Ok picking up this thread again ...
Well I think that allowing non-comitters to contribute to the page is a pretty slim benefit. I couldn't say that we have used that even for a single time in the past. And I think it would be good not to allow this in general. At least I couldn't find a way to release only parts of the site, so it's an all or nothing. And I also couldn't see any "this has changed" so you would actually have to verify the entire page just to make sure no one pushed anything bad. So I think it's a good thing that non-comitters can't contribute to the site. This way we also don't have to worry about signed ICLAs and so. Today I have continued on the new maven-style site and added stuff that was obviously missing: pmc_Flex.rdf, doap_Flex.rdf (don't quite know what they are good for ... looks as if this is some internal stuff for the ASF to gather information about it's own projects) ... at least that's in now. What's still missing is some installer-related stuff. I can see: - track-installer.cmsPage, - all the single-mirror pages - sdk-installer-config xmls I would assume the stuff in the sub-directory would actually come from sub-projects. One thing I couldn't find in the codebase however was the footer of each page containing the Twitter feed, the latest releases, the about flex text and how to subscribe to mailinglists. I couldn't find this content anywhere :-( And I would like to state that I didn't expect this to be a race in which the first one wins. If there are other initiatives to generating the site in git, why not simply setup a separate branch and put that stuff in there? I could change the buildbot job to push to "asf-site-maven" and different solutions could co-exist. All that I want, is the ability to produce the sites of sub-modules such as FlexJS (and it's part) automatically. Chris ________________________________ Von: Alex Harui <aha...@adobe.com> Gesendet: Freitag, 14. Oktober 2016 17:20:05 An: dev@flex.apache.org Betreff: Re: AW: AW: AW: AW: [Website] Progress on the Website-Generation topic On 10/14/16, 12:42 AM, "Christofer Dutz" <christofer.d...@c-ware.de> wrote: > >So if we are comparing the new three-step user workflow: > >1. Do changes > >2. Test them locally with a "mvn clean site" > >3. Commit changes > > >to the current user workflow, which is: > >1. Do changes > >2. Commit changes > >3. Wait for buildbot to stage the changes > >4. Check if the output is what we want in the staging area > >5. Go to the Web-Ui and release all changes at once (Could be a problem >that you release stuff someone else is currently working on, but I guess >that's a more theoretical problem) > I wouldn't mind getting rid of that step 5 in the current workflow, so if there are enough volunteers to re-create our site in Git that's fine with me. Interestingly, I just saw on another list that a supposed advantage of the current workflow is that non-committers can use ASF CMS and its web UI to make changes. Collecting doc changes from non-committers might be a good thing. Our wiki has a separate account system that allows us to do that as well. But would anyone in the community want to leverage that? Maybe using the Web UI is old-school and submitting pull requests is sufficient. So, in sum, no objections from me but I doubt I'll be much help on finishing the port. I am, however, curious as to how this will impact Nick. I think he was in-progress on a web-site revamp. Thanks, -Alex
