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 <[email protected]> wrote:
>
> On Mon, Oct 7, 2019 at 9:37 AM Robbie Gemmell <[email protected]> 
> 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 <[email protected]> 
> > 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 <[email protected]>
> > > 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 <[email protected]>
> > > > 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 
> > > > > <[email protected]>
> > > > wrote:
> > > > > >
> > > > > > Should we move it then ?
> > > > > >
> > > > > > On Fri, Oct 4, 2019 at 4:52 AM Robbie Gemmell <
> > > > [email protected]>
> > > > > > 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 <
> > > > [email protected]>
> > > > > > > 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
> > > > > > > > <[email protected]> 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 <
> > > > > > > [email protected]> 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 <
> > > > > > > [email protected]> 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 <
> > > > [email protected]>
> > > > > > > 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 <
> > > > > > > [email protected]>
> > > > > > > > >> > > wrote:
> > > > > > > > >> > >
> > > > > > > > >> > > > On Wed, 2 Oct 2019 at 16:47, Clebert Suconic <
> > > > > > > [email protected]>
> > > > > > > > >> > > > wrote:
> > > > > > > > >> > > > >
> > > > > > > > >> > > > > On Wed, Oct 2, 2019 at 10:44 AM Robbie Gemmell <
> > > > > > > [email protected]>
> > > > > > > > >> > > > 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

Reply via email to