rjk@mgweed.UUCP (09/12/83)
This is more of a flame than a bug description, but I put it here in the
hopes that future UUCP system designers take heed. It's nice to be able
to see the status of C, D, and X jobs waiting for processing. A few of
us have written monitoring programs to do that - Martin Levy wrote uumon
a year or so ago, and I wrote uusnap to go along with Tom Truscott's
uucp package. Both programs rely on the fact that the work files all
have the same length file name, less the length of the system nodename.
Now comes all these fancy new uucp programs and, yes folks - let's change
the length of the file names. It's practically impossible to scan the
work directories for a count, by machine, of work outstanding. Let's
try to give this some consideration in the future. Better yet, put some
kind of delimiter in the file name such as a tilde or hash or something
so that the length does not matter:
we13~Ab090e
mach~1c1d0f4
mach~f2e3
If this were in the "D" directory, I would know that I have one data file
for or from we13, and 2 data files for or from mach. As it stands now, I
really don't know where the machine name stops and the sequence number
begins.
Randy King
{ihnp4,we13}!mgweed!rjk