Re: [xslt] floor(), ceiling(), and round() functions and negativevalues

[Sander Vesik <sander_traveling yahoo co uk>]

John Fleck wrote:
> On Sat, 2003-03-15 at 08:13, Daniel Veillard wrote:
> 
>>On Sat, Mar 15, 2003 at 08:38:09AM -0600, Ardo van Rangelrooij wrote:
> 
> [snip]
> 
>>[...]
>>
>>>Are these indeed bugs?  If so, I'll file an official bug report.
>>
>>  libxslt doesn't make a difference between -0 and 0, right this may
>>be considered a bug ... The associated questions are:
>>    - does anybody care ?
>>    - does anybody knows how to code the detection in a way
>>      portable across the set of architectures libxslt support ?
> 
> 
> Before we use up precious bandwidth on the arcane question of zero and
> negative zero, no less an authority than my daughter, Nora, points out
> that the notion of negative zero is silly. However, the folks who wrote
> the XPATH specification did not have the wisdom to consult Nora, so the
> specification includes negative zero.
> 

Xpath is hardly the only negative zero case, there is a negative zero in 
most signed magnitud number systems, of which ieee fp is the presently most 
widespread one (another is bcd, yet others are i believe 1-complement 
machines). Its mostly an artefact though there are occurences when it is 
not totaly useless.

> Cheers,
> John



__________________________________________________
Do You Yahoo!?
Everything you'll ever need on one web page
from News and Sport to Email and Music Charts
http://uk.my.yahoo.com