I would think this should be possible, but I don't know that much about Mercurial myself. From a quick web search, I see Mercurial has a conversion extension, has this been investigated?
On Tue, Sep 6, 2011 at 3:53 PM, Upayavira <[email protected]> wrote: > This is me, knowing nothing about Mercurial nor the Wave repo. But I > wonder if you could write a script that checks out the first revision > from Mercurial, and commits it to a local SVN repo. Then updates to the > next revision, and commits that. And so on, all the way through the > repo. In the end, you have an SVN repo ready to be imported. With a > local svn repo, you can 'fake' users, it doesn't worry about > authentication. > > Am I being naive here? > > Upayavira > > On Sunday, September 04, 2011 7:28 PM, "Yuri Z" <[email protected]> > wrote: > > I think that it is technically possible to make clean Hg->SVN migration > > and > > it would be pity to give up the history. I ll try to investigate it some > > more for some more time. Let's postpone the decision for a while (2-3 > > weeks). > > > > On Sun, Sep 4, 2011 at 5:16 PM, Michael MacFadden < > > [email protected]> wrote: > > > > > Honestly at this point, I would vote for just a clean check in. While > we > > > can talk about he use of having the history in the SVN, the fact of the > > > matter is that I don't think it is really important enough to hold us > up. > > > That fact alone has been the only reason why we haven't migrated the > source > > > in nearly 10 months. The fact that no one on the project has spent the > time > > > over the last 10 months to figure out a solution says to me that it > really > > > is not that important. The revision history will stay on Google code > for > > > historical reference if we need it. > > > > > > Imagine if we had just switched over at the beginning of the project. > We > > > would then have 10 months of check in history in the SVN. Most times > when > > > we need to look back at the revisions it's because something had > changed > > > recently. If we had 10 months of history, I doubt we would be going > back to > > > the Google Code Hg very much at all. I think the need for the Hg > history > > > will decrease rapidly over time once we actually make the move. > > > > > > I know it would be nice to have the history, but it seems to be the > road > > > block. If we just bite the bullet and make the switch a few months > from now > > > I don't think it will be impacting us at all. > > > > > > ~Michael > > > > > > > > > On Sep 4, 2011, at 6:13 AM, Yuri Z wrote: > > > > > > > Hello > > > > Yes, you are stressing an important point. I don't think anyone on > the > > > Wave > > > > project would like cancellation of the podling. So, the only solution > > > would > > > > be just complete the migration and move the source code to the Apache > > > Infra, > > > > hopefully along with the Wiki. > > > > However, there are technical issues as well. I already contacted the > > > infra > > > > and the Apache SVN mail lists for assistance on the move from Hg to > SVN, > > > but > > > > it seems like there's no single easy to use tool to do it. There are > > > bunch > > > > of tools that can help, though, but that requires investigation. If > the > > > > infra would provide some tool that would enable automatical migration > > > from > > > > Hg to SVN - that would be really helpful. > > > > Yuri > > > > > > > > On Sun, Sep 4, 2011 at 3:59 PM, Christian Grobmeier < > [email protected] > > > >wrote: > > > > > > > >> Hello, > > > >> > > > >> Wave has entered incubation on 2010-12-01. > > > >> I think it is time to re-open the discussion on the source code move > > > >> from hg to svn again. The project is now 10 months in incubation and > > > >> the sources are still not the ASF. Without sources incubation makes > no > > > >> sense imho. > > > >> > > > >> Can we sum up what exactly is going on and what are the blockers? > > > >> > > > >> I know people are not keen working with SVN, but as long as there is > > > >> no GIT at the ASF, this is the only way to go. If this is a blocker, > > > >> we should discuss the cancelation of this podling. I think it is not > > > >> (or should not) > > > >> > > > >> Are there technical problems - then we should outline whats > expected. > > > >> Maybe infra can help > > > >> > > > >> CHeers > > > >> > > > > > > > > > -- --Matt
