Graph of VCSes before Martin's git package maintenance demo.

martin f krafft madduck at madduck.net
Tue Jun 26 16:53:08 UTC 2007


also sprach Will Thompson <will at willthompson.co.uk> [2007.06.24.2206 +0200]:
> Why did you place monotone on a different branch (ahem) of the
> graph?  I assume it's connected to its weird hybrid
> two-types-of-branch stuff, where any branch can have multiple
> heads which you can then merge (which I have not seen elsewhere),
> but I'd be interested in an elaboration.

Yes, that was the reason. The graph itself was inspired from
a discussion on #revctrl/irc.debian.org on 18 Sep 2006:

  http://colabti.de/irclogger/irclogger_log/revctrl?date=2006-09-18

you may find more information there. Nathaniel put it like this:

  in monotone, branches are named and stored within a single repo,
  and network operations simply exchange information about new items
  (so my box might tell the server "btw, there are some new
  revisions in this branch and that one") a single branch can have
  multiple heads, if you want to merge you do that as a separate
  step.

> Also, darcs' position suggests that you consider it to have been
> superseded by the other DVCSes mentioned on the graph.  Is this
> the case?  If so, why?

Again, Nathaniel explained this:

  darcs is like identical to bk/hg/bzr wrt its branch workflow, but
  if you look at their history models, you have darcs vs. everything
  else as by far the biggest split

It took me a while to understand my own graph again:

  http://people.debian.org/~madduck/talks/vcs-pkg_debconf7_2007.06.21/vcs.png

it's a bad graph anyway. It started out as a chronologic
left-to-right graph, but then that did not work out, so I switched
to something else. The three lines on the bottom are not scales in
that VCS to the right are "more atomic" than those on the left, it's
more a split into atomic vs. non-atomic. The lowest x-axis just
simply separates branching models, and here, SVK, darks,
bk/git/hg/bzr and monotone differ a lot.

I guess darcs is also split from the others because its author
basically did it himself, while bzr and hg and git and so on all
stem from a common theory. darcs is very innovative in that sense
and has been very influential for the others. However, I also find
it not to scale at all. YMMV.

-- 
martin;              (greetings from the heart of the sun.)
  \____ echo mailto: !#^."<*>"|tr "<*> mailto:" net at madduck
 
spamtraps: madduck.bogus at madduck.net
 
la lune, c'est comme les canards
il faut aimer caresser les chats
pour avoir envie d'y aller.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature (GPG/PGP)
Url : http://lists.alioth.debian.org/pipermail/vcs-pkg-discuss/attachments/20070626/af117e9e/attachment.pgp 


More information about the vcs-pkg-discuss mailing list