qyliss changed the topic of #spectrum to: A compartmentalized operating system | https://spectrum-os.org/ | Logs: https://logs.spectrum-os.org/spectrum/
<hyperfekt> paper on lupine linux, which uses kernel mode linux (latest is 4.0 :/ ) and a minimal config for tiny & fast VMs https://dl.acm.org/doi/pdf/10.1145/3342195.3387526
<hyperfekt> i'm surprised i can't find any references to a UML with the kernel compiled into a glibc instead of using tracing
<hyperfekt> s/glibc/libc/
multi has quit [Ping timeout: 240 seconds]
multi has joined #spectrum
julm has joined #spectrum
cole-h has quit [Quit: Goodbye]
<MichaelRaskin> hyperfekt: maybe NetBSD rump kernel?
<qyliss> Pushed a couple of small website additions: https://spectrum-os.org/git/www
cole-h has joined #spectrum
tilpner has quit [Remote host closed the connection]
<qyliss> eep I gotta get started on TWIS
<Shell> TWIS?
<Shell> oh
<Shell> lol
<cole-h> qyliss: I feel so special; 5 mentions in a single piece of mail :D
<Irenes[m]> we have read it. good stuff!
<cole-h> Now I kinda wish I paid more attention in that first stream where you actually build and run cros...
<qyliss> cole-h: by making you feel special i aim to encourage further contributions :P
<cole-h> Hehehe
<qyliss> Follow-up because I forgot a couple of things: https://spectrum-os.org/lists/archives/spectrum-discuss/877dxjpdsv.fsf@alyssa.is/
<cole-h> I can certainly nitpick documentation!
<qyliss> cole-h: there can be another crosvm stream
<qyliss> Actually, the stuff I'm about to start working on would probably make for an interesting stream
<cole-h> (I plan on combing www/ and doc/ after finals to see if there's nothing else I can fix there)
<qyliss> Would anybody be up for a stream tomorrow? I don't want to leave starting this work any later than that, but would be a good stream topic.
<Irenes[m]> sure, sounds nice
<qyliss> let me know times that work for folks and I'll try to schedule it around that :)
<qyliss> Oh, I guess the website should probably mention that there are streams sometimes
<cole-h> Any time works for me, but it'd be good to know sooner rather than later so I can plan my productive hours around it... :P
<Irenes[m]> true, it should :)
<Irenes[m]> my hours during which it's possible to focus on stuff are roughly 18:00-01:00 UTC
<qyliss> Okay, tomorrow evening UTC
<Irenes[m]> cool! looking forward to it
<cole-h> Works for me!
<cole-h> qyliss: So, will you be working off interguest, or master, tomorrow?
<qyliss> Interguest
<qyliss> interguest is where all the interesting stuff is happening
<qyliss> but we'll be looking at crosvm and a Wayland compositor, and trying to observe how they talk to each other
<qyliss> or at least, that's the plan. Streams don't always go as I plan them to. ;)
<cole-h> Hehe
<cole-h> Quick question: how do we get the kernel that cros asks for when `cargo run -- run`-ing?
<cole-h> Will any linux kernel do, or does it need to be ChromiumOS kernel?
<cole-h> Ah, probably `chromiumOSPackages.linux` is what I want
<qyliss> Yep :)
<qyliss> You might also want to check out spectrumPackages.spectrum-vm
<qyliss> Although I'm not sure how well it will work with a crosvm with broking sandboxing...
<qyliss> But at least the source for spectrum-vm will show you what kernel and rootfs and stuff to use
<MichaelRaskin> Re: cloning — maybe recommend adding a spectrum remote and using worktrees?
<qyliss> I'm gonna have the server capacity to handle it, so it's okay.
<colemickens> I keep meaning to write up "how I contribute to nixpkgs with worktrees". highly underrated git feature
<qyliss> I doubt people are gonna read my instructions for how to git clone anyway
<cole-h> Yeah, I've started using worktrees and they're a real blessing
<MichaelRaskin> If you bold the works remote and worktree, people might find out worktrees exist
<cole-h> No more doing the stash-checkout-change-commit-back-pop dance
<Irenes[m]> wait what's a worktree
<qyliss> A worktree is when you have multiple working copies of a single git repository, with different branches (etc) checked out
<colemickens> another checkout from the same singular cloned repo
<qyliss> But all your remotes and stuff are the same
<Irenes[m]> oh cool! handy
<qyliss> For spectrum's nixpkgs, I think I'd actually recommend using git clone --reference over worktrees, because otherwise the branches get confusing.
<Irenes[m]> makes sense
<qyliss> Is "master" based on Nixpkgs master or Spectrum master, for example
<Irenes[m]> still, good to know about
<qyliss> Yeah
<MichaelRaskin> Not much more confusing than normal git branching conventionms
<qyliss> It's very handy when you just want to be able to work on two branches at once
<qyliss> A frequent pattern for me is git worktree add /tmp/whatever
<qyliss> And then I'll use the worktree for some one-off patch or whatever, and then discard it
<qyliss> (Git will notice eventually if a worktree has gone away and clean up internally)