Re: [xsl] single precision bug

[David Carlisle <davidc@xxxxxxxxx>]

> What do you all think of this?

It's the expected behaviour with any numerical computation, you should
never compare for 0 but rather check that the absolute value is less
than some small number, or use format-number to round to a fixed number
pf places before comparing.

David

________________________________________________________________________
This e-mail has been scanned for all viruses by Star Internet. The
service is powered by MessageLabs. For more information on a proactive
anti-virus service working around the clock, around the globe, visit:
http://www.star.net.uk
________________________________________________________________________