Re: [ANNOUNCE] metagit 0.1.2

[Christian Dietrich <stettberger@dokucode.de>]

[-- Attachment #1: Type: text/plain, Size: 1667 bytes --]

Thomas Koch <thomas@koch.ro> writes:

> Would you mind comparing your project with
> http://kitenet.net/~joey/code/mr/ ?

I didn't know mr, when i started hacking metagit (which was based on an
Idea mit mschilli in the German Linux Magazin). My focus was always,
that i can run the same config on all of my machines, because I tend to
have all my documents in single git repos (e.g. every course at
university is a git repo).

To make this happen, i needed to define some modules that list me all my
repositories on the server where my university stuff is. But i didn't
want to see my uni stuff at work, so i needed policies where the repos
should come up. As i see these features aren't implemented directly in
mr.

For my work, where svn with svn:externals is used, I implemented the
svn:externals support, where all the externals are cloned
recursivly. (You can even define if a given command is executed on the
externals or not). This is also a feature mr doesn't have.

And there is the simple, but useful metagit cd, which echos a cd command
to a repository path. So with the bash function

mm () {
        $(metagit cd $@)
}

I can simply jump to the metagit repo with `mm meta', where ever it was put.

On the other hand metagit doesn't implement recording of actions if
you're offline. And of course the most important point that metagit is
worse than mr is, as already mentioned, that it is written in python and
not in perl. 

greetz didi
-- 
(λ x . x x) (λ x . x x) -- See how beatiful the lambda is
No documentation is better than bad documentation
-- Das Ausdrucken dieser Mail wird urheberrechtlich verfolgt.

[-- Attachment #2: Type: application/pgp-signature, Size: 835 bytes --]