Le jeudi 02 juin 2022 à 10:28 -0400, Sandro Tosi a écrit : > > I would suggest ratt-rebuilding all reverse dependencies. Could > > that be > > done? > > there order of thousands rdeps, i dont think it's fair to ask any > individual contributor the time and resources to check that via ratt.
Agreed. If the list of packages to check can't be handled by my modest setup during the night, I won't check them all. > Something i've done in the past (f.e. with numpy and matplotlib) is > leveraging Lucas' archive rebuild infrastructure to run a rebuild the > rdeps with a new package, usually uploaded in experimental. When I upgrade a package with a number of rdeps I deem unreasonable for my setup, I pick some rdeps semi-randomly (mostly random, but having a look at the list to hand-pick a few promising ones, or make sure I drop problematic ones [sagemath comes to mind]). That won't guarantee nothing will break, but does detect glaring issues. And indeed, uploading to experimental and giving people a head's up if there's a thousand rdeps and you can only test 1% is important. Cheers, J.Puydt