We did not switch to Subversion because the people who develop Subversion are not sympathetic to the ideas of the free software movement. That is a sufficient reason, given that CVS works fine.
See the full thread here, and don't miss Karl Fogel's response as well as Jim Blandy's response (Karl and Jim are two of the founders of Subversion).
As a Subversion developer and an advocate of both open source and free software, reading this kind of crap from Richard makes my blood boil. More importantly perhaps, it makes me want to remove the words "free software" from my vocabulary.
Richard, if you really want to help Free Software, you might want to leaf through a copy of "How to Win Friends and Influence People", because you sure aren't winning any friends right now—in fact, you're in danger of losing the ones you've had for years.
I'm really glad you posted this.
ReplyDeleteI too was amazed when I saw that mail from RMS. Never mind reading "How to Win Friends and Influence People", how about just reading "How Not to Lose the Friends and Allies You've Got"!
What really irks me (I've said as much in a private response to him) is the complete lack of documentation for the claim, and the double standard. In a later followup, he wrote:
Several years ago there was a suggestion to use Subversion in some major way on Savannah. People (I don't recall who) looked at the situation and reported that the Subversion developers did not support us as we wished, so we did not promote it.
It's possible I misremembered what they reported a few years ago. If so, I apologize. Before today I did not know who Subversion developers were. You and Jim are certainly free software supporters. I don't know whether Subversion as a project supports free software and GNU the way you personally have.
There is so much wrong there, it's hard to know where to start.
First of all, the credulousness. Someone "reported" that the Subversion project doesn't support free software in the way the FSF wants. It couldn't have been a very concrete report, both because it was wrong, and because it clearly wasn't detailed enough for RMS to remember anything about it now (not even who made the report). Yet based on it, he decided not to use Subversion on the Savannah server.
Second, the double standard. Sure, not every developer in the Subversion project supports every aspect of the free software movement as promulgated by the Free Software Foundation. But so what? The same thing is true of the CVS development team, and the GNU Emacs project has been using CVS for years. So what exactly is the standard here? RMS doesn't spell it out, and I can't figure it out. I don't see any significant political difference between the composition of the CVS development team and the Subversion development team. I've been on both; so has Jim Blandy. I doubt RMS is going to get more credible or better qualified informants than us.
Third, the conditional semi-apology. He shouldn't be apologizing for his memory, but for his credulousness in the first place. The issue isn't whether he remembers the report accurately now, it's why he ever believed the report at the time, let alone continued to believe it so firmly that he was willing to make such a sweeping statement without checking his facts.
And this isn't about whether the Emacs project uses Subversion as its version control system instead of CVS. As I stated in the thread, as an Emacs developer I'd be perfectly happy for Emacs to switch to Mercurial or GIT, or for that matter any other reasonable distributed VC system. A distributed system might even make more sense than Subversion, given the idiosyncracies of the way the Emacs project is run (that is: IMHO overly-strong centralized control, and release processes in which a majority of developers do not participate yet that sometimes interfere with regular development). CVS is holding us back; it's just not the best tool for this particular job.
I do look forward to seeing what decision Richard and the FSF make about running Subversion on their servers, now that the political misinformation has been corrected, though.