> And i did try even, and im not against the message itself, i understand it > is a positive thing, but imho it does need some kind of a ratelimit, maybe > once per process (or better yet; a knob for turning it off).
No. This change is a test. It is an uncommited test. It is in snapshots, but not in the tree. We sometimes trial a change of that sorts in this way, so that we can learn how a subsystem behaves over a large body of software. It is easy to put such a change into snapshots, and collect a large amount of feedback rapidly. Thanks for your feedback about problems it exposes. But no thank you regarding your suggestions that this trial should become user-friendly, we will spend time thinking about the results to make the next diff, but ratelimiting is not interesting. > I'm a user who does almost constantly switch different usb devices attached to > my workstation while hacking, and these messages keep coming up on the > background with no user interaction what so ever, so i either need to focus > checking the console immediately after plugging in, or like whole of last > night due this new feature fckng up on my habbits: grep or scroll dmesg > manually every time.. Then build your own kernels, they will not have the uncommited diff.
