Sorry for keeping this thread alive, but... On Wed, Aug 18, 2004 at 12:16:50PM -0400, Richard Coleman wrote: > Robert Watson wrote: > >Last I looked, my primary concerns with Subversion were: > > > >- Cost to import full FreeBSD history. > > > >- That it promised the multi-way branching and merging in a future > >release, but did not yet have it. > > > >Do you know how things look with respect to the second issue? > > > >Robert N M Watson FreeBSD Core Team, TrustedBSD Projects > >robert_at_fledge.watson.org Principal Research Scientist, McAfee > >Research > > Reading the ChangeLog for the upcoming subversion-1.1 release, it > doesn't look like it. > > http://svn.collab.net/repos/svn/trunk/CHANGES > > My understanding is that the merge functionality in subversion is still > about the same that is currently in CVS (although faster). Saying that the merge functionality is the same is a little misleading. On paper, it appears the same in that it too lacks a mechanism to help prevent repeated merges. In practice, it's much easier because of the way revision numbers work---across the tree. This allows you to use one simple merge command on one revision number to get an whole set of changes across an entire tree. We made the switch from CVS to Subversion at work almost a year ago and never looked back. Or Robert, do you mean something else by multi-way branching and merging? > But there's no denying that subversion is getting better by the day. > For a new project, I wouldn't hesitate to suggest subversion. But for a > large, existing project like *BSD there doesn't seem to be a compelling > reason to switch (yet). But that will change eventually. For a large existing project like FreeBSD where time is a precious resource, the reasons seem even more compelling. Subversion provides an automated solution to much of the manual dog work of CVS... can anyone say "no more repo-copies"? GeoffReceived on Wed Aug 18 2004 - 16:09:47 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:07 UTC