Just as a warning, the macvlan stuff isn’t well documented and seems to have hard limits. I worked with it a couple of years ago and had to eventually read a lot of Docker code to figure some stuff out, and the host was only able to successfully set up 4 macvlan networks at a time - the fifth (and any following ones) were never reachable, even though I used the same scripts as for all other ones.
Things might have improved in the meantime.
The Podman developers did contribute to Docker for a while before starting the project. Docker kept introducing issues and had some fundamentally bad design decisions that they didn’t want to change.
At least try to look into the history of these things before making broad and easily falsifiable statements.
Podman wasn’t built due to NIH. Docker has real problems (though many have been fixed), and Podman was built to fix those.
I am happy with my simple docker-compose setup - one root folder with one subfolder per project containing the compose file and any configuration mounted into the container. Traefik automatically exposes all services I want under a well-known URL using a single line in each compose file. Watchtower updates the containers.
This has been running stable for over two years with probably 2-3 reboots in between. If my current NUC ever breaks I’ll set it up again using Podman instead of Docker, but aside from that I couldn’t be happier!
Very understandable :)
An easy fix for this is to create individual networks for connections. I.e. don’t create one network with Gitlab, Redmine and OpenLDAP - do two, one with Gitlab and OpenLDAP, and one with Redmine and OpenLDAP.
Yeah, if a CEO has to lie to make their product seem better, it’s blacklisted in my mind.
Thank you for the correction! Then it’s also wrong due to Gitea which launched in 2016.
Also plain wrong - Codeberg launched in 2019. Now the question is: did the author just not know better, or is he paid not to know?
Is it pre installed nowadays? I rember having to go to some store and doing stuff to get it…
It’s not pre-installed, but it’s checking one checkbox. Less work than deciding which VM provider to go with.
Mounts and networks should be just checkboxes, dropdowns iirc.
They should be, and yet I’ve rarely seen them work out like that. Usually I have to debug some issues and follow x StackOverflow responses which don’t work properly. Haven’t had any such issues with WSL2 yet.
Terminals are probably better on linux anyway, if we really want the stone age windows tools we can always ssh into it from windows.
… no. Windows Terminal integrates with WSL2 and allows you to open a terminal in Linux without having to set up anything inside of a good Terminal app in Windows. It’s what you’re asking for, but without any setup.
I didnt really get the gui part, linux vm can have, and run GUIs
Yes, but inside of a separate canvas. WSL2 GUI apps run as normal windows.
all the intellij stuff are available for linux natively
Okay, but I’ve tried running them in a VM and in WSL2. It is integrated the best if you run it under Windows and use the native WSL2 integration. Everything else degrades the experience.
Even then iirc they can run with any linux remotely as well, just needs ssh
Yes, and then you have to set everything up. With WSL2 in PyCharm I select “Use WSL2 Python”, it lists all the WSL2 Pythons, and I select the WSL2 Python I want. Is it really so difficult to understand that there is a difference between being able to do something and something just being available without setup?
If you need it to run on windows like native apps, maybe use Xserver via ssh.
Or I install WSL2 and skip all that.
As far as quirks I read some comments in this thread about filesystem being too slow, maybe there are more.
Yeah, you should read up on how WSL2 works. This is not an issue in any different way from VMs. WSL2 is a VM. It’s everything you’re asking for, but standardized, pre-installed and perfectly integrated. I don’t know why you’d recommend spending all those hours when it’s absolutely not necessary.
It’s like telling a beginner “Yeah, do Linux From Scratch, Ubuntu is way too convenient”.
WSL2 is already installed and running without any extra setup required, so all the setup time for the VM is additional time you have to spend. You’ll have to fiddle a lot longer than 45min the first time you set it up if you want parity with the WSL2 installation (bidirectional mounts, bidirectional network access, GUI applications as normal windows, integration into Terminal etc). Until everything is running you’ll probably spend half a day, since you’ll have to first look up how to best do these things for your VM environment. Even more so if you want to use Windows tools with WSL2 integration, like the whole IntelliJ suite.
What features and quirks are you referring to?
And as such it’s perfectly suited for what OP is trying to achieve. A custom VM will be a lot more work than using WSL2.
The Linux Terminal sounds pretty cool, but I can’t imagine it will amount to much.