Minutes of LwDITA syntax call

Minutes of LwDITA syntax taskforce call

08/31/2016

We had our first LwDITA syntax call and these are our notes/recommendations for discussion with the SC.

Present: Carlos Evia, Mark Giffin, Rob Hanna, Keith Schengili-Roberts, Eric Sirois, and Don Day (for a few minutes)

Summary: The taskforce is not opposed to keeping the HTML data attributes originally proposed by M. Priestley as part of HDITA. HDITA should be as close as possible to XDITA (with reuse and even specialization if possible). MDITA, on the other hand, should be more a recommendation than a standard. Standardizing Markdown has failed before, and potential MDITA users have idiosyncratic authoring ways that we should accept with minor modifications.

- Do we need concept, task, reference (and troubleshooting) in MDITA and HDITA?

It would be a good idea to ship those templates in HDITA. MDITA should remain at a basic topic model. Not a standard but a recommendation. We should make HDITA as similar as XDITA as possible. MDITA could fail if we enforce information typing. Developers create Markdown topics in many flavors and it would be a good idea to accept them all with optional recommendations in sections/sub-headings.

- Do we keep data attributes in HDITA elements or try to simplify markup with custom elements?

HDITA authors from marketing communication and web communities will never see the code. They will most likely use a WYSIWYG editor inside a CMS. Data attributes are easy and, as seen in the Evia & Priestley 2016 STC article, work. Any author feeling comfortable enough to use code should probably go to XDITA or DITA. The info type-specifc component of HDITA’s attributes (example the “topic” in <section data-hd-class="topic/example”>) seems unnecessary, but the taskforce made a good case in its defense as it could be useful for HDITA specialization by template.

- Do we need metadata in HDITA and MDITA?

YAML headers can be used as identifiers, prolog containers, and even template indicators. In MDITA, for example, a “task" value in a YAML header can can validate sections and even generate titles. More complex metadata should be processed externally.

- Do we need single-source capabilities in HDITA and MDITA?

Yes in HDITA. Make as similar to XDITA as possible. Not a lot in MDITA. Markdown should be mainly for text and not structure

- Who will use HDITA and MDITA?

From DITA listening sessions, MDITA is for developers. HDITA is for marketing and web communities. We could reach out to marketing communication groups with a survey. We probably need round-tripping of content to use Markdown and HTML5 as source assets but also send back to wikis, GitHub books and repos and native environments.

- Action items:

Evia will work on better examples of MDITA and HDITA with suggested metadata.

We will probably have a second call in mid September.

dita-lightweight-dita message