VMS, UNIX, etc.

[Barry Shein]

Kilroy extols the virtues of VM/CMS over Unix or VMS...

The claim that CMS is fast is a red herring being as Unix runs on the
same hardware as CMS (even concurrently with CMS, you can be logged
into both), unless you're referring to something you're not
explaining.

CMS has its virtues and its vices. One vice has been IBM's insistence
on treating VM and CMS as a second-class product, this has been
getting better recently I hear. One can't really argue with things
like not introducing SNA/VM until years after the MVS version was
available. Then again, not too many of us probably have any use for
SNA, so it may not be that important except as a symbol of IBM's
relative priorities internally. I think anyone who deals with IBM
knows that VM/CMS has always been a "nuisance" product to them. Sort
of like CMS is to MVS in IBM what Ultrix has been to VMS in DEC.

System administration under CMS can be a nightmare. For example, disk
space is allocated in a fixed manner to each user account when the
account is created. This means, for example, if you get 10MB of space
you get all of it committed to your account immediately and to get
more they have to re-format your mini-disk (back up your files, build
a new mini-disk, restore them.) In Unix parlance it would be like
having a separate disk partition for each user, no sharing, no dynamic
allocation.

They don't have any directory structure of disk files, the closest you
can do is have access to more than one mini-disk at once which again
is more like mounting flat partitions than a directory systems, more
like a bunch of MS/DOS floppies.

To share files across user mini-disks you need the READ and/or WRITE
password to the disk. There's no per-file permissions, either you give
away everything or nothing. Not an ideal situation from a security
point of view.

There are public mini-disks where software images are kept and these
are set up so everyone can mount them.

If any other user on the system does use write access to your home
mini-disk then you won't be able to mount it for yourself, basically
you won't be able to login.

XEDIT has its features, it's not terrible to use, nothing to crow
about, sort of like VI basically tho not really because IBM terminals
aren't full duplex so commands aren't interpreted until you hit ENTER.
It's interesting to make a bunch of changes and then hit ENTER only to
have them rejected because you typoed somewhere, even tho some of them
showed up on your screen up till then, then were undone, maybe
partially depending on where the error was detected.

REXX is like DCL or Unix shell programming, I guess whichever bizarre
one you figure out you tend to defend. I wouldn't particularly defend
any against the other, the mysteries of REXX are no less arcane than
the other two (stack stack where's my stack?) They're all very handy
for simple usage.

As far as the wonderful CMS documentation, yes, IBM does a very
professional job, I'm not sure if we're talking about style or content
really, some folks can be wow'ed by style. IBM's documentation makes
VMS's look small and compact, I'm talking perhaps 100 shelf feet for a
basic doc set, real handy. And again UNIX is the only O/S of the three
which has generated any popular press titles in bookstores. Students
generally just have to suffer w/o documentation or use centrally
located copies.

The on-line CMS stuff isn't bad at all, it can get you thru even w/o
documentation, at least basic stuff like using the editor. I'm not
sure it's enough to support a programmer however, that's where these
vendor help facilities always are weak although I can't say I've used
the on-line programming help much.

I don't think CMS is the wave of the future, so what's the point really?

	-Barry Shein, Boston University