[gobolinux-users] Call for volunteers: would you like to be a recipe committer?

[Jonas Karlsson]

On Sat, 21 Jul 2007 09:23:32 +0200, Samuel A. Falvo II <sam.falvo at gmail.com> wrote:

> On 7/21/07, hisham.hm at gmail.com <hisham.hm at gmail.com> wrote:
>> have the SCM not let the user commit something if their local tree is
>> not up to date with the repository they're committing to. That's the
>
> We have this policy where I work (Google),

A sound policy imo.

> and it is a source of significant productivity drains.

Wouldn't breaking the build more regurlarly be a such source as well?

> I personally see patch management tactics as a non-issue.  Neither
> darcs, git, nor Perforce (by extension, SVN, since SVN and P4 are
> virtually identical systems, albeit by accident) can resolve the
> issues I've outlined above.  But darcs and git DO allow me to continue
> submitting distinct patches (instead of one behemoth CL) to my local
> repo, which I can then push (or the continuous build system can then
> pull, as the case may be) when the official repo goes green again.  In
> other words, with a distributed system, I'm not blocked often.  With a
> centralized system, I'm usually blocked.
>
I wonder why noone has mentioned svk (http://svk.bestpractical.com/)
yet. It lets you commit to a local tree, which then can be commited to
the central repository. It's a subversion addon/substitue.

-- 
/Jonas

Using Opera's revolutionary e-mail client: http://www.opera.com/mail/