Re: HTML5 proposes introduction of new family of URI schemes

On 2012-01-23 23:13, Robin Berjon wrote:
> On Jan 19, 2012, at 19:13 , Julian Reschke wrote:
>> Basically, the spec reserves special semantics for infinite number of new schemes, but doesn't want to deal with the maintenance.
>>
>> Instead (stealing an idea mentioned earlier), it could simply define a new URI scheme "web", and manage that namespace itself. At least that would contain the effects of this experiment.
>
> No, that's just punting. Whether it's web+foo:blah or web:foo:blah really doesn't change the fact that a security decision is being made at minting time. The exact same problems are likely to crop up in both cases.

Yes, but the second case doesn't overload the scheme name.

Best regards, Julian

Received on Monday, 23 January 2012 22:59:37 UTC