<html> <head> <base href="https://bugs.documentfoundation.org/"> </head> <body> <div> <a class="bz_bug_link bz_status_NEW " title="NEW - AutoCalculate silent, deceptive fail. Indicative pattern of one edit step lag in AC update noticed." href="https://bugs.documentfoundation.org/show_bug.cgi?id=129199#c11">Comment # 11</a> on <a class="bz_bug_link bz_status_NEW " title="NEW - AutoCalculate silent, deceptive fail. Indicative pattern of one edit step lag in AC update noticed." href="https://bugs.documentfoundation.org/show_bug.cgi?id=129199">bug 129199</a> from <a class="email" href="mailto:newbie-02@gmx.de" title="b. <newbie-02@gmx.de>"> b.</a> <pre>@R. Bingham for those short in time: i think it's a problem of a faulty order in processing the circular dependencies, you can interfere / override that by moving N20 in a row above 20, ctrl-X - ctrl-V, but the bug should be searched and killed anyway ... long version: solved? no :-( but hope i can pinpoint the source! (call me crazy ... ) sorry for long, who doesn't like shouldn't read, after some play around, testing older versions, trying to decrypt the circles and dependencies in your sheet, even installing AOO 4.17 ... i came to one very simple test: copy cell N20, not the cell but the text inside, 'the formula', paste that in e.g N18, in the same way copy P20 to P18, and change it's formula to referencing N18 instead of N20, repeat the tests and cases you studied before. on my system: N18 and P18 working correctly, N20 and P20 failing every second case. that leads to the idea that your bug might be a flavour of the well beloved (only by me?) old! 'autocalculate broken' bug, ... one could have seen that earlier from ctrl-shift-F9 working ... :-( ... new insights from this bug: - the error / bug evolves from a special constellation between dependencies in formulae, and position of cells in the sheet, see experiments below, - in this case the exclusion is not 'total', just 'lagging', - but it depends in any way on a special cell, it's description in the file? it's position in the sheet? former connections to other cells? - your dependencies and circles are quite complex, they need 2 iterations to solve, as arranged in your sheet and worked through by the program they need 3, (can be tested by setting down the iterations in the config till you get err:523 on new input) i see one more very special aspect for your file: the error occurs once two cells (who themselves are referenced elsewhere) reference the same cell (in this case N20 is referenced by M20 and P20) but only! - now it's getting funny - if the referenced cell is in the same row or below! going on to play ... take the file as is, cut out N20 - ctrl-X - place it somewhere above row 20 (references should update automatically), and you're done. what happened? you re-arranged the erroneous order of processing which some programmer implemented earlier, now the circles are solveable in two iterations, and the program works correct. try it, that is not!!! the solution of the bug in the program! - belive me there is one - and not!!! a suggestion not to fix it! imho it's very important to get things like this out, and to FIX it, really find the evil and correct that, NOT to plaster a 'patch' over this special new occurence ... reg. b.</pre> </div> <hr> You are receiving this mail because: <ul> <li>You are the assignee for the bug.</li> </ul> </body> </html>