester.edu (Steve Swales) (09/30/90)
As some of you know, I posted a moderately long list of my favorite
gripes about NOS/VE early in this thread to which there was quite a
reaction. In particular, I received LOTS of great help from Daren
Latham who works for CDC. Thanks, Daren! A few others sent helpful
e-mail (thanks, others), and I got a bit of hate mail (oh, well).
The upshot of all this is that I now realize a lot of the problems I
have encountered are the result of non-current revisions, to the
conservative way our particular site has been administered, and
admittedly to a lack of knowledge about the nitty-gritties of NOS/VE
on my part.
Several of these aspects conspired, in keeping with Murphey's Law, to
make things seem much worse than they CAN be.
Those of you who have been tempted to send me hate mail can interpret
this as a retraction. Those of you who might have more helpful advice,
please don't! While Daren (et al) provided me with lots of great work-
arounds, I can still use all the help I can get :-). One exception:
Several people have mentioned that there are ways to do SHOW USERS-
type operations. Since our system administrators have chosen not
to allow this type of operation, advice about it is of no use to me.
One thing I AM still interested in is if anyone has managed to port
Emacs to NOS/VE (as mentioned is probably possible in the last
article). While NOS/VE has a great editor already, it would still be
of great value to me to have Emacs working there, since I have it on
all of the other OS's I work with regularly (about 6 others in
addition to NOS/VE), and it is a great savings in brain strain not to
have to change editors when moving between machines.
-steve
--
-------------------------------------------------------+"Come, Watson, come!"
Steve Swales (716) 275-0265,-3857,-5101| he cried. "The game is
steve@bat.lle.rochester.edu (128.151.32.111)| afoot. Not a word!
{decvax,harvard,ames,rutgers}!rochester!ur-laser!steve| Into your clothes and
University of Rochester 250 East River Road| come!" S.H.
Laboratory for Laser Energetics Rochester, NY 14623| 'The Abbey Grange'jjn@hare.cdc.com (jj needham 234-4312) (10/03/90)
In article <PYUE.90Sep26204654@ria.ria.ccs.uwo.ca> pyue@ria.ccs.uwo.ca (Paul Yue, CCS) writes: >I've talked to the the Cyber support dude here (who also happens to know >a thing or two about emacs), and according to him, emacs should work >with the new C compiler from CDC. Unfortunately, this new compiler is >priced around CDN$15000, so we have been unable to try it out. > Having worked on the development of CV2 when it was in Canada, I can say that it is a deadly compiler (oh nice plug!) I think the Lehigh guys were porting emacs. It was surprising to see non-CDC Nos/VE customers (note homage to Nos) commenting on the OS. Having lived with the beast for so long, its interesting to see what real users think! If any of you have been reading comp.arch, they are feuding about the viability of 64-bit hardware/OS, hmmmmmm ..... later,jeff Disclaimer: | Jeff Needham If you become any more forgetful, you | Oracle Performance Group will be qualified enough to work in | Control Data - Santa Clara, CA Performance Analysis! | INTERNET jjn@hare.udev.cdc.com
jps@maria.wustl.edu (James Sterbenz) (10/17/90)
Well, its been a while since I've read this group, but seeing the NOS/VE
subject line attracted my interest. Is anyone aware of published
information on NOS/VE and/or Cyber virtual state architecture?
The only information I have is in some of the CDC manuals, which are
not very accessable to the general community. It seems a shame to
have a system so deeply rooted in Multics, and with some very
interesting architecture, be so generally unknown in the research
and academic community. I've described NOS/VE in an operating
systems class I've taught, but was unable to provide generally
available references for reading lists.
--
James Sterbenz Computer and Communications Research Center
Washington University in St. Louis +1 314 726 4203
INTERNET: jps@wucs1.wustl.edu 128.252.123.12
UUCP: wucs1!jps@uunet.uu.net