1 Feb
2009
1 Feb
'09
1:53 p.m.
On Fri, 30 Jan 2009 at 09:49:34 Chuck Guzis said
It surprised me that the micro types took so long to
figure basic
things like SCCS and regression testing out.
I think I can partly explain that. For one thing, though we may all be
brilliant ;) we are not necessarily all that wise! Wisdom hopefully
accumulates with experience, and perhaps with the agony of recreating
something that was lost that should have been carefully preserved -
especially things like batch files, make files and little utilities written
for example because the available software tools could not produce
deliverable binaries in the desired form.
More seriously, most of the places I have worked for the past 45 years have
been what I call "Mom and Pop" operations with all the problems that
produces, including tight budgets and razor-thin resources. BTW, just
because it is a fair size company does not prevent it from being run in "Mom
and Pop" mode. The thin resources in my experience usually result in a very
fast transition to the next project once the current one is done, which does
not leave a lot of time for reflection about what all should be preserved.
Even the quite large company for whom I currently work, though sadly only
until next Friday, :( has run our soon-to-be-dissolved group on razor-thin
resources IMHO.
The other common problem I have seen is that usually no one in the chain
above my immediate manager has the vaguest idea as to the importance of the
work I have just completed. During some unfortunate periods, even he had no
clue!
I've been quite pleased for about the past 14 years to have had the use of
MKS and more recently VSS to provide an archive and history information on
the file servers. Allegedly these are backed up frequently off-site as
well, on tape cartridges I believe, though I have never had an occasion to
test that theory. At one time the IT guy would whine about how much space I
was taking up on the server and I would have to give him a list of folders
that could be written to CD's to make space on the file server. In at least
one case, I recall one of those was partly unreadable when I tried to access
it only maybe a year after it was burned. It has now been probably 10
years - anyone want to bet any of them are still readable? Since they are
sitting in my bookcase at work, anyone care to guess what will become of
them once I leave next Friday? The good thing is that they are all very old
versions and it probably does not matter all that much. No longer needing
to keep card/mag-tape/floppy backups has indeed been pleasant. (I sort of
hate to admit I once even had some punched paper tape backups.) The history
information the source control software provides is invaluable as well, even
though I do keep written notes regarding software changes as they are made.
I've been a bit amazed at how often I've found myself looking in the MKS and
VSS archives to search for the answer to a question like "Exactly when was
_this_ particular tiny change made". These are usually changes that seemed
almost inconsequential at the time, but turned out much later to have been
rather important.
Later,
Charlie Carothers