Re: text-wrap balance

From: Silvia Pfeiffer <silviapfeiffer1@gmail.com<mailto:silviapfeiffer1@gmail.com>>
Date: Wednesday, 19 June 2019 at 08:33
To: Glenn Adams <glenn@skynav.com<mailto:glenn@skynav.com>>
Cc: "David (Standards) Singer" <singer@apple.com<mailto:singer@apple.com>>, Nigel Megitt <nigel.megitt@bbc.co.uk<mailto:nigel.megitt@bbc.co.uk>>, Philippe Le Hegaret <plh@w3.org<mailto:plh@w3.org>>, Pierre-Anthony Lemieux <pal@sandflow.com<mailto:pal@sandflow.com>>, Gary Katsevman <me@gkatsev.com<mailto:me@gkatsev.com>>, Cyril Concolato <cconcolato@netflix.com<mailto:cconcolato@netflix.com>>, Public TTWG List <public-tt@w3.org<mailto:public-tt@w3.org>>
Subject: Re: text-wrap balance



On Wed., 19 Jun. 2019, 1:45 pm Glenn Adams, <glenn@skynav.com<mailto:glenn@skynav.com>> wrote:


On Tue, Jun 18, 2019 at 6:02 PM Silvia Pfeiffer <silviapfeiffer1@gmail.com<mailto:silviapfeiffer1@gmail.com>> wrote:
On Tue, Jun 18, 2019 at 10:32 PM David Singer <singer@apple.com<mailto:singer@apple.com>> wrote:
>
>
>
> > On Jun 18, 2019, at 14:19 , Silvia Pfeiffer <silviapfeiffer1@gmail.com<mailto:silviapfeiffer1@gmail.com>> wrote:
> >
> >
> >
> > While I agree with this and also have no problem excluding this from REC, is still like to encourage the CSS WG to continue with it. Glenn's listing of existing deterministic algorithms in this space should be enough to give us an expectation that it's feasible and technically possible.
>
> I’ve heard offline that people think that there are reasonable algorithms too. That doesn’t make it specifiiable yet, tho.

I don't follow: if it's been implemented and made replicable, it can
be specified.

Just because some algorithm can be specified doesn't mean it will be specified in a form that permits a normative reference, particularly in a reasonable amount of time. Are you willing to wait for 2 or 3 years to elapse to get an algorithm written into a CSS spec, get it tested, and move that spec to at least PR or REC before moving forward on VTT?

We can wait if there's progress (not with REC, but we don't need it for REC).

An alternative would be to pick up your algorithm from TTML for WebVTT as well. I don't mind which we do.

TTML doesn’t do balanced text wrapping, tts:wrapOption<https://www.w3.org/TR/ttml2/#style-attribute-wrapOption> allows “wrap” and “noWrap”. Profiles can optionally specify a requirement for the processor to use UAX14 for the line breaking algorithm, using the #lineBreak-uax14<https://www.w3.org/TR/ttml2/#feature-lineBreak-uax14> feature.

CSS Text Level 3 (WD) specifies<https://www.w3.org/TR/css-text/#line-break-details> that UAX14 must be honoured too. CSS2.1 does not reference UAX14.

Cheers,
Silvia.