"Daniel 'NebuchadnezzaR' Dehennin" <[EMAIL PROTECTED]> writes:
> Hello, > > So, I contact developpers of serveral Emacs modes for DVC tools, here > is the answer of the maintener of git.el. And here's my answer ;-) > From: Alexandre Julliard <[EMAIL PROTECTED]> > Subject: Re: Cogito mode for Emacs > To: Daniel 'NebuchadnezzaR' Dehennin <[EMAIL PROTECTED]> > Date: Sun, 12 Mar 2006 10:43:15 +0100 > User-Agent: Gnus/5.110004 (No Gnus v0.4) Emacs/22.0.50 (gnu/linux) > > Daniel 'NebuchadnezzaR' Dehennin <[EMAIL PROTECTED]> writes: > >> I follow the development of DVC (previously Xtla), I am a quite >> insignificant contributor so I try to open communication between >> projects, the goal of DVC project is to intergrate Distributed Version >> Control in Emacs. >> >> You can find principles at http://wiki.gnuarch.org/xtla#DVC >> >> You may be interessted by the fact that cogito has a backend in >> DVC, do your project and DVC have a chance to cooperate ? > > Well, I looked at DVC before starting mine, but it seemed to be very > tla-specific, It used to be *very* tla specific since we are starting from a 100% tla base. It's a bit less now. The bzr and hg back-ends are starting to be interesting. The goal is clearly to be tla-independant in the long-term. We are ourselves migrating away from tla and baz indeed (but we're still using it for the development of DVC). > for instance things like the inventory mode don't make much sense > IMO for other RCS systems. Sure! But the philosophy of DVC is to let a lot of freedom to the back-ends to implement back-end specific features. So, the tla-inventory mode will remain, but it doesn't force anyone to use it (Indeed, I almost never use it, I find the lint mode and the changes mode much more convenient). But many of the features of all those version control systems look like each other (XXX diff, XXX log, XXX status are very similar for any XXX among bzr, hg, git ...). > I wanted something closer to pcl-cvs since that's what I'm normally > working with, The diff-mode (which is actually more general than just diff, it can handle status as well - Indeed, almost any file-list fits in it) is not far from this. > and from looking at the code it was easier to start from scratch > than trying to adapt DVC... Well, indeed, if you consider a single back-end, it is actually more work to write it genericly than to write it specifically. But code reusability really pays when you think about several back-ends. For example, the bzr back-end currently has a revision-list mode, a status mode, a diff mode, a relatively good integration in ediff, a commit interface, and all this fits in 560 lines of code (344 if you remove blank lines and comments), while the DVC core is already more than 10 times bigger ! > Now I haven't looked at DVC recently, I gather it has become more > generic so maybe it would be easier now to make use of it. I don't > think I'll have time to do that in the near future, but of course > anybody is welcome to take my code and merge it into DVC; and if the > results work well I'd certainly consider adopting it. I'm also lacking time, but this would be great if this could happen ! -- Matthieu _______________________________________________ Xtla-el-dev mailing list [email protected] https://mail.gna.org/listinfo/xtla-el-dev
