Flatpak

91 readers

7 users here now

founded 9 months ago

MODERATORS

cm0002@programming.dev

cm0002@lemmy.world

Fix for KeePassXC browser integration (codeberg.org)

submitted 1 day ago by sockenklaus@sh.itjust.works to c/flatpak@programming.dev

3 comments fedilink hide all child comments

Out of the box a browser installed via Flatpak can't communicate with KeePassXC using the browser extension because of Flatpaks encapsulation.

This can be fixed by adjusting the browsers Flatpak permissions, adding a wrapper for keepassxc-proxy to the browsers container and spawning keepassxc-proxy from within the container.

I built a script that does this for you. Currently supports Firefox, Librewolf and Ungoogled Chromium.

Have fun!

you are viewing a single comment's thread
view the rest of the comments

[–] Matt@lemdro.id 1 points 1 day ago (1 children)

Thank you for the script!

I wish the Flatpak developers would start focusing on the little details to solve problems like this. I have been using Fedora Kinoite for a couple of years and it heavily depends on Flatpaks. It works well, but due to many problems like this, I no longer try to fix Flatpak issues. I always try Flatpaks first, but revert to layering native packages at the first problem. I just do not feel it is worth all of the workarounds.

I will try to remember to give this script a try the next time I do a fresh install before falling back to native packages.

[–] sockenklaus@sh.itjust.works 2 points 21 hours ago

I wish the Flatpak developers would start focusing on the little details to solve problems like this.

I don't know if this really can be considered a problem. In the end Flatpak is working as intended here and encapsulates the browser from the rest of the OS which unfortunately happens to kill the communication with KeePassXC.

I am by far no expert for browser extensions but the fact that an extension can launch executables outside the browsers scope looks kinda sketchy to me.

Flatpak adds another layer of security here, which is great... But I agree with you that the way we have to solve this behaviour in this situation is not very elegant. It would be nicer for example to give the browser the possibility to ask Flatpak for additional permissions that have to be granted by the user.