On 11/08/2010 10:59, Mark Howe wrote:
"Unlikely" in this case is in the order of "maybe once, on any one
planet, at any point in the lifetime of the universe". As one of our
team pointed out when we first considered this solution ourselves, if
that level of probability is an issue, you need to be worrying about
cosmic rays and quantum-level events in your XSLT code.
A little tangential, but when I was in ICL I remember some meetings with
hardware engineers who were implementing an appliance that did
relational joins. As software people, we found it very hard to stomach
the idea that the algorithm they were using had a finite probability of
producing wrong answers (due to accidental hash conflicts). But the
hardware guys were quite unfazed by this: they knew that all their
designs had a finite probability of producing wrong answers, and they
were accustomed to calculating the probabilities and deciding whether
they were within acceptable tolerances.
Michael Kay
Saxonica
--~------------------------------------------------------------------
XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list
To unsubscribe, go to: http://lists.mulberrytech.com/xsl-list/
or e-mail: <mailto:xsl-list-unsubscribe(_at_)lists(_dot_)mulberrytech(_dot_)com>
--~--