I would not deduce anything from this. Most likely the system changes the date whenever the editor accesses the article (or the editor does not use the system in the way it was intended).
Also, it happens that the state is not accurate (e.g., in reality, the paper is already with the reviewers), but there is no way to know that.
Anyway, many journaljournals have a processing time span that is much longer than two months. If after some time the state is still "with editor", you might politely contact the editor to inquire about the process.