[TYPO3-doc] Milestones for the Docs Workflow

[Fabien Udriot]

Hi,

After the reST migration happened, one of the next logical step is the set up of the rendering
workflow that we briefly discuss during our last Doc Meeting. To replace the veteran EXT:ter_doc, we
are aiming to use FLOW3, following the path [1] opened by Sebastian Kurfürst some time ago.

FYI, there is already a prototype that is running (and working) on our testing server using
beanstalkd as queue service. That is good because we can start hacking with some concrete material
already.

We should probably organize an ad-hoc meeting and coordinate people but I am putting these topic on
the ML to collect ideas and feedbacks. Feel free to add your points in the tasks below.

Thinking of the next milestones, I have split them up into three areas:

1) *ReST transformation*.

Output format would probably be HTML, JSON, and PDF. JSON output could be stored directly in the
content repository within nodes (cf. point 2). There is already some existing code for that.

* Finish the current implementation so that it can properly generate and output something more than
just a hello world.

* Define a way to be able to trigger the docs service. I am thinking we can use the RESTfull API of
FLOW3 to achieve that.

* Define and configure the different post hooks that would call the docs service:
  * upon Git / Gerrit merge (I have a skeleton script from Peter Niederlage for doing that)
  * upon TER upload (should be hacked into the TER webservice EXT:ter)
  * upon User request. GUI and concept needs to be clarified and has a "nice to have" priority

2) *Content storing*.

We want to be able to store some data inside the content repository so that we are able to query and
fetch content later on e.g. dress up a list of all manuals.

* Define the model. There is a source of inspiration within a thread above in this ML. There could
be some addition stuff to take from EXT:ter_doc
* Program the Model and Repository.
* Integrate the Model in the queue.

3) *Content output*

* Define what output should be generated.

  * Among other, I have currently in mind the indexing output for Solr. We discussed the topic with
Ingo Renner by the typo3.org relaunch and a first version was programmed based on EXT:ter_doc. This
could be re-use as a starting point.

  * A docs entry point: get inspiration from section "Recently Updated Projects" from
http://readthedocs.org/

  * typo3.org integration could be brainstormed / implemented here as well.

* Program the output, basically.

Cheers,

Fb.

[1] http://lists.typo3.org/pipermail/typo3-project-documentation/2012-May/003754.html