home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!paladin.american.edu!gatech!usenet.ins.cwru.edu!agate!ucbvax!lrw.com!leichter
- From: leichter@lrw.com (Jerry Leichter)
- Newsgroups: comp.os.vms
- Subject: Running LOGINOUT
- Message-ID: <9301051432.AA20133@uu3.psi.com>
- Date: 5 Jan 93 13:31:23 GMT
- Sender: daemon@ucbvax.BERKELEY.EDU
- Distribution: world
- Organization: The Internet
- Lines: 74
-
-
- [On the issue of whether running LOGINOUT to map a CLI is supported,
- Steve Lionel writes:]
- There actually is documentation of the use of LOGINOUT with
- RUN/DETACH; in the VMS V5.0 manual set, it's on page DCL-392 of the
- DCL Dictionary in the description of the /AUTHORIZE qualifier to RUN.
- It says:
-
- When the image to be executed is the system login image
- (LOGINOUT.EXE), ...
-
- This implies to me that it is a supported use to specify LOGINOUT.EXE
- as the image name. However, I say this on my own behalf and not as an
- employee of Digital.
-
- I would call this barely on the edge of supported. This information is hidden
- within the documentation of a qualifier that need not be specified when
- running LOGINOUT - and it actually says just that!
-
- If you check the documentation for $CREPRC, which RUN/DETACH eventually uses
- to create the new process, you find very much the same kind of partial
- documentation. The table of stsflg values (page SYS-98 in the V5.0 book)
- mentions running LOGINOUT as part of the documentation of the NOUAF bit, which
- corresponds to the /AUTHORIZE (well, /NOAUTHORIZE!) qualifier, as well as in
- the documentation of the INTER and NOPASSWORD flags. However, at the end
- of the table we find:
-
- Note that the options PRC$M_BATCH, PRC$M_INTER, PRC$M_UAF, PRC$M_NETWRK,
- and PRC$M_NOPASSWORD are intended for use by DIGITAL software. Complete
- documentation of the possible ramifications of their use is not provided.
-
- Is that documented/supported or not?
-
- There's a further reference to LOGINOUT in the last paragraph of the Descrip-
- tion section:
-
- A created process is unable to run an image that calls the Run-Time
- Library procedure LIB$DO_COMMAND unless the process was created with
- the image argument specifying SYS$SYSTEM:LOGINOUT.EXE ... [which]
- causes a command language interpreter to be mapped into the created
- process....
-
- Again, hints at documentation/support, but just barely. (Actually, this is
- a very strange paragraph. Why does it focus on LIB$DO_COMMAND? There are
- several other CLI callbacks.)
-
- Still, I am not aware of any intention to
- declare this usage as "unsupported"
-
- It would be nice if it were decently documented, rather than being hinted at.
- In general, it's clear in VMS when something is documented/supported, and
- when it just happens to work. I can think of no other example in which the
- VMS documentation tip-toes around the edges but is unwilling to really
- commit itself. There's a big difference between "if you DON'T run LOGINOUT,
- you can't use LIB$DO_COMMAND" (hint, hint - but if you do it you MAY be on
- your own) and the direct "run LOGINOUT to cause a CLI to be mapped".
-
- On this matter, the VMS documentation just doesn't reach it's usual levels.
- (It reminds me of some of my favorite Unix documentation, where I can, for
- example, show you examples of flag values for system calls that are documented
- ONLY in the error messages: "This error is returned if such-and-so occured
- and you didn't specify the this-and-that flag." The "this-and-that" flag is
- mentioned nowhere else.)
-
- and am also unaware of any new
- feature coming in V6.0 that would replace this (and I think I would
- know if there were.)
-
- Well, that's good to hear! VMS development should understand that, whether
- THEY think this technique is officially supported or not, the documentation
- at the very least strongly hints that it is, and MANY people rely on it.
-
- -- Jerry
-
-