To comment on the following update, log in, then open the issue: http://www.openoffice.org/issues/show_bug.cgi?id=74148
------- Additional comments from [EMAIL PROTECTED] Mon Feb 5 07:14:39 +0000 2007 ------- The error also occurs when you use user-defined percentile values, at least when the percentile value is formatted to have NO decimal points at all. Nopne of the percentile columns in the example file have any decimals. And as described earlier, it isn't limited to percentiles and numbers only. Numbers of 0 and 1 turn to FALSE and TRUE if the preceding format is Boolean, which points to this being a general issue, even if unusual and limited to highly specific cases. I am attaching a new file to this where some of the behavior is showing. Explanation follows. In Test4.ods, the original columns are C (Boolean), D (unformatted), K (percent, user-defined to no decimals) and L (unformatted). Column C had original values of TRUE and FALSE added and column D had 1 and 0. Column K had 100%, 0% and 50% entered and column L values of 1, 0 and 0.5. The file was saved, closed and reopened and the bug manifested with the visible results for columns E and L. Then, values 1 and 0 were added to column C and values 1, 0 and 0.5 to column J, the file was saved, closed and reopened, leaving the final results as they are now. As we can see, adding new values to the left will corrupt original percentile data while leaving numbers corrupted by percentile format to their corrupted form. Adding numbers to the left of a Boolean column will not corrupt the Boolean data, but numbers corrupted by Boolean format will stay corrupted. Try following these steps yourself and see if you can reproduce the results. I also note that this bug does not affect all data types. I did a quick and dirty test with a new file and ran through all of the predefined variable types (date, currency, time, scientific, fraction) and it seems that data corruption occurs ONLY when percentile and Boolean formats are involved. None of the other data types have any effect on the following column. Further, it also seems that percentile format is able to corrupt e.g. currency format but only partially. Currency values between 0 and 1 will be corrupted to percentile values but values above 1 (tested with value of 100) stay uncorrupted. That could lead to some really nasty results if this happens in a production environment and especially if it happens across the board. The columns F (percentile), G (currency), H (unformatted) in Test4.ods are a rather graphic demonstration of the last case. --------------------------------------------------------------------- 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]
