Lennart PoetteringHere's the third published video of my FOSDEM talks, about the Varlink IPC system:
https://video.fosdem.org/2026/ub2147/NFNKEK-varlink-ipc-system-keynote.av1.webm
This was the last one of the three. Enjoy!
Sebastian Wick@pid_eins should have counted the context switches of a portal call for maximum effect :p
@swick i am curious, for some simple portal how many context switches would you guess? (For a single portal gated request, ideally non-interactive if there is such a thing)
best case: 6 context switches [request: app -> dbus-proxy -> broker -> portal frontend] -> [response: broker -> dbus-proxy -> app]
but with auth required on every connection: 8 context switches [request: app -> dbus-proxy -> broker -> portal frontend -> [credentials: broker -> portal frontend]] -> [response: broker -> dbus-proxy -> app]
and with a backend call: 12 context switches [request: app -> dbus-proxy -> broker -> portal frontend -> [credentials: broker -> portal frontend] -> broker -> portal backend] -> [response: broker -> portal frontend -> broker -> dbus-proxy -> app]
@pid_eins oh, and with varlink and appd it can be exactly two because no more dbus-proxy, no more frontend/backend split, and no call for the creds/permissions.
@swick yippiiiehh!
Ryan BrueHow does appd fit into this? Varlink alone seems like it'd be enough to replace dbus. Would appd be for authentication and such?
P.S. we should definitely expose varlink interfaces as alternatives to DBus if possible in the nearer future than appd ;p
@ryanabx well, appd would allow us to go from unix socket to app identity.
I am a bit worried about people exposing all kinds of varlink interfaces to sandboxed apps. the nice part about portals is that it forces everyone through one place where we make sure everything functions correctly together without poking holes in the security. we really have to retain that.
@swick aaaaaah
Sergey Bugaev
S1m