Or, alternatively, instead of updating the contents of asf-site and removing master, per the JIRA discussion we could change the site to be published from master instead of asf-site ourselves by adding a .asf.yaml file: https://issues.apache.org/jira/browse/INFRA-19235
On Mon, 7 Oct 2019 at 16:09, Robbie Gemmell <robbie.gemm...@gmail.com> wrote: > > We can update the contents of the asf-site branch to suit our needs, > the only requirement from infra is that the published content is in > the content dir. The only thing we needed them to do would be change > the default branch. > > Daniel will be talking about the new abilities offered by the more > recently added asf.yaml stuff which can do various other things and > override the existing website publishing stuff: > https://cwiki.apache.org/confluence/display/INFRA/.asf.yaml+features+for+git+repositories > > We can certainly look to add more of the stuff like staging etc if > folks wanted, but I'd say its seperate from this change, and I dont > personally see a real need for it right now. > > Robbie > > On Mon, 7 Oct 2019 at 15:31, Clebert Suconic <clebert.suco...@gmail.com> > wrote: > > > > On Mon, Oct 7, 2019 at 9:37 AM Robbie Gemmell <robbie.gemm...@gmail.com> > > wrote: > > > > > > I'm not sure I see a need for creating another branch or asking infra > > > to assist in the update itself. We are effectively just adding the > > > source and history to the asf-site branch, which is where the site > > > bits have to go, byt overwriting it, which we can do. The only thing > > > we should really need infra to do is change the default branch setting > > > to asf-site after the update is done, and then master can be removed. > > > > We could certainly do that. I just wanted to make sure Infra would > > have no issues with the change. > > > > Daniel is talking about a yaml file I never heard about, even talking > > about staging website (which could be useful as part of our release > > processes). > > > > I'm getting confused through the JIRA. and I need help :) > > > > > > > > On Mon, 7 Oct 2019 at 13:51, Clebert Suconic <clebert.suco...@gmail.com> > > > wrote: > > > > > > > > I will do the right thing. I will create a branch called future-master. > > > > And ask infra to update everything. > > > > > > > > I will send a separate thread with a clear title. > > > > > > > > On Mon, Oct 7, 2019 at 6:42 AM Robbie Gemmell <robbie.gemm...@gmail.com> > > > > wrote: > > > > > > > > > A necessary change would be updating the readme to describe the > > > > > updated build + publish process. I've raised a PR against your test > > > > > branch with suggested changes: > > > > > https://github.com/clebertsuconic/activemq-website/pull/1/files > > > > > > > > > > Robbie > > > > > > > > > > On Mon, 7 Oct 2019 at 10:33, Robbie Gemmell <robbie.gemm...@gmail.com> > > > > > wrote: > > > > > > > > > > > > Do you mean move things over to the asf-site branch? Seems > > > > > > reasonable > > > > > > to me. I'd start a separate thread though giving folks a heads up > > > > > > that > > > > > > e.g you will do it tomorrow/some-other-point unless discussed > > > > > > otherwise, so its clear when the change might happen and that lazy > > > > > > consensus can be taken as only a few folks have commented on this > > > > > > thread. > > > > > > > > > > > > On Fri, 4 Oct 2019 at 18:02, Clebert Suconic > > > > > > <clebert.suco...@gmail.com> > > > > > wrote: > > > > > > > > > > > > > > Should we move it then ? > > > > > > > > > > > > > > On Fri, Oct 4, 2019 at 4:52 AM Robbie Gemmell < > > > > > robbie.gemm...