[reSIProcate] representing qvalues as integers instead of floats

[Alan Hawrylyshen]

On Aug 11, 2005, at 11.15, Rohan Mahy wrote:

>
> I think it would be a good idea to migrate the q parameter to this  
> type instead of the float type.  Some folks have pointed out that  
> using a float for a q-value results in the gratuitous addition of  
> floating point code to platforms that may not otherwise "need" it.
>
>


Given that the ABNF defines q-values as:

qvalue = ( "0" [ "." 0*3DIGIT ] )
        / ( "1" [ "." 0*3("0") ] )

It's clear to me that they aren't really a float, but a value from  
0-1000 inclusive.  I think it's a good idea.

A