Dear Pavel,
In message 20141009230559.GB25685@amd you wrote:
v2: added tags to the subject v3: added diffs to previous version . (From memory, but IIRC something very similar to this happened before).
Yes, this happens when people repeatedly ignore to read the patch posting rules.
I'd argue that if only changelog is updated, it is _not_ a new version of patch, and does not need changelog diff. Or maybe be less strict policy / less strict enforcement of the policy in trivial cases.
You ignore the fact that you are supposed to miimize the work load you create on reviewers. If I try to review a patch, I need to understand what has been changed compared to the previous version. Yes, of course I can apply both versions and run a diff (diffs over diffs are too difficult to read in general), but this costs time. And each reviewer has to spend this time.
On the other hand, you should know exactly what you've changes, so it is minimal work for you to add such an explanation. It you use the provided tools for the task (patman comes to mind), this is even well supported.
So what is better for the community: if you spend a little time once, or if every reviewer spends much more time trying to figure out what you might have changed? So it all depends whether you take the egotistic or the community point of view....
Best regards,
Wolfgang Denk