@gmail.com> > > > > > > > wrote: > > > > > > > > > > > > > > > What you did is just what the exiting build script did, so that > > > > > should > > > > > > > > be fine if it was previously. I tried with and without the > > > > > > > > --incremental and it didnt make a difference here (wasn't much > > > > > > > > different time wise either). Strange that you saw this, and > > > > > > > > strange > > > > > > > > you now dont, but at least its working now. > > > > > > > > > > > > > > > > On Thu, 3 Oct 2019 at 22:07, Clebert Suconic < > > > > > clebert.suco...@gmail.com> > > > > > > > > wrote: > > > > > > > > > > > > > > > > > > I did not see the error again. I guess it was an effect of > > > > > > > > > working > > > > > > > > > late hours.. I messed something up. > > > > > > > > > > > > > > > > > > I was playing with --incremental on jekyll perhaps that's what > > > > > caused it. > > > > > > > > > > > > > > > > > > > > > > > > > > > Playing with a single branch would make it a lot easier. It > > > > > bothers me > > > > > > > > > to keep moving between branches as my IDE gets crazy on > > > > > > > > > indexing > > > > > files > > > > > > > > > (don't ask.. I prefer IDE than VIM... I use vim emulation on > > > > > > > > > idea > > > > > > > > > though ;) ) > > > > > > > > > > > > > > > > > > > > > > > > > > > If more people could double check this is okay.. perhaps we > > > > > > > > > can > > > > > > > > > replace asf-site by my test branch. > > > > > > > > > > > > > > > > > > On Thu, Oct 3, 2019 at 3:09 PM Clebert Suconic > > > > > > > > > <clebert.suco...@gmail.com> wrote: > > > > > > > > > > > > > > > > > > > > I copied the folder from the branch. > > > > > > > > > > > > > > > > > > > > Check out asf-siite > > > > > > > > > > Move it. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Checkout the test branch > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Moved it back. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Thu, Oct 3, 2019 at 7:31 AM Robbie Gemmell < > > > > > > > > robbie.gemm...@gmail.com> wrote: > > > > > > > > > >> > > > > > > > > > >> I dont see that happen when I run the build, and hadnt > > > > > > > > > >> actually > > > > > ever > > > > > > > > > >> seen the 'typechange' file status before. > > > > > > > > > >> > > > > > > > > > >> The content/schema/core/activemq-core.xsd is actually a > > > > > symlink, and > > > > > > > > > >> some googling suggests the 'typechange' status mainly > > > > > > > > > >> happens > > > > > when > > > > > > > > > >> such link is replaced with an actual file. If I do that > > > > > deliberately, > > > > > > > > > >> I then see the same status you do. Some posts suggested it > > > > > > > > > >> can > > > > > happen > > > > > > > > > >> with certain copy commands, that have dereferenced the > > > > > > > > > >> link and > > > > > copied > > > > > > > > > >> the referenced files content. How did you initially > > > > > > > > > >> populate the > > > > > > > > > >> 'content' dir your subsequent status is showing as being > > > > > updated? > > > > > > > > > >> Running the build afresh, or copying prior build output? > > > > > > > > > >> > > > > > > > > > >> Robbie > > > > > > > > > >> > > > > > > > > > >> On Wed, 2 Oct 2019 at 22:53, Clebert Suconic < > > > > > > > > clebert.suco...