- From: Glenn Adams <glenn@skynav.com>
- Date: Sun, 11 Jan 2015 07:57:38 -0700
- To: Koji Ishii <kojiishi@gmail.com>
- Cc: Xidorn Quan <quanxunzhen@gmail.com>, W3C Style <www-style@w3.org>, CJK discussion <public-i18n-cjk@w3.org>
- Message-ID: <CACQ=j+fxJwjT=QpSsanwHNRasjXUh-kfo8hqe_niaxMJcFTXxQ@mail.gmail.com>
On Sat, Jan 10, 2015 at 11:40 PM, Koji Ishii <kojiishi@gmail.com> wrote: > I put this onto the wiki[1]. > > Naming of the value looks fine, and I support the use case. However, > it looks like we don't receive interests to implement in UA at this > point, and I18N WG expressed they wish to keep the Level 1 as minimum > as possible. > > As long as we don't see strong objections to the naming, and as long > as TTML/WebVTT uses some translation rules to render in UA, As Xidorn > pointed out, they should be ok to translate the 'outside' value to a > CSS selector. > > Glenn, one more thing you may want to checkout is whether > text-emphasis-position has the same issue or not. I suspect it does, I > remember seeing such in movies, but your people must know it better. > yes it does; i think i mentioned this earlier in the thread > [1] https://wiki.csswg.org/spec/css-ruby-2 > > /koji > > On Wed, Dec 17, 2014 at 2:20 AM, Glenn Adams <glenn@skynav.com> wrote: > > > > On Tue, Dec 16, 2014 at 9:01 AM, Koji Ishii <kojiishi@gmail.com> wrote: > >> > >> So TTML doesn't use UAs to render? > > > > > > Both TTML and WebVTT assume that some entity, e.g., server, client side > > scripting, or embedded UA functionality, will either translate to an > > available directly supported renderable form, e.g., HTML+CSS, SVG+CSS, > etc., > > or will render directly. Consequently, there is generally an attempt > made to > > choose an expression that can be readily translated into an > {HTML/SVG}+CSS > > equivalent that preserves the intended presentation semantics. > > > >> > >> JLREQ clearly states that its > >> coverage is limited to regular books such as novels, so I'm not > >> surprised it's not in there, but no motivations/use cases in Web nor > >> UA engines concerns me a bit to define it. > > > > > > That is a good point to make, because many folks perceive JLREQ as > > articulating a wider range of requirements. It is a good reminder that it > > does not, and that we should not *solely* rely on JLREQ to define all > > requirements, but need to look elsewhere as well. > > > > The semantics I am discussing here is outside of the explicit JLREQ > scope, > > but remains a real world requirement. So the group will need to weigh > how to > > proceed. > > > >> > >> > >> Maybe we should create a wiki page for CSS Ruby Level 2 and put it > >> there, unless Xidorn or any implementers wish to implement. Does that > >> work? > >> > >> /koji > >> > >> On Tue, Dec 16, 2014 at 12:02 AM, Glenn Adams <glenn@skynav.com> wrote: > >> > > >> > On Mon, Dec 15, 2014 at 7:51 AM, Koji Ishii <kojiishi@gmail.com> > wrote: > >> >> > >> >> As I see it in almost every movie I watch, and in every Karaoke too, > >> >> I've been wondering why there were no requests so far. > >> >> > >> >> Naming wise, 'outside' looks a good candidate to me. Other candidates > >> >> we tried for the text-align property was 'first-last' or 'first > last'. > >> >> It was removed for other reasons but there were a few concerns on the > >> >> naming during the review, so they might not be good candidates. > >> >> > >> >> Are you just asking naming ideas that fits well to CSS, > >> > > >> > > >> > If someone has a better name than 'outside', I'm sure we could use it > >> > with > >> > TTML. > >> > > >> >> > >> >> or also asking > >> >> as a possible addition of this value to CSS Ruby? > >> > > >> > > >> > Since this is a real requirement that isn't discussed by JLREQ and > isn't > >> > addressed by the current css-ruby draft, it seems natural that such a > >> > value > >> > should be added. It may also help when TTML is translated to HTML/CSS > >> > for > >> > rendering. > >> > > >> >> > >> >> > >> >> /koji > >> >> > >> >> On Mon, Dec 15, 2014 at 7:17 AM, Glenn Adams <glenn@skynav.com> > wrote: > >> >> > > >> >> > > >> >> > On Sun, Dec 14, 2014 at 2:54 PM, Xidorn Quan < > quanxunzhen@gmail.com> > >> >> > wrote: > >> >> >> > >> >> >> On Mon, Dec 15, 2014 at 2:05 AM, Glenn Adams <glenn@skynav.com> > >> >> >> wrote: > >> >> >>> > >> >> >>> Recent work on supporting deployed Japanese Subtitling/Captions > in > >> >> >>> TTML > >> >> >>> indicates a requirement to support ruby positioning on 2 line > >> >> >>> subtitles/captions where the first line uses right/above (before) > >> >> >>> and > >> >> >>> the > >> >> >>> second line uses left/below (after). We have addressed this in > TTML > >> >> >>> by > >> >> >>> introducing an 'outside' keyword, which is interpreted as > 'before' > >> >> >>> for > >> >> >>> lines > >> >> >>> 1 through N-1 and 'after' for the Nth (last) line. > >> >> >> > >> >> >> > >> >> >> I guess it would make more sense to combine ::first-line with > >> >> >> ruby-position than to define a new value. It would have different > >> >> >> behavior > >> >> >> from what you defined here when there are more than two lines, > but I > >> >> >> wonder > >> >> >> if either way gives an ideal result in that case. > >> >> > > >> >> > > >> >> > TTML doesn't have selectors, so that option isn't available there. > >> >> > Even > >> >> > if > >> >> > it were available, it doesn't seem quite proper to distribute the > >> >> > semantics > >> >> > of positioning in that fashion, i.e., to rely upon first line > >> >> > overrides. > >> >> > But, yes, that remains an option with CSS. > >> >> > > >> >> >> > >> >> >> > >> >> >> > >> >> >> - Xidorn >
Received on Sunday, 11 January 2015 14:58:27 UTC