[patch] Use /etc/mailname if available

[Aaron Bentley]

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

John A Meinel wrote:

> user at hostname
> 
> Is technically correct. Though I guess it should really be "user at fqdn".

Okay, it may be technically correct in some sense, but if you look at
it, the purpose of the committer ID seems to be

1. identify the committer
2. provide a means of contacting the committer

abentley at lappy (yeah, there are some commits with that) does not serve
either purpose well.

For identification, the 'lappy' part isn't specific enough (I can't be
the only person on the web who likes Homestarrunner) and neither is
'abentley' (though I do manage to get that username on most web sites).
 It may make sense to grab the full name out of /etc/passwd, but you may
get 'Debian User', etc.

For 2, it's not a routable email address, and I don't think bzr has any
way of guessing which address to use.  It comes down to preference,
anyway. (abentley at panoramicfeedback.com? aaron.bentley at utoronto.ca?).

I don't have a perfect solution.  It might make sense for bzr to write
and warn when first used:

$ bzr init
note: setting committer id to "Debian User <user at localhost>".  If you
want to change this, edit ~/.bazaar/email.
$

> Also, if you include the admin overhead, bzr is *way* faster to setup
> than svn. The problem is most people will come to it with everything
> already setup by the admin. And I don't know that we have a good way to
> do that for bzr.
> Perhaps a "bzr initialize" plugin, which interactively (possibly with a
> simple gui) helps you setup the stuff you care about and get going.

Yeah, I think that would be nice.  Maybe along the lines of the Debian
exim setup, so that it lets you declare what you want to do, and then
prompts you to fill in the data.

If we assume checkouts, repositories, bound branches and lock-step
development are implemented...

Which situation best describes you?  (You can always adjust these
choices later)

1) I just want to create a branch in this directory
2) I want to store my branch data in another location
3) I want to do lock-step development (cvs/svn-style) on this branch

And after 2 or 3:
1) The branch location is fast and local.
2) The branch location is fast and remote, but I don't need to work offline.
3) The branch location is slow or I can't reach it when I work offline.

After 3
1) I prefer to store my local branch data in the working tree
2) I prefer to consolidate all my branch data in one place

etc...

Aaron
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFDpttX0F+nu1YWqI0RAs2UAKCAi0evBoWAJ7h8HKoye+en+21dMQCfUXqC
GVzeMlNW8/OdKCIHtPDKbVw=
=eHWr
-----END PGP SIGNATURE-----