Christophe Fergeau

> Regarding this, I propose implementing a "mock" segment client that will be returned if the initialization of the real segment client fails. Can be a separate "NoopClient" which is...

Can you try with `crc setup --bundle ...` or `crc config set bundle ...` to see if this works? I would move the `.crcbundle` out of `~/.crc/cache` for these tests....

I've rebased it at https://github.com/cfergeau/crc/tree/stus-change-event and added a fixup commit to fix compliation of the first commit in the series.

I'm not sure step 3. is correct in the testing instructions: > 3. start crc with crc start Don't we need to start it with the REST API for now...

Few more things I noticed during testing, if VM exists, and then I call `delete`, I get ``` id: data: {"status":{"CrcStatus":"Error","OpenshiftStatus":"","OpenshiftVersion":"","PodmanVersion":"","DiskUse":0,"DiskSize":0,"RAMUse":0,"RAMSize":0,"Preset":""},"error":"Cannot load 'crc' virtual machine: no such libmachine vm: crc"}...

> What is holding back this PR? Some discussion around https://github.com/crc-org/crc/pull/3903#discussion_r1476130780 would have been useful Apart from this, as said multiple times, the comit log must be a lot more...

Yes the problem is the `nosuid` mount.

> Ideally version is taken from the bundle but it might be just clusterVersion which will work for microshift,openshift and OKD side instead having more filed around all different preset....

> My suggestion is to use crc config to store the path of the .crc directory which can be set and unset using the crc config command. This makes validation...

> `INFO A CRC VM for OpenShift 4.11.7 is already running` Run `crc delete` before retrying `crc start` to make sure it's recreated from scratch. `lookup api.crc.testing on 192.168.8.1:53` this...