@gmail.com> wrote: > > > > > > > > > >> > > > > > > > > > > >> > I have a branch where I'm using a single branch: > > > > > > > > > >> > > > > > > > > > > >> > https://github.com/clebertsuconic/activemq-website/tree/test > > > > > > > > > >> > > > > > > > > > > >> > > > > > > > > > > >> > The only concern I have so far is that any time I build, > > > > > > > > > >> > I > > > > > get a > > > > > > > > > >> > change into a schema: > > > > > > > > > >> > > > > > > > > > > >> > Changes not staged for commit: > > > > > > > > > >> > (use "git add <file>..." to update what will be > > > > > > > > > >> > committed) > > > > > > > > > >> > (use "git checkout -- <file>..." to discard changes in > > > > > working > > > > > > > > directory) > > > > > > > > > >> > > > > > > > > > > >> > typechange: content/schema/core/activemq-core.xsd > > > > > > > > > >> > > > > > > > > > > >> > > > > > > > > > > >> > > > > > > > > > > >> > Although I would consider that anyone would check stuff > > > > > > > > > >> > before > > > > > > > > committing.. > > > > > > > > > >> > > > > > > > > > > >> > Any ideas? > > > > > > > > > >> > > > > > > > > > > >> > On Wed, Oct 2, 2019 at 1:03 PM Justin Bertram < > > > > > jbert...@apache.org> > > > > > > > > wrote: > > > > > > > > > >> > > > > > > > > > > > >> > > If it's simpler and documented that sounds like a win > > > > > > > > > >> > > to me. > > > > > > > > > >> > > > > > > > > > > > >> > > > > > > > > > > > >> > > Justin > > > > > > > > > >> > > > > > > > > > > > >> > > On Wed, Oct 2, 2019 at 11:41 AM Robbie Gemmell < > > > > > > > > robbie.gemm...@gmail.com> > > > > > > > > > >> > > wrote: > > > > > > > > > >> > > > > > > > > > > > >> > > > On Wed, 2 Oct 2019 at 16:47, Clebert Suconic < > > > > > > > > clebert.suco...@gmail.com> > > > > > > > > > >> > > > wrote: > > > > > > > > > >> > > > > > > > > > > > > > >> > > > > On Wed, Oct 2, 2019 at 10:44 AM Robbie Gemmell < > > > > > > > > robbie.gemm...@gmail.com> > > > > > > > > > >> > > > wrote: > > > > > > > > > >> > > > > > Thats what we do at Qpid and what I've seen other > > > > > > > > > >> > > > > > projects do, I find it simpler overall. > > > > > > > > > >> > > > > > > > > > > > > > >> > > > > Tl;DR: > > > > > > > > > >> > > > > I'm all +10000000 on this... what we need to > > > > > > > > > >> > > > > change on > > > > > infra. > > > > > > > > > >> > > > > > > > > > > > > > >> > > > > Can we keep the history from master > > > > > > > > > >> > > > > > > > > > > > > >> > > > > > > > > > > > > >> > > > We wouldnt need to change much to get everything on > > > > > > > > > >> > > > the > > > > > same > > > > > > > > branch I > > > > > > > > > >> > > > dont think, and it seems like we can do so without > > > > > > > > > >> > > > infra: > > > > > I > > > > > > > > just tried > > > > > > > > > >> > > > on your last commit as a test and was able to > > > > > > > > > >> > > > force-push > > > > > the > > > > > > > > asf-site > > > > > > > > > >> > > > branch in the website repo. > > > > > > > > > >> > > > > > > > > > > > > >> > > > So basically it seems it needs something like: > > > > > > > > > >> > > > repopulate > > > > > the > > > > > > > > asf-site > > > > > > > > > >> > > > branch (or a test branch) with the master history > > > > > > > > > >> > > > locally, > > > > > > > > simplify > > > > > > > > > >> > > > the build script to only build and nothing else, fix > > > > > > > > > >> > > > up > > > > > the > > > > > > > > .gitignore > > > > > > > > > >> > > > file appropriately, build things, commit, and [force] > > > > > push. > > > > > > > > > >> > > > > > > > > > > > > >> > > > Assuming of course others agree that it is what > > > > > > > > > >> > > > should > > > > > happen? > > > > > > > > > >> > > > > > > > > > > > > >> > > > Robbie > > > > > > > > > >> > > > > > > > > > > > > >> > > > > > > > > > > >> > > > > > > > > > > >> > > > > > > > > > > >> > -- > > > > > > > > > >> > Clebert Suconic > > > > > > > > > > > > > > > > > > > > -- > > > > > > > > > > Clebert Suconic > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > > > > Clebert Suconic > > > > > > > > > > > > > > > -- > > > > > > > Clebert Suconic > > > > > > > > > -- > > > > Clebert Suconic > > > > > > > > -- > > Clebert Suconic