Posted by Neredbojias on 03/23/07 04:35

On Tue, 20 Mar 2007 11:16:23 GMT Andy Dingley scribed:

> On 20 Mar, 08:42, Neredbojias <inva...@example.com> wrote:
>
>> Eh? The "inventor" defines the "correctness" without adjustment of
>> progress or evolution?
>
> There is no standard for favicon, largely because M$oft won't play
> ball. The web is built by layering one simple standard on top of
> another. M$oft's half-baked implementation of favicon broke existing
> standards in at least 3 ways, so it's just not possible to integrate
> it into the grand scheme of things. (From memory, these are: insisting
> on a fixed URL location within a site for the icon document, using a
> magic name for the rel atribute that had forbidden spaces in it, and
> using an icon format that didn't have a defined MIME type (easily
> fixed)).
>
> Some smart puppy should go and draft an RFC for favicon with a clever
> backwards-compatible method for making it work transparently in both
> worlds. As that's beyond this bear's little brain, even a clear
> documentation of the existing position would be worthwhile. However
> trying to bring M$oft to heel is a thankless task and usually asking
> for your work to be dumped on by them in the next release.

A standard could surely be advantageous, though I may suggest that a "non-
sensible" standard is hardly better than no standard at all. Regarding
favicons, it seems to me that the way Mozilla (and Opera I believe)
impliment it is much superior to the way IE does it. Therefore, any
hypothetical standard based on the IE-method is hardly worth the effort. I
might propose a meta line change from 'link rel="icon"...' to 'link
rel="favicon"...', however.

--
Neredbojias
He who laughs last sounds like an idiot.

[Back to original message]