Documentation Revamp / Organisation

[Stefterv]

The examples and some of the documentation around Processing are kind of all over the place

The Goal

Have one repo where documentation lives: processing-website

The current issues

• processing-website
  • updates processing4's keywords.txt based on content generated through processing-doclet
  • has to manually update content/examples and it is unclear that this data is partially read-only, the script applies processing-examples on top
  • content/contributions is also readonly and is no longer compatible with the new processing-contributions pipeline
• processing4-javadocs does not automatically update with each release
• processing-examples does not contain the images/livesketches/metadata that the website needs.
• processing-doclet is a separate repo that will create the parsed javadoc comments for the website

Suggested actions

• Get rid of any instances where the developer has to make sure repositories live next to each other on disk
• Migrate processing-doclet into the processing-website and make it run on build
  • Research further into migrating / replicating the custom doclet into Gradle and make it run w gradle
• Delete processing-doclet repo
• Migrate processing-examples into the processing-website
  • Tested, still need to figure out how to merge the folders whilst keeping history
• Archive processing-examples
• Remove all read-only files from processing-website
• Move the generation of keywords.txt back into the processing4 repo
• Add an automatic update trigger for processing4-javadoc to the processing4 repo release action (or migrate processing4-javadoc into the processing4 repo)
  • Migrate Javadoc from Ant to Gradle

Related PRs/Issues/Other documentation

#1204

[Stefterv]

Let me know if there is anything I missed @catilac @SableRaf

[SableRaf]

I tried to summarize the current architecture (thanks Stef for your feedback)

[SableRaf]

I can't be completely sure we're not missing any obscure dependencies but this should cover the big ones. I fully support any efforts to simplify and automate this legacy tangle of repos.

[Stefterv]

For the examples directory seems like we can merge in the history

[Stefterv]

For the Processing docket, I will move that functionality into the processing4 repo along with the functionality of the keywords.txt from the website

[catilac]

I can't be completely sure we're not missing any obscure dependencies but this should cover the big ones. I fully support any efforts to simplify and automate this legacy tangle of repos.

Yeah this is the thing I'm the most concerned about with this one. So I think we should move slowly and thoughtfully