> On Jun 15, 2017, at 7:27 PM, Michael Catanzaro <mcatanz...@igalia.com> wrote:
> 
> One issue we've noticed is there are many different nonstandard ways that 
> websites use to find icons. [1] details many of these: the Microsoft way, the 
> Apple way, the OpenGraph way... and even then the icons can be put in 
> different places... is this the problem Brady's code is designed to solve?

The WebCore::IconDatabase was extremely limited.

It would look at only the last <link rel="icon"> or <link rel="shortcut icon"> 
(which are standardized in HTML5) and consider it "the one icon" for a page.
If the <head> specified none, it would consider the default /favicon.ico the 
one icon.
And all it could load/store is that one icon.

The new mechanism already expands upon this by allowing for *each* <link> icon 
to be considered individually.
It also expands upon this by having built-in support for apple-touch and 
apple-touch-precomposed icons.

There's nothing stopping us from adding support for all of these icon types 
with only small changes to WebCore. Of the list in that blog post some might be 
considered relevant to Apple, others probably not - But LinkIconCollector is a 
tidy self-contained class that can easily have per-platform implementations.

tl;dr - Yes, this is basically the problem the new mechanism solves.

Thanks,
 Brady
_______________________________________________
webkit-dev mailing list
webkit-dev@lists.webkit.org
https://lists.webkit.org/mailman/listinfo/webkit-dev

Reply via email to