TTS note title

[mattgarrish]

One issue I forgot to raise for the initial extraction is what do we name this thing. For now, I've named it "EPUB 3 Text-to-Speech Enhancements".

The oddity here is that this is neither new (it succeeds from 3.2) nor a rec-track 3.3 specification, so it's just as odd to number it "EPUB Text-to-Speech Enhancements 1.0" as it is to number it "EPUB Text-to-Speech Enhancements 3.3".

I don't think the name I have right now works, though, since it doesn't have a incrementable number, so open to other ideas on what to call or number it.

[iherman]

We have named multiple renditions as "EPUB Multiple-Rendition Publications 1.1", although, I believe, the same arguments apply. Maybe, to be consistent, we could call it "EPUB Text-to-Speech 1.1".

[mattgarrish]

although, I believe, the same arguments apply.

Not really. What is defined in the multiple renditions spec is all brand new stuff - rendition selection attributes, mapping documents, grammar for the metadata.xml file and whatever else. There's no legacy of any of that stuff existing in the 3.0+ that you would ever refer to.

It'd be odd to number it to match multiple renditions and not have a "1.0" anywhere, though.

Maybe we can mash everything together and get "EPUB 3 Text-to-Speech Enhancement 1.0".

Even "EPUB 3 Multiple Rendition Publications 1.1" sounds better than what we have, since the specification is not applicable to EPUB 2 or a future 4+. (Suggesting we rename that one, not a typo!)

[iherman]

although, I believe, the same arguments apply.

Not really. What is defined in the multiple renditions spec is all brand new stuff - rendition selection attributes, mapping documents, grammar for the metadata.xml file and whatever else. There's no legacy of any of that stuff existing in the 3.0+ that you would ever refer to.

What I meant was: multiple rendition was part of EPUB 3.2, and we took it aside due to poor implementation. There are similarities in this respect... (O.k., I know, there was a separate IDPF 1.0 version, so that is different.)

It'd be odd to number it to match multiple renditions and not have a "1.0" anywhere, though.

That is true

Maybe we can mash everything together and get "EPUB 3 Text-to-Speech Enhancement 1.0".

Works for me

[mattgarrish]

Closing this issue as I've update the title in 4f667e2 and the short name in 633ad4b

[iherman]

The issue was discussed in a meeting on 2021-06-24

• no resolutions were taken

View the transcript

2. Text-To-Speech

See github issue #1715, #1717.

Wendy Reid: last week we discussed what to do with PLS and SSML in the spec. We aren't getting rid of them, but are separating them into their own separate document.
… the issues were just doing this

Matt Garrish: yes, we pulled them out, and added a little intro type prose
… we talked about what to name these documents in one of the issues
… but like we said last time, these parts just felt a little underspecified (and still do)

Wendy Reid: I started doing a little bit of research into what RS might have implemented TTS, this is ongoing

Matt Garrish: DAISY has some readers that can do TTS, but even when it comes to PLS they don't implement it exactly as per spec
… about naming, we were going for something that tied it to our spec, but also gave it a 1.0 version number

Ben Schroeter: VitalSource has TTS, but not sure if they use SSML and PLS

Brady Duga: same with us, but we just use OS level TTS functionality