Hi Mark,

On 10.05.2016 04:21, Mark A. Flacy wrote:
Hmm.

I've used (in order, more or less), PLS (which I expect nobody to know), 
Clearcase, RCS, CVS, Arch, TLA, HG, BZR, and Git. I won't
claim to have used svn in any real sense.

The first 4 of that list were centralized version control systems and so not 
applicable to this discussion.

Of arch, tla, hg, bzr, and git, I'd say that I enjoyed bzr over hg for a rather 
long time. Now that I've used git, I think that git
has the correct distributed model over hg and bzr.

Once you realize that git doesn't require history to be immutable (but once 
you've shared history, you have a problem if you change
it), you'll find that you can do a lot of things that simply are not possible 
with hg and bzr. (That's not entirely true; the patch
queue for hg allowed you to do some things that you can do with git but with a 
tremendous amount of pain.)

Mr. Bächle, you should try to use git for a couple of your internal projects.


No need to get formal... ;)

I am using git for several projects (some own and some open-source), and I don't mind it. As I tried to explain before, I'm not opposed to making the switch to git for the SCons repo...but I'm trying to make sure that we're doing it for the right reasons.

And when single persons claim that there is a "hindrance" for them in contributing to SCons currently, because there's so much syntax that is hard to remember I start to wonder: with how many issues at the same time are these users juggling? Because if one works on at most one bug at a time, one should be able to get away with a "linear series of commits":

  hg clone ...
  # edit
  hg commit
  hg pull / push

which are all the same as in git. One can even throw in a "hg add" and it won't hurt, but only remind you that the file is already under version control. ;)


It's not about my personal preferences, it's all about the project.

Best regards,

Dirk


P.S.: One of my personal preferences is: I *don't* want the history in my repos to be mutable...maybe that's why git doesn't seem to be so more powerful than hg to me, and why I still consider them to be "on par" regarding functionality. At least for the work I have to do with them...

_______________________________________________
Scons-dev mailing list
Scons-dev@scons.org
https://pairlist2.pair.net/mailman/listinfo/scons-dev

Reply via email to