On Thursday, 24 April 2003, Jeff Squyres wrote:

The problem is NOT stderr, but more general...
    

Ah -- I see.  So you're more interested in being able to "short circuit"
parts of the target dot files to avoid running unnecessary things when LAM
is being booted.

Exactly!

Can you give a few examples?

As I mentioned in my original posting, we have an institute-wide dot profile that takes care of users' needs on a variety of platforms.  For instance, we initialize the terminal type with "tset"; if $TERM is not definied or not a known value according to terminfo the user is asked for his terminal type.  This would block LAM from booting.  Another example: Some users use UNIX primarily to run simulation models.  They are asked which model to run and what X window System display to use for the GUI if the $DISPLAY variable is not set; they need not necessarily every get a shell prompt.

We short-circuit these things for batch environments on our multiprocessors. NQE and OpenPBS provide the "hooks" we need to detect the fact that the login is occurring in a non-interacive context.  In such a case we just do the core initialzation (e.g., paths, license files, etc.) and skip everything.  IMHO LAM should provide such a hook as well.

As for the 7.0 release, I'm sorry I'm too late...

It's not a question of being innocent -- the issue is twofold:

a) can we support such an environment variable in all boot environments
(LAM 7.0 will support rsh/ssh, Globus, bproc, tm).

As a newbie I cannot oversee the consequences for Globus, bproc and tm.

b) we're well into the release process of 7.0, and new features ... are not allowed so that we can have controlled testing ...

I understand fully, but you can't blame me for trying. :-)

I saw you initiated a change request.  As I mentioned in my initial posting we have a quick-and-dirty workaround, so I can patiently wait for the feature to appear in a future release.  Thanks for your help!

Cheers,
    Irv.