Interesting point although I am not sure I understand what the limitations are. If I were using a Subversion repository, would I be bound be the limitations of Subversion (no offline versioning) and just looking at git's face? This seems to be a bit against common perception which is more along the lines ignore git's face but admire the power of the tool.
Could you clarify a bit?
On the other hand, it should be rather easy to migrate to a new system, totally agreed. In the worst case, I'd keep running one system until the article is published and start the next project using some other system. Compared to the size and complexity of the code base you guys are dealing with, an academic publication is laughably simple in structure.
Thanks
Christoph
On Sat, Mar 1, 2008 at 3:44 AM, Jay Soffian <
jaysoffian@gmail.com> wrote:
One thing to keep in mind is that whatever decision you make isn't totally cast
in stone. It's pretty straight-forward to convert from Subversion to a DVCS
(well, at least to git or hg, and I'd imagine to bzr as well), and you can also
convert between the DVCSs with varying degrees of ease.
Another option is using git as an svn client. You can setup an svn repo, let
most folks use the normal svn clients, and for anyone who wants the ability of
offline commits, just let those folks use git's svn integration.
j.