On Don, 20 Feb 2003, David Hyatt wrote:

> I'd be curious to see a list of sites that block Gecko.  Not one of the 
> top 1500 sites on the Web blocks Gecko as far as I know, and if you're 
> talking about banking sites, well, they'll block us for being Safari 
> (and are well-aware of any UA games we try to play).

Nope. not Mozilla, but Netscape 6.x. Both might identify themselves as 
Gecko, but Netscape 6.x was really broken. 

> I much prefer claiming to match Gecko by default and then automatically 
> spoofing based off an exception list for those sites that only work 
> with WinIE.

I tend to agree. Do you plan on defaulting to a list of wellknown sites that 
only work with WinIE or do you let each user individually figure out?

> Anyway, I think the key thing to take away from this discussion is that 
> we should include something in our UA strings that allows someone to 
> easily detect a KHTML-based browser.  The rest of our strings (and our 
> philosophies regarding default emulation of WinIE vs. Mozilla) may 
> differ, but we should make it so there's an easy check that recognizes 
> both Konq and Safari.

I disagree. If we don't agree on the same default, we end up bloating the 
code with every quirk we find in the default. We can hardly render both MSIE 
and Mozilla compatible and be faster than both browsers. 

Anyway, I don't disagree anymore switching to a Mozilla like UA string. The 
key thing in this discussion for me was that the general layout of the UA 
string of our two browsers should be the same. If you absolutely need the 
double bracketed string parts, then please tell. Otherwise IMHO it would be 
better to just include it in the first bracketed string part, at the same 
position in both browsers, having the same field layout. Look, changing the 
UA string is not so difficult for you, as you don't have a stable release 
yet. We have about 5 major public stable releases which we break 
compatibility with. 


-- 
Dirk (received 331 mails today)