I expect Docker Desktop to lose relevance for Windows workloads as soon as containderd on Windows reaches feature parity with its Linux version.
That said, the current options are pretty low level (i.e. trying to get nerdctl running), versus the developer experience that many are familiar with.
I don't doubt that it will have some effect on Docker's market share, but saying that "Docker Desktop is pretty dead now" is perhaps getting ahead of oneself.
If anything, Podman and Rancher will have more influence in the near term, because they're (more) fully fledged and can be used today, instead of relying on software that's still very much in early active development and might have plenty of breaking changes. Not that their desktop offerings are there yet, either.
It would probably be nice to hear more about why you think this is! I've certainly heard of some having to move away from Docker Desktop.
However, at the scale where you need a license (250 employees or 10 million $ in annual revenue) it's not quite as big of an issue, especially at their current pricing per seat: https://www.docker.com/pricing/
> stick to standard open source tools like Colima etc...
Sticking to open source is a great idea!
I think mentioning that Colima runs on macOS and Linux only at the moment is also a good idea: https://github.com/abiosoft/colima
A large market share of the Docker Desktop installs are Windows in particular (since it's "the one way" how most install Docker nowadays, as opposed to not really needing a GUI or the supporting tools on Linux).
In another comment I mentioned Podman Desktop as a mostly viable alternative: https://github.com/containers/podman-desktop
Then there's also Rancher Desktop as well: https://github.com/rancher-sandbox/rancher-desktop
Regardless, it's nice to see reputable orgs behind the open source projects as well, which gives a bit more credence to their chances of surviving for the years to come.