- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Tue, 4 Mar 2014 15:45:17 -0800
- To: John Daggett <jdaggett@mozilla.com>
- Cc: www-style list <www-style@w3.org>
On Fri, Feb 28, 2014 at 12:20 AM, John Daggett <jdaggett@mozilla.com> wrote: > Sort of along the line of Jonas's question, it's unclear to me precisely how/when > newly created FontFace objects are added to the appropriate > FontFaceSet (e.g. document.fonts). Is the intent that they are > automatically added at construction? If so, then the spec needs to > state that clearly. If not, then it should state that an explicit add > is required. Construction doesn't mention anything about automatically adding it to the document's font source, so it doesn't do so. I can add a note if you think it would be useful as a reminder. > The spec *really* needs references to the Promises API. How about a > normative reference to Dominec's spec [1] and an informative reference > to the HTML5Rocks article on Promises [2] for now? Something is > better than nothing. I realize the spec is still in flux within the > ES6 group. I was waiting for things to settle down, but I can go ahead and add those links for now. > It seems odd that FontFace.load() has a return type of void but > FontFaceSet.load returns a Promise. Wouldn't it be more consistent to > have them both return a promise? Yeah, other people have cited this as something weird. It's a path-dependent thing; I originally didn't have the "status" attribute, so .ready() was the only way to observe the load status. I've removed .ready() from FontFace and just made .load() return the [FontStatusPromise]. ~TJ
Received on Tuesday, 4 March 2014 23:46:05 UTC