This is unfortunately fairly normal. There is usually a lot of things that have happened before the corruption is noticed. The last one I saw was more straightforward, in that there was (I think anyway!) a single mistake due to an odd part where power and ground got reversed in schematic while correct in breadboard. However even with that, I have not been able to create a sequence of changes that cause the corruption. There is some specific set of actions that causes the corruption (it looks like a delete did not complete for some reason leaving enough elements around to cause the corruption but I don’t know why), but it isn’t at all obvious what the sequence is. Until we can reliably reproduce the problem we can neither trace the issue to fix it, nor be sure any fix has actually fixed the problem. It may be that we need to add the facility to save the undo stack in Fritzing to a file so that once the corruption occurs it would be possible to step back through the undo stack to see how the issue occurred and hopefully reproduce it by reapplying the changes. I’m however not sure that is feasible, I expect it may be difficult to write the undo stack out to a file and that file may get very large, usually for no purpose because this happens very rarely.
Peter