RE: uuid question

> Hi all,
> 
> Does p:uuid generate *one* UUID, or separate UUIDs for each of the
> matching nodes? I think it is the former, but it is not 100% clear in
> the spec.
> 

Also, I think we are missing something in the declaration of p:uuid. If
you want to generate level 3/5 UUIDs (MD5/SHA-1 hash based), you need to
be able to pass an input value for the UUID algoritm. Perhaps we could
do the same thing as with p:hash and add the "value" option to p:uuid
(...or we could simply say that level 3/5 UUIDs are not supported).

-

I also have one general (and heretical?) question about p:uuid and
p:hash: Why do we have these steps at all? They are just a special form
of p:string-replace, ...so what if we had p:uuid() and p:hash() as XPath
extension functions? I know it would mean to define a nice APIs for
these functions, whether and how they accept additional parameters etc.,
but in the end, I think we could gain much more flexibility. But it is
probably too late for considering such a big change, plus there may be
other serious arguments against this which I don't see... But I am just
curious about your opinion.

Regards,
Vojtech

Received on Friday, 5 December 2008 12:27:17 UTC