Stories
Slash Boxes
Comments
NOTE: use Perl; is on undef hiatus. You can read content, but you can't post it. More info will be forthcoming forthcomingly.

All the Perl that's Practical to Extract and Report

The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
 Full
 Abbreviated
 Hidden
More | Login | Reply
Loading... please wait.
  • You won't regret learning git. It's a fine solution to the problem of version control. But, we are blessed at the moment with many good version control tools. For all the complaints, subversion does work well. Mercurial is in heavy use by Sun for Solaris, OpenJDK, etc. Ubuntu are investing heavily in bzr. There's a lot of choice out there, and they're all pretty good (when compared to CVS, RCS and SCCS which came before).
    • Actually, for all its primitivity, SCCS wasn’t on the wrong track, at least according to the information I could find about how it worked.

      Even RCS is OK for single-user tracking of a handful of files in one directory. That may not sound like much – but it gets two things right: local repository instead of a server; and zero overhead for setting up a repository (not even a foo init). If someone wrote decent merging and push/pull support for it, it would actually not be bad at all.

      CVS is where things really went off the rails, by using RCS as its underlying model, accepting almost all its limitations with a shrug, and then breaking all the things that RCS did get right. Instead of addressing collaboration and experimentation by providing merging and push/pull, it tried to do so by going totally into the weeds. CVS is pretty much wrong in every possible way.

      Subversion is better than CVS in the sense that it fixes the limitations that CVS inherited from RCS. However, it does nothing to address the flawed premises that CVS was based on; on top of that, both of its physical repo formats are worse than CVS’s (quite a feat!).

      As for the DVCS alternatives: I don’t consider them all equally good choices. There are only two that have the right model of version control – git and Mercurial. Of the two, git has a much better technical basis. All others are based on partially flawed premises – although certainly they’re all far and away better than the CVS/Svn model.

      • Well, if you really want to investigate SCCS, you could poke at CSSC [sourceforge.net]. :-)

        Actually, OpenSolaris probably has a copy of the originals, but I can't find it after a quick look.

        • I saw that before. I’d be interested in a short conceptual overview of the SCCS model, but not interested enough to put in the effort to deduce that from the software. :-)

      • If I understand correctly, SCCS was one of the major inspirations for BitKeeper, as it was a major part of TeamWare.

        • Yeah; this is the sort of fragmentary hint about SCCS of which I’ve picked up a couple, which lead me to say that my understanding is it was decent, if comparatively primitive. But I can’t say any more than that since I haven’t seen any substantial exposition of how it worked.