https://bugs.freedesktop.org/show_bug.cgi?id=86216
--- Comment #6 from m_j_mal...@hotmail.com --- Here is a second, even simpler spreadsheet that 4.2.7.2 cannot sort properly. Ask it to sort by column B ascending with A1-B2 as the range. As for the process described for altering a spreadsheet to remove confidential data, I have examined it. The process would not remove the uniqueness of the numbers of the spreadsheet. I might recommend that LO add a 2 functions to anonymize data in a range. The first function goes through each cell in the defined range. It examines the contents and replaces it and all identical occurences of that data (either numerical or text) with a random equivalent everywhere in all tabs of the spreadsheet. For occurrences found within the defined range, they are marked as "done" and are not doubly randomized. Text becomes random text of similar length. Numbers become random numbers (never zero) of similar parameters. So $5.45 becomes $10.71, but never 6.024E24 for instance. The second function does not make global searches and replaces, the data in each cell is just replaced with random data of similar nature. Therefore two occurences of $5.45 become different amounts. One might then randomize a spreadsheet with little manual labour while maintaining the action of MATCH and SUMIF operations. If the linking of these functions were not preserved in the randomization of my spreadsheet, it would completely change the nature of sorting it. However, someone will no doubt point out there is some other archane system for making suggestions. Returning to the problem of 4.2.7.2 not sorting correctly, In this reference: https://bugs.launchpad.net/ubuntu/+source/libreoffice/+bug/1389858 Lukebenes reportedly supplied a spreadsheet that 4.2.7 could not sort properly (the same way 4.2.6 did) with either setting of the secret setting. Luke (lukebenes) wrote (in part) on 2014-11-11: ... To be clear, with 4.2.7 there is no way to return to the old style sorting routine. Some spreadsheets that worked fine in Excel and 4.26 require UpdateRefs=True, while others require it to be false. Others are broken with either UpdateRefs setting because of bugs introduced in this new sorting routine. Is it any wonder users are confused? I recommend taking his sample spreadsheets from that thread test_offset.ods and Calc bug.ods and verify those. This matter is so urgent, I hope someone can use a little imagination and use these sources in other bug reports as tests. -- You are receiving this mail because: You are the assignee for the bug.
_______________________________________________ Libreoffice-bugs mailing list Libreoffice-bugs@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/libreoffice-bugs
