inline
On Thu, Nov 4, 2010 at 1:34 PM, Frans Bouma <[email protected]> wrote:
> > Is what you say obtained from experience, either running a project in a
> DVCS
> > or contributing to one?
>
> running a couple in svn and earlier in CVS (*shiver*).
I am sorry, the experience really does not translate.
> if you think I'm talking about a perspective of a VSS user, you're
mistaken. We use svn for years, with multiple repositories, multiple trunks
(15 or more per version), merging fixes from live branches to new feature
branches every week (if applicable). This used to be problematic as svn
didn't track changes across merges, but since 1.5 it does that and merging
is really simple. The only conflicts we get now are the ones you also get
with git (same lines of code is edited in both branches).
Again, the experience is not applicable.
But again, it's about setting a standard, checking if code obeys
that standard and moving on. As there's no standard, everyone can commit
whatever code that works, and I think that also happens (as when it doesn't
fail a test, it must be working correct code). I just wondered, whether this
would become a bigger problem if people start pulling changes from off-main
trunk branches more often than not. I think it will. Others, think it
doesn't. Well, one way to find out, right? :) Sad thing is: if I am right,
it's hard to change it back later.
Still not following you. Can you explain how the use of DVCS means that
there is no standard?
FB
