Re: how to estimate speed of a transformation

Hi,

> > Furthermore store the results of calculations in variables if you need
them more often.
>
> Is this advice true? If I need the same 'select' ten (or ten thousand)
times in a program,
> isn't it just simpler to write that select and let the implementation
memoize it?

Yes it would be simpler, but what if the application isn't.

> Its value will not change since the data is the same. Why should I
manually
> 'cache' values if the data model allows to do it consistently and
automatically?
>

I agree if applications would cache it automatically. I wouldn't have to do
it by hand, but I'm don't write parser code and I have no idea how easy or
difficult such a thing would be. However I do know, from other threads
started by you actually, that many possible performance optimizations are
not implemented by all parsers. Such is true, if I understand it correctly
for tail recursion which is optimized by Saxon and jd.xslt but not by other
parsers.

Cheers,
Agnes



 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list