To comment on the following update, log in, then open the issue: http://www.openoffice.org/issues/show_bug.cgi?id=91358
------- Additional comments from [EMAIL PROTECTED] Fri Jul 4 18:40:47 +0000 2008 ------- Just to confirm this in R: > x <- 29513736*92842033 > print(x, digits=22) [1] 2740115251665288 While Excel and Calc do get it wrong, the MS Windows Calculator gets it right, too. [Tested on the Calculator that ships with Win2k ;) ] So, I doubt that 15 digits accuracy and precision can be kept any longer in 2008. It might have been great in the '70s, it might have been right in the '80s, but it well overdue in 2008. For other interesting readings, I point to Kahan's website: http://www.cs.berkeley.edu/~wkahan/ There are a lot of interesting articles, I especially recommend the 56 pages long: How Futile are Mindless Assessments of Roundoff in Floating-Point Computation? http://www.cs.berkeley.edu/~wkahan/Mindless.pdf While not directly relating to this issue, it does cover Excel's bad behaviour and a lot of Floating Point mistakes. The introduction to: How Java’s Floating-Point Hurts Everyone Everywhere, is also excellent. Especially the section devoted to floating-point exceptions. http://www.cs.berkeley.edu/~wkahan/JAVAhurt.pdf Hope this will help to make Calc a better mathematical tool, and one trustworthy. At least, Calc should sense such overflow-errors and report them, e.g. by displaying a red marker besides the number, and when the user clicks on this marker, display: *Result inaccurate due to overflow.* --------------------------------------------------------------------- Please do not reply to this automatically generated notification from Issue Tracker. Please log onto the website and enter your comments. http://qa.openoffice.org/issue_handling/project_issues.html#notification --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
