> 1) Favicon requests put minimal load on servers either way.
Kind of like a mime at a party, who goes up to every single guest,
stands in front of them, and starts performing. There's only minimal
interference with the guest - but sooner or later the mime's going to
get punched in the face and thrown out the door. <grin>
> 2) If publishers want an icon displayed, and they care about bandwidth,
> favicon.ico is actually BETTER.
Because it promotes not having to do any HTML coding and removes
the responsibility of the publisher? I can't see how this is good.
It's the site that should control content and behaviour, not the
client.
> of additional requests, but even those can be minimized with an empty
> file called "favicon.ico".
Which means that every site in the world would have to go to the
extra step of creating this file simply in order to alleviate the
demands of the client. This is not at all unlike parents having to
child-proof their house in order to accomodate their potentially
harmful curiosity. Are you saying that Mozilla is a misbehaving child
that needs to be protected from? Surely it's far simply to just
change Mozilla's behaviour.
> 4) The debate has nothing to do with web standards, because both
> favicon.ico and <link> tags are valid no matter how you slice them.
Since when has .ico ever been anything more than a Microsoft
proprietary "standard"?
> But then again, this debate was never about what is best for anything.
I surely think that <link> is better than favicon.ico.
> It's just a bitch session where the *REAL* targets are Internet Explorer
> and whoever enabled the favicon.ico feature in recent Mozilla builds.
Okay, that's true too. Yes, I *am* complaining about IE (although
I've always complained about IE and Microsoft so that's hardly news)
and Mozilla whose philosophy should not be letting it copy this bad
behaviour.
Jason.
