[Tts-project] does the TTS vs a11y split still make sense?

[Paul Gevers]

Hi,

On 13-02-18 00:47, Samuel Thibault wrote:
> Sergio Oller, on jeu. 01 févr. 2018 18:15:59 +0100, wrote:
>> 2018-01-31 21:49 GMT+01:00 Paul Gevers <[1]elbrus at debian.org>:
>>> Wouldn't the TTS packages receive more exposure in the realm of the
>>> a11y team
> 
> I'm not sure about this.  I mean, it's not necessarily obvious to people
> that text-to-speech can be found in accessibility packages.  Making TTS
> a separate team is also a way to express that it needs its own set of
> maintainers, so the accessibility team can concentrate on purely a11y
> stuff.  Put another way, it avoids the reasoning "a11y people handle
> these TTS packages".  Sure, there is a lot of overlap in the teams, but
> the idea was that perhaps we can drag people into the TTS team for pure
> TTS reasons, who may not have been dragged into the a11y team.

I wonder if most teams work so effectively in Debian.

>>> (e.g. due to the fact that the a11y team is explicitly
>>> listed on udd.d.o/dmd/)?
> 
> Uh, I didn't even know that it was there.  Did we really drag people
> that way? :)

No idea. But I think there are some people now part of the a11y team on
salsa that I wasn't aware of they were lurking (if that is what they are
doing).

>>> AFAIK, the last couple of years the uploads of TTS packages have been
>>> done by members of both teams.
> 
> Yes, there is a lot of overlap.  That doesn't mean the teams necessarily
> have to merge :)

Sure, but my point is that there is additional effort on our/my part.
E.g. UDD¹ can only show the overview for three addresses, so TTS is
currently missing from *my* default overview (two a11y addresses and my
own). Also, we'll need to migrate the e-mail list sooner or later
(support is suggested to last for another year, depending on demand).

>>> *If* the split is still desired, I *guess* it makes sense to migrate
>>> festvox-mbrola, flite, eflite, espeak, espeak-ng, speech-tools and svox
>>> (and maybe more) to the TTS team, but I highly doubt it is worth the
>>> effort.
> 
> It would make sense indeed. With cicero, espeakedit, pcaudiolib, svox,
> the list would be complete. There's also sonic, but that's really
> something for a11y: making the speech so fast is only for speech
> listeners :)
> 
> I would however not care too much about taking the effort of moving them
> all just for the sake of it.  If TTS people feel they'd like to maintain
> some of them, then we can just move them.

OK. I'll not migrate them for the sake of it. But when I am touching a
package anyway, I'll consider moving them. If we want to keep the two
teams, I think it is good to send the right message about (intended)
ownership.

> All that said, in the end for me the question is: did the existence of
> the TTS group drag some people into it that would perhaps not joined the
> A11Y group?  If yes, I'd say we should keep it.  Otherwise we can merge
> indeed.

Nice metric. The TTS team on Alioth does contain people that are not
member of the a11y team on Alioth (although the picture on salsa isn't
so clear yet). So indeed, let's keep it then.

Paul

¹
https://udd.debian.org/dmd/?email1=elbrus%40debian.org&email2=pkg-a11y-devel%40lists.alioth.debian.org&email3=debian-accessibility%40lists.debian.org#todo

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 488 bytes
Desc: OpenPGP digital signature
URL: <http://lists.alioth.debian.org/pipermail/tts-project/attachments/20180213/4b46bc05/attachment.sig>