https://bugs.documentfoundation.org/show_bug.cgi?id=170865
--- Comment #20 from Telesto <[email protected]> --- (In reply to Laurent Balland from comment #13) > I am not in favor of removing this confirmation dialog because it contains > an important information: the number of sheets that will be removed. It is a > quite common scenario when there are several sheets selected, that you may > forget about it, and deleting active sheet will destroyed all selected > sheets. The dialog reminds you that several sheets are selected. Argh, sorry I overlooked that. So, the whole 'references' stuff (like comment 2) is bit of a sidetrack. There is no direct relation to the presence of references and the dialog A) The dialog will appear as long as there is content - (including stuff like a cell border, cell background color) - somewhere on sheet when deleting So the dialog is from this perspective an "Information dialog" as mentioned in comment 16. *) I makes clear some content being present. Although this might be only formatting *) It shows the number of sheets being selected, when deleting. Preventing unintended deletion * It prevents unintended deletion, when delete is picked from context menu instead of delete sheet B) However, the dialog can also be interpreted as permanent unrecoverable - action (see Excel). However, this is commonly untrue in case of Calc: undo works. So dialog is misleading from this perspective. Yet, there are exceptions: like references (not sure if there are more cases), where undo is 'incomplete' (broken references). Which might be even by design (looking at Excel) > I have difficulty to image a use case where there would be references to an > empty sheet, but this corner case is not treated. True (In reply to ady from comment #19) > (In reply to Telesto from comment #16) > > Case A > > 1. Open Calc > > 2. Insert a couple of sheets > > 3. Select one (or more) sheets -> No warning dialog at all > > > Empty sheets, so it does not count. True, is bears not much relevance here (didn't notice before). Although I perceive this as aberration (bug), not a feature. It's an unnucessary inconsistency. Probably not by design. Sure the result is not dramatic, but still. As the following applies * It prevents unintended deletion, when delete is picked from context menu instead of rename sheet *) It shows the number of sheets being selected, when deleting. Preventing unintended deletion * A presence of single formatting on a sheet can already trigger the dialog. So barely no distinction between totally empty and nearly empty. I tend to say: it should ideally be included, IMHO. But well that's somewhat beyond this bug > No, UNDO is not "manageable". Either Calc is able to revert to the prior > status after a specific action, or it can't. We, users, are simply "lucky" – > thanks to developers, not a random event – that in many cases a mistake can > be undone. Unfortunately, (too) many users expect this to be the case in > every situation, and this is not actually true. Generally speaking: this are likely bugs. Although not quite certain about broken references. > No additional confusing messages should be added. The confirmation dialog is > not a tutorial regarding what exactly happens when deleting a worksheet. As > soon as someone finds out that something else is not completely and 100% > undone correctly, the request will come to add more details, warnings, > info... I agree. No change is better compared to what is proposed. The dialog should be seen as 'Information Dialog'. Although it 'implicit' Warning Dialog (unrecoverable action). The ambiguity is the biggest problem. However, no idea how that can be improved. > Regarding the possibility to completely skip the confirmation dialog, it is > certainly an expert configuration only. No common user should have to > confront the situation. Completely-aware experienced users should take > responsibility of deciding to skip one infrequent confirmation dialog. And > if deleting a worksheet is so frequent that a user is tired of the > confirmation dialog, then that user is most probably over-confident > regarding this confirmation dialog in the first place. Agreed. Although I'm not even in favor of another expert setting. It adds complexity to the code for a corner case. -- You are receiving this mail because: You are the assignee for the bug.
