- From: Felix Sasaki <fsasaki@w3.org>
- Date: Wed, 17 Apr 2013 15:32:44 +0200
- To: Jirka Kosek <jirka@kosek.cz>
- CC: Yves Savourel <ysavourel@enlaso.com>, public-multilingualweb-lt@w3.org
Am 17.04.13 11:46, schrieb Jirka Kosek: > On 17.4.2013 11:10, Felix Sasaki wrote: > >> It has three drawbacks: >> >> 1) Global rules are different than "real" defaults in terms of precedence. > We have special section for HTML precedence (6.4 - > http://www.w3.org/TR/2013/WD-its20-20130411/#html5-selection-precedence), so > we can modify it if necessary. But with this "special purpose" precedence, things can get really unclear: - how to handle HTML embedded in other content formats - how to handle HTML not labelled as HTML5 - how to handle legacy HTML converted to HTML5 later, that is what expectations should (ITS local / global rules) authors have? That is my main concern. Best, Felix > > Also we can create two rule files one providing defaults, second > providing extraction of local ITS values. Then we can say something like: > > 1. Implicit local selection in documents (ITS local attributes on a > specific element). *Local selection is defined by the following rules...* > > > 2. Global selections in documents (using mechanism of external global > rules or inline global rules), to be processed in a document order, see > Section 5.2.1: Global, Rule-based Selection for details. > > 3. Selection via inherited values. *Inheritance rules are by provided by > the following rules...* > > 4. Selections via defaults for data categories. *Defaults are provided > by the following rules...* > > Then implementation is very easy -- it will just put global user-defined > rules in between rules for 1. and 3. and will not interpret local ITS > datacategories inside HTML. > >> 2) If we go this way we won't be able to speficy "how to work with >> translate in HTML5" for "local only" implementations. > We can say that "local only" implementations MUST behave as if the > following rules will be used. Implementation doesn't have to support > global approach, it just have to behave in the same way as is described > in default rules. > >> 3) we will have a disalignment with the behaviour of HTML5 "translate" >> implementations in browsers. > Why, we can model rules in a way it matches this. Agree with this one. > > Jirka >
Received on Wednesday, 17 April 2013 13:33:21 UTC