Warnings about broken instance in `limactl list`

[nirs]

Description

limactl list may log unneeded warnings if running while an instance is created and started by another program. The command already shows the status of the instance as broken so the warning is not helpful.

Not bening able to find ~/.lima/cluster/ha.sock is an expected condition when the command is running in parallel to instance creation.

% limactl list
WARN[0000] instance "cluster" has errors                 errors="[failed to connect to \"/Users/nir/.lima/cluster/ha.sock\": stat /Users/nir/.lima/cluster/ha.sock: no such file or directory]"
NAME       STATUS     SSH            VMTYPE     ARCH       CPUS    MEMORY    DISK      DIR
cluster    Broken     127.0.0.1:0    vz         aarch64    2       3GiB      20GiB     ~/.lima/cluster
krunkit    Stopped    127.0.0.1:0    krunkit    aarch64    4       4GiB      100GiB    ~/.lima/krunkit
libnbd     Stopped    127.0.0.1:0    vz         aarch64    10      16GiB     100GiB    ~/.lima/libnbd
podman     Stopped    127.0.0.1:0    vz         aarch64    4       4GiB      100GiB    ~/.lima/podman
uvkc       Stopped    127.0.0.1:0    vz         aarch64    8       8GiB      100GiB    ~/.lima/uvkco 

[jandubois]

Personally I've also been annoyed by validation warnings from limactl ls. It happens when you create an instance with a version from a branch that has schema changes, and then you move back to the release version (or just another branch). But I guess this is something only Lima developers run into, so maybe not worth changing.