The article at Perl.com describes the design and implmementation of svk, a distributed version control system built on top of Subversion.
Distibuted Version Control with svk
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
So Subversion is just CVS with the kinks worked out, right? This sounds like they’re trying to cobble BitKeeper-like distributed development functionality on top of what is essentially CVS. What a mistake. I can’t imagine it will work very well since Subversion was not built from the ground up for distributed development. What a shame.
I think the real solution to this is to use GNU Arch (http://www.gnu.org/software/gnu-arch) instead of Subversion. Unlike Subversion Arch was built from the ground up to support BitKeeper-like distributed development.
There’s clearly NOTHING cobbled onto CVS, Subversion is a new implementation from the ground up. It is _not_ essentially CVS, it just provides similar working principles because people know them, and they have proven their merits in the past.
Concerning Arch: I’ve never used it, so I’m not making any judgement here. But one thing that stopped me from exploring it is: where is the Windows support? If it wants to be widely used, then make it widely usable.
Arch definitely needs Windows support. For one, I’m using Subversion for my project, and Windows support is a big factor.
-Vesko
svk only uses the subversion fs layer for storing versioned resources locally. And the merge support is already as good as arch. Read the article about how the design decisions before over-advocating arch again.