https://bugs.documentfoundation.org/show_bug.cgi?id=163441
--- Comment #3 from ady <adylo811...@gmail.com> --- Using attachment 197056, and following the steps (starting from step 08): (In reply to Martin from comment #0) > 08. Set E104 to =E102/E103 and copy across to column G > > --- test 1 - cells dont' recalculate > > 09. All of E104:G104 should be 100 > 10. Hit F9: All cells update, E102:G103 DO NOT recalculate. > > --- test 2 - clicking into an affected cell fixes the recalculate > > 11. Click any cell in E102:G103 > 12. Hit F9: The clicked cell updates, the rest of E102:G103 DO NOT > 13. The check in row 104 below the clicked cell is NOT 100 > > --- test 3 - clicking out of the affected cell breaks the recalculate > > 14. Click elsewhere, eg E104 > 15. Hit F9: All cells update, E102:G103 DO NOT recalculate. > > --- test 4 - deleting rows fixes the recalculate > > 16. Delete 5-10 rows from the top of the sheet. > 17. Hit F9. All cells update and E102:G103 DO recalculate. > 18. All of E104:G104 should be 95 > > --- test 5 - replacing the rows breaks the recalculate > > 19. Press Ctrl-Z > 20. Hit F9. All cells update and E102:G103 DO recalculate, but this only > happens once. > 21. All of E104:G104 should be 100 > 22. Hit F9: All cells update, E102:G103 DO NOT recalculate. I do not reproduce the behavior described in comment 0 (and quoted here) when using: Version: 25.2.0.0.alpha0+ (X86_64) / LibreOffice Community Build ID: c3c29d31d77ff93aa50634cfd51c62d12dc0f6ec CPU threads: 4; OS: Windows 10 X86_64 (10.0 build 19045); UI render: default; VCL: win Locale: en-US (es_AR); UI: en-US Calc: threaded In my test, I have not used OpenCL, no hardware acceleration, and no Skia. There is a chance that changing some combination of those settings, the problem would show up, but I have not tried them. Maybe the OP might want to test such settings; please report back. As for "test 6", I have not tested it. > > --- test 6 - ctrl-Y, ctrl-Z fixes/breaks the recalculate > > 23. Hit Ctrl-Y, F9 works as expected. Hit Ctrl-Y, F9 works once then it > doesn't > > --- end of tests -- You are receiving this mail because: You are the assignee for the bug.