Mysteriously Cloned Cells

[djl47]

I have a number of large spreadsheets that I edit and update. Each spreadsheet contains multiple sheets. Each sheet consists of two formula columns on the right and a text column to the left.
The following does not occur all of the time but often enough that I have noticed a pattern:
Open one of the worksheets on the right
Scroll down a thousand or so rows and double click in the text cell to edit the text. Make a change. Note the row number.
Control F to launch the search UI. (leave the cursor in the cell being edited)
Enter a search string
Make sure search in all sheets is checked
Run the search
Every now and then the contents of the cell where I started is cloned into the same row and column on the left hand sheet.
I began noticing this when some searches landed on text that was far below the end of the sheet and was identical to text on the same row and column in another sheet.
I have not seen any text cloned from left worksheets to right worksheets, only when I start in a right hand worksheet, reach the end of the document and continue the search.

My workaround is to make sure I click the cursor into an empty cell before starting the search.

[acknak]

We discussed this back here: http://user.services.openoffice.org/en/ ... =9&t=27172

I filed a bug report that (so far) has not been officially acknowledged: Issue 109084: search while in edit mode can cause data to be overwritten

If you take a look at the report and agree that it does seem to be the same problem you're seeing, maybe you could register on the bug tracker and add a comment and/or a vote (up to two) on that issue. More activity might raise the visibility a bit.

[djl47]

The problem described in the linked thread is similar and I suspect it is the same problem. I need to spend some time seeing if I can reproduce your problem description. In my example I did not place the contents of the cell on the clipboard. The software automagically cloned the content of a cell located on of sheet 3 row x column c to sheet 2 row x column c.
I think one of the key factors in causing this is to start the search using control f while leaving the cursor in the cell being edited. Give me a few days to tinker with this and I'll vote for your bug when I can add some cogent content beyond a the vague description I have now.
Thanks

[acknak]

The software automagically cloned the content of a cell located on of sheet 3 row x column c to sheet 2 row x column c.

Is it possible that you had both sheets selected?

When you enter data, it will go into the same cell in every selected sheet, all in the same action.

[djl47]

The cell was cloned into one other sheet among up to seven sheets. It's a good thing to be aware of though. I didn't know about this feature. I'll try it tonight.

[djl47]

I tinkered around with the problem and can consistently produce it now.
Open the document
Click on sheet1
Position the cursor on a blank cell in column c. Note the line number.
Enter some data in the cell.
control F to bring up the search menu. Make sure search entire document is checked.
Enter sam or tom in the search window and start the search.
The data you entered will be moved from the cell on sheet1 to the same cell on the sheet where the string is found.
+++
Click on sheet1
Position the cursor on a blank cell in column c. Note the line number.
Enter some data in the cell. Click on another cell, then double click back into the cell you modified. Add something to the cell.
control F to bring up the search menu. Make sure search entire document is checked.
Enter sam or tom in the search window and start the search.
The data you entered will be copied from the cell on sheet1 to the same cell on the sheet where the string is found.

[acknak]

Ok, thanks for posting the steps--I see it also.

It looks like the same problem to me: the core steps in both cases involve using the find function while entering data, before pressing Enter. If the find function changes the active sheet, then the data being entered end up in the other sheet.

[djl47]

Stopped by to grab the attachment I created to test this five years ago. It is sad that a data corrupting bug is marked trivial in the bug report. It's caused me enough grief over the years.

[acknak]

I wouldn't put too much importance on the "trivial" label. I suspect it's there because the sequence necessary to trigger it seems unlikely, so the bug is unlikely to cause problems for many users.

Developer resources are very scarce; there are lots of higher-priority issues.

Constructive approach: add a brief comment to the report to keep the issue "alive".

[djl47]

The sequence to trigger it is very likely. Most users will never connect that sequence with the data corruption that they discover days or weeks later.

[acknak]

I'm not saying it is or isn't likely to happen, only that it seems unlikely. I've never run into it (I don't search while entering data) and I don't remember it coming up on the forum except here. Sure, maybe that's because no one notices; it's possible.

If the issue is interfering with your work, add a comment to the issue and make a case for raising the priority. Something brief and concise is more likely to be read.

As I hear it, resources are scarce enough that even regular releases are delayed, and only the highest priority issues are getting any attention.