Re: [dita] Review of "Add a multimedia domain #27"

Keith,

Apologies for the long delay on this. Things were a bit crazy before I went on vacation last week and this fell through the cracks.

Re: the decision to add nested elements with true/false values. In DITA 1.3 and lightweight DITA, we’re not allowed to add arbitrary attributes to specific elements; specializations of @base are global and can’t be targeted without extensive constraints. Since one of the drivers for this domain is consistency between DITA 1.3, DITA 2.0, and LwDITA, we have to use the ‘least common denominator’ approach here, which means using nested <data> specializations. Also, of course, it’s impossible to have an attribute without a value in XML. That’s the main reasoning behind this design.

Note that the @value attributes on the nested elements are all #IMPLIED. My thinking here – though I now realize I never made it explicit in the proposal – was that if the @value attribute is missing, “true” should be assumed. Thus, the mere presence of one of these elements will be sufficient to turn that option “on”. For example:

<media-controls/>

</video>

The notes about the examples are well taken and I’ll address those in the Stage 3 proposal.

Chris

From: <dita@lists.oasis-open.org> on behalf of Keith Schengili-Roberts <keith.roberts@ixiasoft.com>
Date: Monday, March 5, 2018 at 12:48 PM
To: "dita@lists.oasis-open.org" <dita@lists.oasis-open.org>
Subject: [dita] Review of "Add a multimedia domain #27"

I had already sent this separately to Chris directly, but I think I am supposed to send this to the group.

The following was originally meant as a stage 2 critique, but I understand from Chris that the following should prove useful for his work on stage 3.

--Original message --

The proposal looks great, and I checked everything against the latest spec for HTML5 and everything stands up, ditto the arguments for the attributes that are deliberately being excluded.

A question regarding the associated elements: your DTD and example code suggests that media-controls, media-autoplay and media-loop (for example) should be value="true|false", but I note that the equivalent in HTML5 is more like <audio controls="controls"> or <audio loop="loop"> or <audio controls loop> (all equivalents for "true"). While I'm not advocating the W3C's "loopy" approach to this, am wondering if it might be worthwhile to be more explicit as to why we are taking this approach.

My only other quibbles are with the examples. In the first instance, you reference an audio file ("An audio element with a single source") but the example uses the video element rather than audio. From a technical perspective I think this is possible, but I think the intention was to show an example using the audio element. Also, from a best practices perspective, isn't it always a good idea to throw in fallback? That would make a revised audio example look like this:

The other examples are fine, I would just like to see more of them.

Suggested code for audio with multiple sources/audio types:

<audio data="">    <media-source value="mysong.wav" type="audio/wav" />
   <media-source value="mysong.ogg" type="audio/ogg" />
   <fallback="Audio playback is unavailable" />
</audio>

Suggested code for audio with audio controls explicitly turned on, and automatic looping disabled:

Cheers!

Keith Schengili-Roberts

Market Researcher and DITA Evangelist

IXIASOFT

825 Querbes, Suite 200, Montréal, Québec, Canada, H2V 3X1

tel + 1 514 279-4942 / toll free + 1 877 279-4942

robertsk@ixiasoft.com / www.ixiasoft.com

dita message