Re: qvalue *, was: Issue 113

Frank Ellermann wrote:
> Julian Reschke wrote:
>  
>>> The problem of "*" or "*/*" or other Accept-* stars,
>>> where the implicit or explicit <qvalue> is not the
>>> smallest <qvalue> in the Accept-* list.
>  
>> It's only a (potential) problem if clients send it.
>> Do you have any evidence this happens in practice?
> 
> I've no evidence that it does NOT happen in practice.
> 
> I know for sure that a "quick locale switcher" add-on
> of my UA managed to add some weird chrome://whatever
> string to Accept-Language until I fixed it manually.
> 
> <about:config> claims that my current Accept-Charset
> might be iso-8859-1,*,utf-8 (sic!).

Which UA?

My FF3 says:

Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7

...so I don't see any problem here.

> And <http://delorie.com:81/> tells me that it is in
> fact Accept-Charset: windows-1252,utf-8;q=0.7,*;q=0.7

For me it says:

Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7

which matches what LiveHeaders reported.

> So yes, this obviously happens, PEBKAC or otherwise.

PEBKAC?

>> The case where "*" has qvalue bigger than an other
>> value in the list is an edge case. I don't see why
>> a client ever would send it.
> 
> I also don't see why my UA should ever wish to send
> garbage, nevertheless it does.  And RFC 2616 doesn't
> even say that this is garbage, so arguably it is not.

I'm not convinced yet that your UA *does* send garbage. <about:config> 
apparently leaves the qvalues out, so you really need to look at the 
HTTP traffic.

BR, Julian

Received on Monday, 4 August 2008 09:55:17 UTC