At Flock 2016 in Krakow, Poland, I had the privilege of updating the community on the status of the Fedora Docs Project.

I made a small presentation and moderated a discussion in the Hackfest: Fedora Docs Learn and Hack panel. Unfortunately, my co-presenter and Fedora Docs Project Lead, Pete Travis, could not attend this year.  Therefore a lot of the conversation reflected my opinions and what I have gleaned from others.

The presentation slides are online. Unfortunately, the session wasn’t recorded or transcribed, so I wanted to try and present the conversation here. I am not attributing any comments in order to avoid mistakes. Additionally, I am working from my memory and the memory of other attendees, so omissions are accidental.

Two focuses for the Docs Project

There was a FAD in May 2016 to formulate ideas for moving the project forward. Two big ideas came out of this meeting:

First, a desire to move to a topic-based style of writing. This changes our writing to thinking about “every page is page one”.  This style is shorter and refers to pre-requisite steps and knowledge as needed. This makes it easier to submit new material as writers don’t have to figure out how to fit their contribution into the narrative flow of a large book. Lastly, it is easier to consume this more directed and self-contained writing and it will score better in search.

Second, tools were debated. DocBook and publican seem to have led to problems with contributions, lots of friction in the project, and longer on-boarding.  Additionally, there are the problems associated with a relatively unmaintained upstream. Prior to the FAD, nb had moved our repositories from FedoraHosted to Pagure as part of the effort to join the new git forge. At the FAD, a lot of tools were analyzed and considered. In the end, the discussion led to the idea of using the work flow provided by Pagure and building the site with Pintail.

The real value came with the questions and comments raised as part of the discussion. In no particular order, these major points were raised.

Outcomes from discussion at Flock

How do we move to topic-based writing? Is there a plan?

After the FAD, there was not a finalized plan, but planning began during this session. Many ideas were mentioned and consensus seemed to form around just writing new topic material. Folks from the GNOME project pointed out several reasons, including efficiency, for why they rewrote their materials when shifting from books to topics. Another benefit of starting from scratch is that new material can be written in a priority order, possibly based on search keywords.

One challenge is right now is that we have no place to put and publish these new topics. It was quickly pointed out that this is the “tools problem” and it has many solutions. Instead of letting tools be a blocker, it was proposed to have people just write. “Write it in any format, any markup, any program, even on paper and just send it to us. We will get it published.” This turned into a discussion of how many topics are also good for the Fedora Magazine. So the suggestion was made for folks to consider submitting material there first and we can pull it back into docs later. Additionally, folks can email the docs mailing list or me. I am super excited to tell you I got an email with a topic 35 minutes after the session ended

Continue the conversation in this thread.

Should we stop publishing the current guides now?

The requirement to keep publishing the current guides feels very self-imposed. Continuing to publish them is a challenge for the new tooling as it has to be built to accommodate the past and therefore slows down the future.

Additionally, publishing the current books spreads our resources very thinly, if not past the breaking point. It also creates inertia which prevents the move to topics. Confusion can result from this as well because contributors don’t know what to update (old books or new topics).

Lastly, there is a growing belief in the larger documentation community that no docs is better than old docs. Here this is a direct reference to the fact that we don’t republish all the docs for every release and we don’t thoroughly review every doc that is published. Versioned docs are important, but some old materials are probably going to cause problems (i.e. references to yum or iptables.)

One proposal was to have a “flag day” where we stop updating the current docs and another day (or same day) where we stop the publication. this would definitely need to be moderated for versions not yet end-of-life.

Continue the conversation in this thread.

Translation needs clarity on how to get updates published and the process.

This seemed like a communication problem between the two projects that needed to be resolved with better docs on the process and hand-off procedures. Because the tooling proposal will hopefully include continuous deployment, this may become a lot easier in the future.

Continue the conversation in this thread.

Tooling, tooling, tooling

There was strong consensus around changing to new tools and markups. In fact, most of the tooling conversations were held in small groups near the end of the meeting. There was a desire to continue to see a drive to simpler contribution and publication.

The only significant question was around the community, upstream adoption, and contributor base for Pintail, which is the central tool in the new processing flow. People were concerned that there wasn’t evidence of enough adoption and contribution to prevent the project from being at risk of going either unmaintained or slowly maintained.

Continue the conversation in this thread.

Add your voice

You are strongly encouraged to continue these conversations on the Fedora Docs Mailing List, in IRC in #fedora-docs on freenode, and in our weekly IRC Meeting on Mondays at 1400 GMT in #fedora-meeting on freenode.


Image courtesy João Silas – originally posted to Unsplash as Untitled, modified by Justin W. Flory.