https://bugs.kde.org/show_bug.cgi?id=446566
--- Comment #15 from Fred <d...@email.com> --- (In reply to Jack from comment #14) > *(&^&^( I wrote a whole reply, but got a "mid-air collision" and closed too > quickly before telling it to post anyway. So.... > I have been unable to replicate the issue, either with a version compiled > from git head of the 5.1 branch, or with the ArtixLinux 5.1.2 version. > > First, I hope you meant "thread" instead of "threat." :-) > > Second, can you check/confirm all the detailed steps you took? After > hitting the X to remove the tag, you need to click the check to accept the > changes to that split. Then you need to click OK to accept the changes to > the transaction. Just hitting "Enter" will Cancel instead of OK. > If you actually did all that, and the tag is not removed, then can you > create a small kmy file that shows the problem? Since you demonstrate the > problem with two different versions, I'd like to see if the problem might be > in the data, although I can't think of how that could work. Specifically, > can you create new transactions that demonstrate the problem? Does it only > happen with multiple splits? (So far, I've only tried with two splits.) Jack, finally the mystery is resolved. Previously I DID NOT hit the 'confirm' action button (√) and went straight to 'OK'. Somehow I ignored this step. Having said that, IMHO, this 't h r e a d' can now be closed. Thanks for your patience. PS: When the a user made changes in the item, perhaps these action button could be more eye-catching, e.g. through a coloured border around them. This way, stupid users, like me, would see better what their have to do ;-) (nice improvement) -- You are receiving this mail because: You are the assignee for the bug.