Re: [xsl] Streaming best practice: no user-defined functions with node parameters ... Do you agree?

[Wendell Piez <wapiez@xxxxxxxxxxxxxxx>]

On Fri, Nov 15, 2013 at 5:56 AM, Wolfgang Laun <wolfgang.laun@xxxxxxxxx> wrote:
> On 15/11/2013, Costello, Roger L. <costello@xxxxxxxxx> wrote:
>
>> In fact, it might be a while before XSLT processors can handle any
>> user-defined functions, so it might be a good idea to completely avoid
>> user-defined functions in your XSLT programs.
>
> But I've been using user-defined functions for quite some time...
> -W

Indeed. Roger overstates the case, perhaps forgetting that stylesheet
functions are a very useful feature in XSLT 2.0 and are implemented in
any conformant XSLT 2.0 processor. Unless we mistake his sense.

Yet I agree that the discussion of functions and streamability has
been useful, and the current proposal to restrict streamability to
functions accepting only atomic arguments appears clean and nice (in
the old-fashioned sense). But does it mean we won't be able (when
streaming) to call functions even with temporary trees or snapshots?
Not that I know how to draw that line....

Cheers, Wendell