Re: Widget URI ED

On 29 Apr 2009, at 17:32, Robin Berjon wrote:

> I've made a first pass at an ED for widget URIs (which is needed, as  
> you can tell from the I18N discussion). It's in need of tightening  
> up editorially wise but the basic ideas are there:
>
>  http://dev.w3.org/2006/waf/widgets-uri/


Thanks for getting this started.

1. Please clearly distinguish between URIs, URLs, and URI references.   
What you mostly mean are, e.g., relative URI references.

2. What's the use case for including the UUID in the base URI?  And  
what is the expectation how an implementation will deal with an  
absolute URI reference, i.e., that includes a UUID *different* from  
the current one?  (There might be a really nice cross-instance  
communication feature in here.  It's not clear it's quite there.)

3. I think the construction of the origin is largely sane, but, again,  
would like to go into relevant use cases in more detail before saying  
that it's ok.  Also, I'd like to see  some clarity around the notion  
of "instance" -- e.g., are you assuming that several instances of a  
widget can run within the same environment at the same time, but with  
several UUIDs?

> Share and Enjoy!

... and I thought that levrogyrous dahut's in the French Alps were  
Somebody Else's Problem (tm).

Received on Wednesday, 29 April 2009 16:02:04 UTC