I don't think there's much percentage in moving to the CMS with a structure like that of Commons. That said, checking in the whole mess, as Phil suggests, seems perfectly doable and should not preclude updating parts of the tree in quite a similar fashion as how updating a given component's site is done today, except no ssh to mino. Am I missing something fundamental?
Matt On Mon, Dec 10, 2012 at 1:47 PM, Phil Steitz <phil.ste...@gmail.com> wrote: > On 12/10/12 11:40 AM, Phil Steitz wrote: > > On 12/10/12 10:50 AM, Ralph Goers wrote: > >> All the sub-sites are hooked off the main site. I would have no idea > how to migrate anything without migrating the main site first. > > Having now looked at [1], it looks to me like we can solve the > > immediate problem using svn pub-sub. The docs are not clear and > > they may not allow it, but in theory, we could in fact do this > > incrementally - start an svn repo and move the "mutable" portions > > there incrementally, understanding that only changes to the > > svn-migrated stuff will be propagated. If infra barfs on that, then > > we just commit the whole mess starting at the top level commons > > site, following the Ant example linked on [1]. Then to make > > changes, we follow the process that has been in place for the main > > ASF site for ages - maintain a local checkout of the generated site, > > re-gen when you want to update and check in. > > > > I know playing with the CMS might be fun and interesting; but a) we > > have no volunteers to do this and b) we do not have time to migrate > > all of the content. > > > > Therefore, I suggest we just follow the Ant route; possibly moving > > incrementally if infra allows that. > > Let me modify the proposal to make it simpler and more palatable to > infra: > > Just do like Ant - check in the whole mess. > > Phil > > > > Phil > > > > [1] http://www.apache.org/dev/project-site.html > >> I suppose it is possible to point to the sub-sites in their current > location but in logging we found it more beneficial to simply copy the > content under the main site in the CMS. > >> > >> Are all the sub-sites built with Maven? Any that are not could move to > the CMS as part of the main site. > >> > >> Ralph > >> > >> > >> On Dec 10, 2012, at 8:23 AM, Phil Steitz wrote: > >> > >>> On 12/10/12 7:55 AM, Ralph Goers wrote: > >>>> That is what we did in logging. Changing it at the end is fairly > easy. However, we don't have much time for testing. > >>> Do we really have to do it all at once? > >>> > >>> IIUC (which is quite possibly false), the intent here is to get > >>> everyone onto svn pub-sub and kill off the rsync processes from > >>> p.a.o to the live site. The stake in the ground is that the rsyncs > >>> are going to stop Jan 1. Do I have that right? > >>> > >>> Why can't we move to svn pub-sub incrementally somehow, > >>> understanding that until individual components move, their sites > >>> will be read-only? So basically, when you decide to make changes to > >>> a site, you get it set up for svn pub-sub? Note I am including the > >>> main site in this - i.e., it does not have to "move" until it needs > >>> to be changed. It would seem to me (again, may be missing something > >>> critical) that we could build the newly definitive source svn repo > >>> incrementally, with publishing always sourced from there, but "old" > >>> directories on the live host remaining until they get clobbered. In > >>> the worst case, if the live host *must* include only svn-published > >>> files, we could seed the new repo with the "old" content. But I > >>> don't get why that has to be the case; because if it is, we are in > >>> worse shape than Christian suggests - come Jan 1 we will be dark if > >>> we don't get everything moved to svn pub-sub. > >>> > >>> Sorry if above is naive. The intent is to avoid a huge amount of > >>> fiddling in a short period of time when quite a few component sites > >>> have not changed and will not change for some time to come. > >>> > >>> Phil > >>> > >>> > >>>> Ralph > >>>> > >>>> On Dec 10, 2012, at 4:34 AM, Christian Grobmeier wrote: > >>>> > >>>>> On Mon, Dec 10, 2012 at 1:32 PM, Gary Gregory < > garydgreg...@gmail.com> wrote: > >>>>>> Well, we have to start somewhere. This is going to be a lot of work, > >>>>>> we have many components, unless you see a short cut. > >>>>> I thought we would create: commons-test.apache.org > >>>>> move all components to there and then make a switch from > >>>>> commons.apache.org to the new site > >>>>> > >>>>> > >>>>> > >>>>> > >>>>>> Gary > >>>>>> > >>>>>> On Dec 10, 2012, at 7:13, Christian Grobmeier <grobme...@gmail.com> > wrote: > >>>>>> > >>>>>>> On Mon, Dec 10, 2012 at 1:07 PM, Gary Gregory < > garydgreg...@gmail.com> wrote: > >>>>>>>> Bah, let's pick one component and start there and skip a test > site. > >>>>>>> But then there is only one component visible under the new > commons.a.o? > >>>>>>> > >>>>>>>> Gary > >>>>>>>> > >>>>>>>> On Dec 10, 2012, at 3:08, Christian Grobmeier < > grobme...@gmail.com> wrote: > >>>>>>>> > >>>>>>>>> starting from 1. January. Just saw a final reminder from Infra. > >>>>>>>>> > >>>>>>>>> Commons is surely a LOT of work. > >>>>>>>>> > >>>>>>>>> I would like to suggest we act immediately. > >>>>>>>>> > >>>>>>>>> In other terms: let us request a commons-test cms where we can > try things > >>>>>>>>> out and prepare the new sites. > >>>>>>>>> > >>>>>>>>> As Ralph Goers has already mentioned, we have a similar > structure in > >>>>>>>>> logging land (1 main site, several sub sites) which might fit to > Commons > >>>>>>>>> too. > >>>>>>>>> > >>>>>>>>> Basically we have the Main-Site under CMS control; the subsites > are > >>>>>>>>> generated via Maven. The target of this generation is then > copied to > >>>>>>>>> another svn tree from which it will be taken and published. > >>>>>>>>> > >>>>>>>>> Obviously we will have to generate sites for each component > then, which > >>>>>>>>> will be a tough job with x-mas arriving. > >>>>>>>>> > >>>>>>>>> Thoughts? > >>>>>>>>> > >>>>>>>>> -- > >>>>>>>>> http://www.grobmeier.de > >>>>>>>>> https://www.timeandbill.de > >>>>>>>> > --------------------------------------------------------------------- > >>>>>>>> To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > >>>>>>>> For additional commands, e-mail: dev-h...@commons.apache.org > >>>>>>>> > >>>>>>> -- > >>>>>>> http://www.grobmeier.de > >>>>>>> https://www.timeandbill.de > >>>>>>> > >>>>>>> > --------------------------------------------------------------------- > >>>>>>> To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > >>>>>>> For additional commands, e-mail: dev-h...@commons.apache.org > >>>>>>> > >>>>>> > --------------------------------------------------------------------- > >>>>>> To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > >>>>>> For additional commands, e-mail: dev-h...@commons.apache.org > >>>>>> > >>>>> -- > >>>>> http://www.grobmeier.de > >>>>> https://www.timeandbill.de > >>>>> > >>>>> --------------------------------------------------------------------- > >>>>> To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > >>>>> For additional commands, e-mail: dev-h...@commons.apache.org > >>>>> > >>>> --------------------------------------------------------------------- > >>>> To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > >>>> For additional commands, e-mail: dev-h...@commons.apache.org > >>>> > >>>> > >>> --------------------------------------------------------------------- > >>> To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > >>> For additional commands, e-mail: dev-h...@commons.apache.org > >>> > >> --------------------------------------------------------------------- > >> To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > >> For additional commands, e-mail: dev-h...@commons.apache.org > >> > >> > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > For additional commands, e-mail: dev-h...@commons.apache.org > >
