[kepler-dev] SVN or GIT
cxh at eecs.berkeley.edu
Wed May 21 15:53:08 PDT 2008
The biggest issue I see is:
How hard is it to use git out of the box with Eclipse/Netbeans etc.?
One advantage of CVS is that it works with Eclipse without downloading
extra software. Last I looked, Eclipse requires a plugin for SVN.
This makes it more difficult for new users to get started. Eclipse is
hard enough to use as it is.
How hard is it to set up Eclipse with git vs SVN?
When it comes to tools, I see a tension between cutting edge tools and
ease-of-use. We do students a disservice if we don't train them to
use tools they are likely to see in the future. On the other hand,
who really benefits from using git (or SVN)? It is really probably
the professional software engineers who are building releases. I'm
not so sure that the scientist in the field is going to see benefits
unless we make it _very_ simple to download, configure and build.
I see a very high risk of building a very flexible build system that
software engineers like me like, but that a scientist cannot or
will not use.
Just my $0.02.
I'm basically fine with going with git over SVN, but do we lose Maven
support? Do we end up on the cutting-edge with git and end up bloody?
However, we would need to decide soon (this week) since I'm about to
work on migrating Ptolemy to a new version control system.
Debating git vs SVN seems reasonable to me.
We could have a telcon either today or tomorrow about this.
I'm committed to moving Ptolemy to a new version control system.
I'm also fine with following Kepler's lead moving Ptolemy to either
svn or git.
There was talk at the kepler core meeting of moving the kepler CVS
repository to SVN. We may want to move to GIT instead. It seems to be
the up-and-coming successor to SVN. It supports distributed
repositories and claims to be much faster for distributed development
teams. Here's a howto for SVN users: http://git.or.cz/course/svn.html
Kepler-dev mailing list
Kepler-dev at ecoinformatics.org
More information about the Kepler-dev