Hi! On Mon, Jun 22, 2009 at 09:58:04PM -0500, Eugene Prodeguene wrote: >[...]
>http://www.openbsd.org/why-cvs.html >Because none of the above mentioned will allow for 70+ developers to >update ~1.2GB/~140,000 files of source code, allow anonymous checkouts, >has an available web based interface and interfaces with ssh. Now, that page uses outdated comparisons. Ok, cvs works better than sup... Wow. sup is such a current comparison ;-) Now, git allows even more efficient updates and checkins, both authenticated via ssh, and anonymously. For anonymous readonly access you don't need ssh, because git uses sha1 hashes for everything. (Ok, the use of sha1 becomes more and more debatable, of course.) You can even gpg sign tags if you like (but unsigned tags are also available, and tags and branches are *cheap*, no churning through thousands of files for creating a tag or branch!). Of course it has a standard web interface, too. And it's more convenient, because you can look at changes in changesets instead of spreading *one* commit over all the files touched as if it were independent changes. git is very efficient both for deep history and quite large trees. Just as an example (other dvcs seem to perform well too on those lines of comparison). Of course svn offers similar features too (web interface, of course, svn via ssh or own service or https/dav). It has some drawbacks (I'm not a great fan of it anymore): non-first-class tags/branches, bloated working-directory format, bigger repository (IIRC git has one of the most compact repository packing algorithms). >[...] Kind regards, Hannah.
