Re: Change definition of URL to normatively reference IRI specification using a well-defined interface

On Apr 7, 2010, at 12:12 AM, Julian Reschke wrote:

> On 07.04.2010 00:45, Ian Hickson wrote:
>> On Tue, 6 Apr 2010, Julian Reschke wrote:
>>>
>>> But defining these two productions doesn't require defining  
>>> parsing; you
>>> simply build them from the more basic elements.
>>
>> It would be most convenient if the IRI spec could define this.  
>> Naturally,
>> if it does not then the HTML spec will have to work around this
>> limitation. However, since it is far easier to define this kind of  
>> stuff
>> in one place rather than spread it around, I don't see why we would  
>> want
>> the IRI spec to not define it. Productive cooperation comes from  
>> working
>> groups working together to achieve the common goal of improving the  
>> Web.
>
> Motherhood and apple pie :-)
>
> On the other hand, interfaces shouldn't contain more than it is  
> needed. "hostport" and "host-specific" appear entirely redundant to  
> me.

If we can get agreement to provide an interface roughly similar to  
what Ian proposed, then I'm sure there's room to negotiate the  
details. I'm really interested in hearing whether this general  
approach is acceptable, before we deep-dive on specific details.  
What's your position? Right now, I don't know what you think of any of  
this other than that you question some of the proposed outputs of the  
"parsing an address" algorithm.

Regards,
Maciej

Received on Wednesday, 7 April 2010 07:24:05 UTC