On Jul 8, 2008, at 12:19 PM, Wendell Piez wrote:
... depending on where you're going, you might have a concern about
tomorrow as well as today.
In my experience designing languages the engineerings that end up
using it do all kinds of thing the designers never anticipated.
I asked the question because I care about tomorrow.
It is also my experience that language designers (I know, I am one)
use "should" language way too much.
In fact, I was after rather more specific information; unrelated to
any opinion about what I "should" be doing. I essentially asked for
the optimal way to place result-document output in two locations. So
far, I have been offered no quantified reason for doing what I propose
that actually would impact scaling or robustness.
However, I can think of one or two possible issues. XSLT garbage
collection, for example, may not - for reasons I may be unaware of -
be unhappy. I'd expect it to be fine, but it seems reasonable to ask.
There may be error handling issues that people have encountered
processing a number of collections. Again, it seems reasonable to ask.
In general, all my XSLT is server side and is not generating "dynamic
pages" so I do not expect to have either scaling or robustness issues
later. And while I do process file sets of some size, each set is
constrained to a logical document. Imagine something that dealt
Michael Kay's new book (which I am finding very helpful BTW - and
would love to know just which tools were used to produce it).
With respect,
Steven
--
Dr. Steven Ericsson-Zenith
Institute for Advanced Science & Engineering
http://iase.info
http://senses.info