R. Boujbel

> The behaviour has been there the since the introduction of the option in [345b475](https://github.com/ocaml/opam/commit/345b475a81fa2c3b3cc79eed4a9263e2c615364f). This commit introduces an explicit exception to the rule of checking for the result, just...

~~queued on #5349~~ rebased

related to #6530. As we update the whole switch, we end up with a case where we give the wrong information for an operation. It is another argument to this...

Tested the PR, working as it should be :+1:

I'm not sure if it should fail, but if not, there is no way to show the message to user as it is embedded in sandbox script. It is done...

Discussion on dev meeting: We don't want to enforce failling jobs on package that we don't maintain, but the information can be used to know which maintainer/project to ping. Ideally,...

I've extracted commenting in another PR (#6646) to unblock this one.

> There are definitely some parts that can be improved (e.g. every project reinstall their dependencies, even if they have been installed before because a new local switch was created),...

related to #5130 too

Update: ``` $ opam ci logs prof_spacetime prof_spacetime: multiple build failures found with different configuration parameters. Please refine the command to select exactly one of the following: opam-ci logs prof_spacetime.0.1.0...