Re: sccs2rcs to perl

[Riley Williams]

Hi Michael.

> Hi Riley -- Thanks for the reply.  At least now I know that my mail
> is being seen by someone other than myself.

NP.

> Let me address your points at they go...

>> I'm not on the CVS development team as such, just being a very happy
>> user of CVS...

> Oh, don't be fooled -- I'm a very happy user of CVS too. I just use
> SCCS in my "day job" because "that's what we've always used", but
> that's changing these days -- mostly with the help of the sccs2rcs
> perl script that I included before.

As stated in my original email, I never managed to get on with SCCS on
the only occasion I was expected to use it, so can't really comment on
it in any sort of meaningful way...

>> ...but I would like to offer one possible reason for the lack of
>> comments: Perhaps those on the list here are like me and don't use
>> SCCS in the first place, so have no use for the script.

> True, that could be a reason, but you may be missing the fact that
> I'm not advocating the *addition* of a script, but a *replacement*
> of a script.

If you think about it, you'll soon realise that the difference you refer
to is irrelevant to anybody who doesn't use SCCS - if they don't use
SCCS, then they will have no interest whatsoever in the existence of a
script to convert from SCCS to CVS, much less in how well (or even
whether) the script in question works.

> If you check your CVS distribution, you'll find in the contrib
> directory a sccs2rcs script *already* there. However, it is written
> in csh which is A) slow B) one more dependency for your .deb or .rpm
> to depend on. The perl script that I'm requesting consideration
> addresses both of these issues since the perl script runs about
> twice as fast as the csh version and, since there are already perl
> scripts in the CVS distribution, adds *no* additional dependencies.

To get any comments on either your script or the original one, you'd
need to find somebody else who uses both SCCS and CVS and needs to
convert from SCCS to CVS. Other than that, to be honest, nobody's going
to be interested for the simple reason that they've no reasonable means
to test either script.

>> Another possibility is the length of the "To:" and "CC:" list in
>> your email - I know several people who have their system set up to
>> auto-kill any emails with more than three names in those two
>> combined simply as a way to cut down the amount of mail they have
>> to handle, and your email would not have made it to any of them.

> I have sent this email already twice before -- once to bug-cvs and
> once to user-cvs with very reasonable To: and CC: lines so I don't
> think this is the reason.

I'm on the bug-cvs list but not on the user-cvs list, and I've never
seen your email before the one I replied to. I've been on the bug-cvs
list since last September, so wouldn't have seen it if posted before
then, but if posted since, it apparently didn't get through to that
list, which could explain the lack of response...

> I've also written to Larry Jones (who seems to be the primary person
> committing to the CVS source tree) directly and I got *no* reply -
> not even a "go away" or "not interested"!

That much I can't comment on as I don't know Larry's circumstances.
However, if they're anything like me, he probably doesn't have the time
to even read, much less reply to, emails about things he's not directly
interested in.

To put that in context, allow that (after my spam-filter has finished
with them) I receive between 500 and 700 emails a day, mostly from the
Linux-Kernel or the UK-Genealogy mailing lists, and if I even read every
email I received, I'd have little time to do anything else. As it
happens, I read every mail posted to the Linux-8086, Linux-Hams and
Bug-CVS mailing lists, but I only reply to ones where I can make a
positive contribution of some sort. With most of the mailing lists I
receive, I decide whether to read a particular email based on (a) the
sender is somebody important on that list (I read EVERY email from
either Linus Torvalds or Alan Cox on the Linux-Kernel list), or (b) the
subject line catches my attention. If neither of those apply, the email
doesn't get read - it's as simple as that.

Best wishes from Riley.