Re: Generic-Resources-53: URIs for representations

On Oct 9, 2006, at 09:46, ext Dan Brickley wrote:

> Patrick Stickler wrote:
>
>> While one cannot conclude that a given URI denotes a  
>> Representation because
>> it seems to always return the same octet stream, if any URI ever  
>> returns
>> different octet streams then one should be licensed to conclude  
>> that that
>> URI does *not* denote a Representation.
>
> *ever* is rather strong there, don't you think? eg. how would you  
> model
> a hacked server that was compromised for a week, returning unusual
> content instead of its typical behaviour?
>
> Forever is a long time...

I think that when speaking about architectural issues, malfunctions can
be kept out of scope, with the clear understanding that if/when things
break, strange thigns happen -- though strange happenings would  
certainly
be an indication that something is broken.

Perhaps the above could be modified to "if any URI ever returns
different octet streams then one should be licensed to conclude that  
that
URI does *not* denote a Representation, or that the server is
malfunctioning in some manner.

>
> The only way I can see around this is, if you want to preserve the
> claim, is by appeal to some notion of a legitimate or successful
> derferencing, ... but that has its own problems.
>
> Dan
>
> ps. I like the clock example. Simple but with nice complexity (eg.
> language negotiated content too).

Agreed.

Cheers,

Patrick

Received on Monday, 9 October 2006 14:46:07 UTC