Proposals/feedback for a distributed version control system for cutecash?

[John Ralls]

On Mar 30, 2010, at 10:14 AM, Derek Atkins wrote:

> John Ralls <jralls at ceridwen.us> writes:
> 
>> There's not much point to a read-only git repo, as it's pretty easy to just
>> use git-svn instead.
> 
> Actually, there WOULD be a point -- which is testing our migration
> strategy to make sure all the SVN history is correctly in the repo.  It
> would let us create the git repo, test it, and then swap over to it
> later.  A small point, but still a point...  And a point that git-svn
> doesn't really provide.  Actually, doesn't git-svn provide ReadWrite
> support if you have write access to SVN?
> 

Git-svn is indeed two-way if one has commit auth on the svn repo (with limitations noted in the git svn manpage). 

There are other differences that will need to be addressed, but it doesn't look like history is one of them. Commit hooks is probably the biggest concern, followed by svn treating tags as branches, while they're different beasts entirely in git.

The read-onlyness of a test repo would make it challenging to test those commit hooks, I think. 

Anyway, googling "svn git migration" turns up lots of blog posts with others experiences. If we get serious about this, we can severally experiment with different strategies and see what gets us closest. 

Let's get 2.4 released first.

Regards,
John Ralls