Re: [xsl] Random UUID in pure XSLT?

["Christophe Marchand cmarchand@xxxxxxxxxx" <xsl-list-service@xxxxxxxxxxxxxxxxxxxxxx>]

In my opinion, all those functions that depend on external things, as 
existence or non-existence of a file, content return by an http request 
can not be specified as deterministic, as implementors are not 
responsible of the returned value.

If specification defines these function as deterministic, as an 
implementor point of view, I should cache the result calculated by the 
first call, and always return the same content, without querying again 
the underlying resource ; this is not convenient, especially if the 
resource is a very huge resource.

I think these functions should not be deterministic.

Best regards,
Christophe

Le 12/11/2020 C  11:33, Norm Tovey-Walsh ndw@xxxxxxxxxx a C)critB :
> Dimitre Novatchev dnovatchev@xxxxxxxxx <xsl-list-service@xxxxxxxxxxxxxxxxxxxxxx> writes:
> > It seems to me that tagging the function unparsed-text() as "deterministic"
> > was rather inaccurate... And definitely not useful, if we need hacks as
> > workarounds for its limitations...
> Do you think fn:unparsed-text() should be different with respect to
> determinism than fn:doc() (and fn:document(), come to that)? Or do you
> think they should also be nondeterministic?
>
> (I donbt feel strongly one way or the other, Ibm just wondering about
> the distinction.)
>
>                                          Be seeing you,
>                                            norm
>
> --
> Norman Tovey-Walsh <ndw@xxxxxxxxxx>
> https://nwalsh.com/
>
> > If you don't have the time to do it right, where are you going to find
> > the time to do it over?