Re: Coniguring 301/302 redirects to track an fspath rename

It was civil until CMike got on his high horse- you need to police yourselves a bit better before you start accusing outsiders of misbehavior- that's how civility starts degrading in a project. The fact is that a robust solution involves coding thisup as part of the generic exception handlingmechanics for whatever svn ops can deal with301s.  What's currently in 1.7 that we're pretendingautomates the redirect problem is inadequate forany serious svn http server administrator trying to track moves.  This is not an idle question- Daniel actually tried putting a Redirect block in for the openejb -> tomee rename and all it wound up doing is destroying any attempts to crawl the project's pre-move history.  We had to back out the change just to let our git-svn mirror pick up the new tomee location and DTRT with the history. Please go ahead and test the current server behavior for svn.apache.org to see how existing svn clients respond pre-and-post podling graduation.  The server support and the client-side error reporting are already there. >________________________________ > From: Branko Čibej <brane@wandisco.com> >To: "dev@subversion.apache.org" <dev@subversion.apache.org> >Sent: Tuesday, February 5, 2013 1:52 PM >Subject: Re: Coniguring 301/302 redirects to track an fspath rename > >a) I know it's not a call for infra to make >b) if you can't lead a civil conversation, you may as well send it to >/dev/null for all the good it'll do. > >-- Brane > >On 05.02.2013 18:36, Joe Schaefer wrote: >> AFAICT if I really wanted this feature myself >> I could write a perl script that wraps svn client calls >> and DTRT with the error message in less than 20 LOC. >> You don't have to act like dicks towards people >> who are trying to take advantage of newer features >> you ostensibly support- the only reason we're discussing >> relocation features is because that's all you currently >> offer.  If you start supporting in-repo switch ops, >> hallelujah. >> >> >> >>    ------------------------------------------------------------------------ >>    *From:* Joe Schaefer <joe_schaefer@yahoo.com> >>    *To:* Branko Čibej <brane@wandisco.com>; >>    "dev@subversion.apache.org" <dev@subversion.apache.org> >>    *Sent:* Tuesday, February 5, 2013 12:08 PM >>    *Subject:* Re: Coniguring 301/302 redirects to track an fspath rename >> >>    1) that's not a decision for infra to make. >>    2) the svn client even with 1.6 now reports >>    the proper error message indicating the location >>    of the new url.  All infra needs is for the svn client >>    to provide a config option to actually interpret >>    the error status and take meaningful action itself. >>    This doesn't sound like rocket science to me, >>    no matter how screwed up the current level of >>    support for redirects may be. >> >> >>        ------------------------------------------------------------------------ >>        *From:* Branko Čibej <brane@wandisco.com> >>        *To:* dev@subversion.apache.org >>        *Sent:* Tuesday, February 5, 2013 11:50 AM >>        *Subject:* Re: Coniguring 301/302 redirects to track an fspath >>        rename >> >>        On 05.02.2013 16:07, C. Michael Pilato wrote: >>        > So, I hate to be the bearer of bad tidings, but this entire >>        exercise >>        > is a giant waste of time and energy. >> >>        I was going to say it'd be much better overall to create >>        podling trees >>        in SVN in the location they'd have when they graduate, as has >>        already >>        been half-way done with mailing lists. >> >>        -- Brane >> >>        -- >>        Branko Čibej >>        Director of Subversion | WANdisco | www.wandisco.com >>        <http://www.wandisco.com/> >> >> >> >> >> > > >-- >Branko Čibej >Director of Subversion | WANdisco | www.wandisco.com > > > >
Received on 2013-02-05 20:05:09 CET