[net.cms] UNIX over CMS anyday

jf@sri-unix (06/01/82)

I stumbled upon the UNIX vs CMS discussions by accident (not being
a regular reader of net.cms), but I am also facing an O/S controversy
which will include comparisons of UNIX with VM/CMS and other helpings
of IBM alphabet soup.  So I would definitely appreciate any comments
(since I have not used CMS very much, I could use some help on my
offensive strategy [which is anti-CMS]).  

Speaking as an experienced UNIX/TOPS/MCP/SCOPE user (in that order)
I was shocked at some of the omissions in CMS.  At the Berkeley
Computer Center, MTS was considered, but some requirement to contribute
to an MTS user software pool was felt to be too stringent.  Since
the Berkeley Computer Center was (and still is) a heavy UNIX shop, I
am surprised that they were able to put up CMS.

Also, I have only seen net.cms from article 7. If anyone has the back
issues of net.cms, I would appreciate it very much if they mailed a
copy to me. Thanx.
				jrfalcone
				hplabs!jf

jf@sri-unix (06/01/82)

I stumbled upon the UNIX vs CMS discussion on net.cms (not being a
regular reader of it) and I am very interested to here from users of
both systems as well since I will be facing an O/S controversy which
will include comparisons of UNIX with VM/CMS and other portions of
the IBM alphabet soup.  I am particularly interested in the omissions
of functionality in CMS as well as any unique features it might have.

Speaking as an experienced UNIX/TOPS/MCP/SCOPE user, I was shocked 
to see what is missing from CMS, especially since it had been represented
to me by my former superiors at the UCBerkeley Computer Center as the
greatest thing since sliced bread.  Berkeley's conversion to IBM for
mainframe computing needs was particularly surprising since they had
been largely a UNIX shop, and some had suggested a network of VAX 780s.
Once the VAX had been more or less eliminated, the big dispute settled
on the O/S and they picked CMS over MTS allegedly because an MTS license
required some contribution to a user software library.  Whether the reason
was valid or not, I am amazed that the UCB Computer Center (not to be 
confused with the EECS dept!) went with CMS when they had a user community
that was accustomed to UNIX conveniences.  If anyone knows of distinct
CMS advantages, I would like to hear about them.  I personally don't
believe that any possible flavor of CMS could hold a candle to
Berkeley (EECS) VAX UNIX.  It is ironic that the leading VAX UNIX
shop is also a 4341 CMS shop.

Also, I missed the early articles of net.cms (before article 7).
If anyone has a copy of these, I would really appreciate their sending
a copy to me.

				JRFalcone
				HP Labs
				hplabs!jf

geoff (06/02/82)

Does anyone at Berkeley know why CMS was picked over UTS
(Amdahl's UNIX) or maybe Bell's 370 UNIX?

mark (06/02/82)

I believe Berkeley also runs UTS - this is one of the advantages of VM,
you can run several operating systems at once.  Bell's version of UNIX
for the 370 has not been released yet.

G:cliff (06/03/82)

Amdahl's UTS is in limited use here at the Berkeley Computer Center.  There
are two virtual machines running it, one for computer center staff and
the other for the computer science department to fiddle with.  We (CC staff)
have not done extensive evaluation of it that I know of, but at first
glance it has several problems.
	-It communicates well only with IBM 3270 type terminals, whose
         architecture prohibits 'raw mode' type communication.  All
         existing UNIX full screen programs (most sadly vi) are 
         completely impossible to bring up.  There is a full screen 
         editor, but it is not anywhere near as useful as vi.

        -Even 'lint'ed C programs generally require modification to run
         under UTS.  For some reason the Amdahl people have changed some
         specs of some system calls, including one of the ones to do
         with time.

        -From our point of view, it is sufficiently different from the
         UNIX that our people use all the time that it would require a
         lot of effort towards user-education to train them.

I should point out we have really done very little it, and I am not one
of the people with a say in our policy making.  I think that there are
a couple of people in the Computer Science dept who may be doing
something useful with their machine.  
     We have not run significant benchmarks or tests to see how CMS and
UTS bog eachother down (that I know of).
       -Cliff 

G:cliff (06/03/82)

ps  I think CMS was picked over UTS because of the enormous amount of
     software that is supported and running under CMS.  SPSS, SAS, SIR
     FOCUS, IMSL, etc.....  There is a segment of campus that wants
     IBM specific software available somewhere, as well as our hordes of
     UNIX users.