Project Description

Our SUSE Support knowledge base sucks. It has a clunky UI, only has a What you see is what you (don't) get editor, doesn't show version history. I hate it, you hate it, and our customers hate using it and would rather pick up the phone and bother us.

This project's aim is to move away from Salesforce Knowledgebase to a FLOSS solution that allows:

• versioning and diffs
• defined formatting
• better text search
• tag and category organization (let's at least try to match wiki functionality)
• reverse link search and link validation (I don't know exactly what it's called, but the thing in wikimedia that allows us to see what links to what so we don't create orphan links)
• prettiness

What I'm envisioning is a repository of asciidoc markdown files in GitLab that are turned into XML Docbook articles, and we can apply the standard SUSE product documentation style sheet to create documents in a pdf and html format. You want a .pdf to send to the customer? No sweat. You want .html for the website? Easy.

How beautiful and consistent! How easy to use and learn!

Versioning and standard git best practice allow the review and quality control that's needed. Changes made by contributors would be tracked and approved by editors with privileges and responsibilities to publish by pulling into the master branch. From that, the KB article would be regenerated. Documentation as code.

DAPS written by Frank Sundermeyer et al. from Documentation is likely the best solution for this.

Goal for this Hackweek

I would like to have a prototype to show support management to convince them that it would be possible and desirable to migrate away from Salesforce knowledge base.

We'd build a tool chain that would flow asciidoc from a SUSE GitLab repo into XML Docbook and generate a downloadable pdf and an html file. Or even epub. You laugh, but having reference documentation on an ebook reader is a great help in the datacenter.

Since we don't want to lose all the work we've done, the real challenge is to come up with a migration tool chain as well. It seems as though it is possible to pull XML from Salesforce (whether DocBook valid remains to be seen) as well as some metadata.

We'd need to validate and tweak the result by hand in Support, but I hope we will do this once, and we'll be FREE from proprietary nonsense and join the GNU generation.

Resources

GitLab/GitHub repo soon.

I'm looking for buddies who have XML docbook pipeline experience, knowledge management experience, good ideas on how to make this idea into a better design that's easy to maintain, and are willing to lend a hand. :)

Particularly useful would be input and experience from the Product Documentation team, which operate the XML docbook templates on which we would like to piggyback.