Comment 3 for bug 592148

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Parth Malwankar wrote:
> So is the message "Will continue to try until XXXX, unless you press
> Ctrl-C." (and the related timeout code) actually used for any specific
> case or should this be removed?
>

"Will continue to try" is bogus, which I mentioned (I believe) on the
other bug that you posted. That should be removed.

Though if I think about it, I think a *short* timeout would be
reasonable (10s). I'm not entirely sure why we picked 5minutes as the
global default.

I'm pretty sure it is rare that people encounter a branch on Launchpad
that is locked by real contention (such that waiting for the lock to
release is what you want to do). If you are pushing, then only one of
the two people pushing can succeed (eg. even if you wait on the lock,
then your basis *should* be out of date and you'll need to pull &&
merge/update before you can push again anyway.)

As such, if there is a lock it is:

a) Stale, and will never be released anyway
b) Something is being updated, and you'll probably need to serialize on
that being completed before you can do what you wanted anyway.

The only caveat is that updating branch.conf currently requires a write
lock, so there are a couple things that could be done 'in parallel',
where waiting for a lock could be reasonable. But rare enough that I
don't think we need to focus on supporting it.

John
=:->

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkwQ/OAACgkQJdeBCYSNAAPt/ACeOt/XZrlJ7iaFV7e0oS7ajUJa
6yMAn3mK7s89QnxhhPBWh/YXLSY0fDMl
=NbmV
-----END PGP SIGNATURE-----