Mike Cramer wrote: > > CaT wrote: > > > Why should he have to? I thought the point to Mozilla (or at least one > > of them) was to NOT make a broken client. A client that requests > > something other then what is on a webpage is broken. > > If you are talking about an HTTP client, I don't think that is true. I > don't think you'll find anything in the HTTP spec that says a client can > only requests objects referenced on an HTML page. In fact, I don't think > you'll find much about HTML in any HTTP client spec. Requesting > favicon.ico may be somewhat impolite, but it is definitly not an example > of "broken" behavior. > > It's amazing how much anger this feature is generating. Particularly > since it amounts to giving content providers *MORE* control over the > presentation of their content than HTML allows directly without breaking > HTML, HTTP, or anything else.
Wrong. Getting *MORE* control with <link rel="icon"/> doesn't break HTML, HTTP, or anything else. /Jonas
