this bug has two parts really. one part is external to ruby, the other is 
internal to it. the external part is a problem with the kernel not conforming 
to ieee 754 floating point standards.... the register has a smaller significand 
than is expressed in the standard. It should have 16 significant digits that 
are at all times exactly correct. the 17th digit should be the earliest digit 
with rounding errors. Only digits after the 17th digit should have random 
(trash) bits:
8.4000000000000056 8434188608080148696899414062500000e+00
The zeros should fill all the way to the space I inserted.

When I inquired about the ruby part over at redmine.ruby-lang.org --
ruby1.9.1 has a recognized bug which is fixed in the current trunk
(1.9.3 trunk). building the current  trunk of ruby also fixes some
problems with File object mangling utf-8 paths, and should really be
adopted with all due haste over 1.9.1.

-- 
Luid: floating point operations seems not to match specification
https://bugs.launchpad.net/bugs/586087
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.

-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to