REF warning

[TAB]

If B1 refers to A2 and row 2 is deleted, B1 is no longer valid and shows the #REF! error message. OK. If I see it, I can undo the deletion.
But if column B is hidden, the message is not shown, and I may not realize I made a mistake until it's (much) too late.
Suppose column W suddenly fills up with #REF! as a result a wrong deletion or edit. If I am working on column A, column W is off screen, and I don't see all these #REF!.

Conclusion: Calc should warn if an edit operation is going to cause mayhem. Or, at least, warn if it just did, so that the user can undo the operation.
Does it?

[kingfisher]

Have a look at Tools >Detective. I've never delved into it but it traces dependents.

[Dave]

Other spreadsheets behave in a similar manner. So, it should be expected and care taken when designing the spreadsheet. it has been my personal experience that when programming at any level as much time or more is spent working around the limitations of a language as with the finite list of tools available. Same with the spoken language if you think about it. One must simply be aware of the limitations, and work within them and around them when designing the program.

David.

[TAB]

Have a look at Tools >Detective. I've never delved into it but it traces dependents.

Yes. But that does not solve my problem: accidentally causing a REF error. In theory, I know my spreadsheet and its dependencies, but...
Same if you try to close an unsaved document, or delete a style: OpenOffice checks if you really mean it.

[kingfisher]

You could set up a range of cells which reflect cells likely to be affected, so that you could check for errors in one location. There's no gui setting, AFAIK. It might be possible to write script to set up a listener but I imagine it would not be a trivial matter.

[TAB]

Other spreadsheets behave in a similar manner.

So? Should Calc behave as badly as others?

So, it should be expected and care taken when designing the spreadsheet.

Even a well-designed spreadsheet is subject to error, and the best designer can be tired or distracted. For example, I might delete an 'empty' row --that contains hidden data that are referenced in other cells.

it has been my personal experience that when programming at any level as much time or more is spent working around the limitations of a language as with the finite list of tools available. One must simply be aware of the limitations, and work within them and around them when designing the program.
David.

What do you mean by 'programming ' or 'designing the program'? programming Calc so its does certain things for the user, or using Calc?
Since Calc immediately recalculates, and displays REF error messages (in visible cells, anyway), it should very simple to program Calc to pop a message at the first such error: 'Warning: ref error in cell ..; Continue?/Cancel?'. I am not asking for a major enhancement.
I can be 'aware of the limitations', but I will be more comfortable with a friendly software that warns me if I make a mistake. Un-hiding 2 columns and discovering 76 cells filled with REF! is pretty unsettling --when and how did THAT happen???
I

[kingfisher]

To initiate such a change, you would have to file a request for enhancement. See the thread about filing a bug report on the Tutorials board.

[Weatherlawyer]

Other spreadsheets behave in a similar manner. So, it should be expected

What a dumb reply.

I lost a lot of work redoing a sheet using the undo button before finding out that I was also undoing everything posted on sheets two and three.

So because it happens on other office-ware, it is all right? In which case why does OOo take pride in not having poor code?

The obvious work around is to open the stuff in a new sheet and save that and then go and pay with the undo, now I know. Not that that is much help when a giant copy will sometimes cause a white out. As when that happens all you can do is close the programme and maybe lose the lot.

And if I was that careful, I'd not make the first mistake in the first place.