https://bugs.documentfoundation.org/show_bug.cgi?id=171017
--- Comment #16 from Colin <[email protected]> --- (In reply to m_a_riosv from comment #14) > IPOV, the issue with the array formula: > {=IF($Z3=0; 0; INDEX($A$2:$I$999; $Z3; {3|4|5|6|7|8|9}))} > Is about INDEX not being calculate when Z3=0, so no bug there, IMO, because > only one '0' when the condition it's true. You can use: > =IF($Z3=0; {0|0|0|0|0|0|0}; INDEX($A$2:$I$999; $Z3; {3|4|5|6|7|8|9})) > > Aslo, can be use a formula like: > {=INDEX($A$2:$I$999; $Z3; {3|4|5|6|7|8|9})*($Z3>0)} > It works fine for me because INDEX it's always calculated, what can be a > performance issue. My "misunderstanding" is clearly that I assumed the undo simply reverted to the previous "snapshot" rather than "uncalculating" the current sheet, whereby the IF then seems to have a different "TRUE". Is it that the array undo is element by element so testing the residue in elements 2-7 doesn't find 0 - exit - but a real number - process it? Also, would you consider the report should be NOTABUG and I simply follow @ady 's advice and formulate a more precise report - or is it more likely that the plethora of UNDO bug reports renders my own experience redundant? -- You are receiving this mail because: You are the assignee for the bug.
