Well, despite the initial "haha" reaction, I feel bad for the Subversion project team for all of the attention/politics they're going to get from this.
As you can see on the ticket, they'll already getting the usual Apache internal politics of "that's not how we did it in the 90s!" reaction to any infrastructure/tool changes.
And, with showing up on HN, they'll likely have everyone snickering at how Subversion is so obviously inferior to git that even their own dev team doesn't use it anymore.
When really they're just trying to do their job, with the tools that work best for their situation. (I assume, I have no internal knowledge of their decision, I'm just assuming good intentions.)
If it is, it's incredibly well played by Greg Stein. He's riffing off the oft-lambasted git/svn politics, private lists, etc. bureaucracy of Apache, which, to me, made the ticket seem really believable.
...will have to wait until 4/2 to see whether I'm a sucker or not. Although perhaps that means that I am either way. :-)
Obviously Subversion is important from a historical perspective, there are so many repositories out there will need to be supported for some time, but it does seem like a flawed enterprise to keep actively developing Subversion. Wouldn't that time be better spent on a platform that's viable, like git, Mercurial, or some other experimental platform that might make a big breakthrough?
Everyone's bailing on that project, including Subversion itself. Isn't that a sign you should, you know, move on?
Even CVS is, sadly, still in use. Subversion, being far superior, is likely to itself remain in use in even more places well into the next decade. It'd be nice to avoid bitrot and see pain points further smoothed over.
We still use subversion for binary file tracking - specifically Altium files and the like. In thus case, we heavily use the locking feature. Git does not fit well when merging changes is nearly impossible. In addition, the local copy size for Git would be way too large.
I'd somehow missed this! Looks awesome. We already have started using git for some side projects that are public. Again, no central locking database, though I suppose we could use something out-of-band.
I know of a huge, highly profitable financial services company that was very pleased to announce (internally) their successful migration from CVS to SVN, in 2011. There's no way they're ready to move on.
I have doubts anyone could be so incredibly stupid as to raise an issue in the INFRA JIRA as a joke. Note that requests are made to the ASF infrastructure team by creating issues in this project, such as provisioning and maintenance. Obviously @jimjag is taking a dim view of the request, but that's for different reasons; the (perceived) lack of transparency by the PMC when making this decision. However I think getting the infrastructure team to migrate a repository just for the LOLz would be seen even less favourably.
Seriously? Downvoting? Did you actually read the discussion?
"I would like to point out that the new community dynamic for git-based projects is different in many salient ways from a traditional subversion project. Voting in private, or even decisions by fiat of the chair, are most welcome in the new regime. It is good to see the subversion project embrace this new mode of project development in their smart exodus from the chains of the past."
Joe Schaefer
As you can see on the ticket, they'll already getting the usual Apache internal politics of "that's not how we did it in the 90s!" reaction to any infrastructure/tool changes.
And, with showing up on HN, they'll likely have everyone snickering at how Subversion is so obviously inferior to git that even their own dev team doesn't use it anymore.
When really they're just trying to do their job, with the tools that work best for their situation. (I assume, I have no internal knowledge of their decision, I'm just assuming good intentions.)