Re: [ slightly OT ] anoncvs or cvsup?

[Woodchuck <djv_(_at_)_bedford_(_dot_)_net>]

On Sun, 30 Jun 2002, Brian Szymanski wrote:

> Hi y'all...
>
> My ISP has recently decided it will charge per-Gig rates over some certain
> limit.

Morons.  Better they counsel the subset of users with pr0n habits,
or smoke out and hang any spammers.  Lazy devils won't do traffic
analysis...

> limit. Since I like to keep my sources up to date, I (currently) have
> three machines pulling anoncvs updates each night (one tracks current, one
> 3.0, and one 3.1, so I can't just pull the updates to one machine and then

I'd wonder if it is necessary to update 3.0-stable and 3.1-stable on
such a frequent basis?  One sees so few changes to either.  Maybe monitoring
the daily changelogs would suffice, with a weekly, or monthly, or even
"Oh, that's interesting!" run?  No sense paying to only update
spelling fixes in man pages.

> exchange the updates locally via NFS/cvs/whatever). The CVSup webpage
> makes a lot of fuss about how fast it is, but what I'm wondering is how
> bandwidth efficient it is as compared to anoncvs. Anyone run in to a
> similar situation? I'd rather not do CTM, because that would be a big PA

CTM would probably eat the bandwidth for breakfast. (Just guessing: it's
not hard to improve on email.)

I believe cvsup would be more efficient than anoncvs, though; my
impression is that it does a lot more "work" on the clientside,
anoncvs seems to send a (large) fixed amount of data to the server
regardless of the size of updates sent back.  I base this on informally
watching the traffic going out over my ppp link.

> to forward the emails around to the right machines and such... Any ideas?
> Thanks in advance, sorry if this is a little off-topic...
> Cheers,
> Brian

Well, compared to some other topics aired lately...

Dave
--
        In each of us, there burns a soul of a woodchuck.
        In every generation a few are chosen to prove it.