Re: [xri] GCS Spoofing

[William Tan <wil@dready.org>]

Hi Mike,
> I've been following the discussion and find it very interesting.  I'm
> starting to wonder if we are delving into areas that need to be fixed
> generally instead of just in the context of XRI. 
These are areas that need to addressed, but it is generally agreed that 
there is no silver bullet, let alone an easy "fix".

>  I'd also like to know
> what sort of effort the folks involved expect this sort of analysis to
> take.  This topic is sounding very time-consuming to me.  
>
> So I guess I'm asking if (1) XRI is the right place to do this level of
> analysis and specification and (2) if we decide XRI is the right place,
> how long might this take?
>   
For IRI, the only real problematic character is the '/' (well, ':' and 
'@' may be used to spoof IRI authority as well, but is less serious). 
XRI, however, reserves several GCS characters and some delimiters as 
syntax characters having special meanings. So, yes, we do have to 
address this at XRI-level. But the question is how it is best handled.

I am dodging the "how long" questions because unless we have a clear 
proposal, we can't estimate effort required.

wil.