<plumm> I overrode the source (version and src url) and gave the proper sha after using the fake one
<plumm> but im still building 2.0.8
rb2k has joined #nix-darwin
supersandro2000 has quit [Disconnected by services]
supersandro2000 has joined #nix-darwin
rb2k has quit [Ping timeout: 256 seconds]
rb2k has joined #nix-darwin
rb2k has quit [Ping timeout: 258 seconds]
rb2k has joined #nix-darwin
rb2k has quit [Ping timeout: 240 seconds]
rb2k has joined #nix-darwin
plumm has quit [Quit: My Mac Pro has gone to sleep. ZZZzzz…]
rb2k has quit [Ping timeout: 260 seconds]
rb2k has joined #nix-darwin
rb2k has quit [Ping timeout: 264 seconds]
angerman has joined #nix-darwin
<angerman> the nix-darwin install doesn't appear to create /run/current-system/sw/bin for me :-/
<angerman> might be realted to using nix-daemon in multi-user mode?
eraserhd2 has joined #nix-darwin
eraserhd has quit [Ping timeout: 265 seconds]
klardotsh has quit [Quit: WeeChat 3.0]
klardotsh has joined #nix-darwin
<LnL> nah, that's separate from the installation type
<LnL> does /run exist?
__monty__ has joined #nix-darwin
<supersandro2000> angerman: had the same issue
<supersandro2000> IIRC when the script runs the ln some path does not exist yet
<angerman> LnL: the best I could come up with is that it happens when rebuild doesn’t Finnish
<angerman> LnL: I did set network.hostname, which fails pretty far to the end, and then the link isn’t generated.
<angerman> But the error “looks” benign and not fatal. Little did I know.
philr has quit [Ping timeout: 240 seconds]
ehamberg has quit [Read error: Connection reset by peer]
prusnak has quit [Read error: Connection reset by peer]
pasukon has quit [Read error: Connection reset by peer]
ehamberg has joined #nix-darwin
prusnak has joined #nix-darwin
pasukon has joined #nix-darwin
hamishmack has joined #nix-darwin
rb2k has joined #nix-darwin
<LnL> wel if the build doesn't finish nothing happens outside of the store
<LnL> and the toplevel /run is setup separately
<LnL> networksetup is one of the last steps before everything is done
rb2k has quit [Ping timeout: 272 seconds]
rb2k has joined #nix-darwin
rb2k has quit [Ping timeout: 264 seconds]
rb2k has joined #nix-darwin
<dhess> abathur: presumably if we run https://github.com/NixOS/nix/pull/4532/files on existing systems, we won't have an issue with the 11.2 -> 11.3 update, right?
<dhess> abathur: I ask because, while you're qualifying this as being helpful if you haven't already updated 11.1->11.2, I suspect it will be useful even to those who already have.
<dhess> so you might just want to drop that qualification and say, "Hey, macOS users, run this script ASAP to avoid future problems with macOS updates." :)
<dhess> abathur: and thank you, by the way. I'm sure that was a lot of unfun work to track that down.
<abathur> dhess I'm not sure on that front, it would be nice to get feedback both ways; I'm currently trying to download a big fat beta 11.3 installer I guess
<abathur> I tried downloading the update with softwareupdate --download but I can't find the damn file
<abathur> and I figured out early on in this mess that it was better to add a partition to hold installers while I iterate rather than download every time
plumm has joined #nix-darwin
<dhess> abathur: I think I will probably run it on all of our systems (11.2). It can't hurt at this point :)
<abathur> dhess: it is both a lot of time and not; a lot of wall time, but most of it is waiting for things to install/update, so it has at least not eaten my life for several days
<abathur> I've also iterated into the partition optimizations to speed up cycles a bit which has helped
<abathur> the first several cycles I was reinstalling the beta from a bootable usb 2.0 drive, getting a partition set up to handle that task shaved quite a bit of time off of the reinstall phase :)
rb2k has quit [Ping timeout: 260 seconds]
<abathur> dhess: you might want to hold on those, bauer brought up a good point in the thread that may entail an update
rb2k has joined #nix-darwin
<dhess> Yeah I saw that, thanks
hamishmack has quit [Ping timeout: 264 seconds]
ehamberg has quit [Read error: Connection reset by peer]
hamishmack has joined #nix-darwin
ehamberg has joined #nix-darwin
<gchristensen> LnL: do you know if it would be a problem to update the builders to a more modern system?
<LnL> depends what version I guess
<gchristensen> Catalina, perhaps?
<LnL> I doubt there would be much of a problem with that at this point
<gchristensen> I'm a bit worried that having such an old system is a bit of a risk
<LnL> having the next sdk update bump straight to 10.14 might also be the way to go, so that would most likely also make that the minimum supported
<LnL> what are we at right now?
<gchristensen> HS
<LnL> ah, yeah then updating wouldn't be a bad idea
<LnL> but there's not really that much of a difference for us for 10.13-10.15 so mot a big deal either
<gchristensen> ok cool
<gchristensen> I'll try to get catalina woring :)
<LnL> great!
{`-`} has joined #nix-darwin
<LnL> in theory 10.16 would also be nice since it catches more impurities, but there are still issues in various ecosystems AFAIK
<gchristensen> by 10.16 do you mean 11, bigsur?
<LnL> yeah, isn't it only aarch64 that returns 11?
<gchristensen> ah
<gchristensen> I believe you :)
<LnL> I don't really know either, might have been just a beta thing (I haven't upgraded)
<gchristensen> I'm hesitating on big sur because it appears to require several binary blob kexts that I haven't chased down source for
<LnL> yeah, it's too early anyway
<gchristensen> good point, catalina should be well trodden by now
<LnL> it's just that with the stdenv changes + big sur we're even more isolated from the host system making it even less likely for things to break in the future
<gchristensen> eah
<gchristensen> right
<gchristensen> I realized our OVMF files we boot with are binary blobs that I might not have built myself :x
<LnL> unless Apple decides to throw us another weird curveball I expect we won't need any changes for 11.1/12
<gchristensen> cool
<LnL> don't we have those in nixpkgs somewhere?
<gchristensen> fingers crossed
<gchristensen> yeah but they're possibly patched
<gchristensen> oh, well, not anymore. they used to be posibly patched
<LnL> I could be misremembering but I think I ran some version with an nixpkgs OVMF at some point
<gchristensen> hmm
<gchristensen> I will try with Catalina and let you know :P
<gchristensen> my notes trying with HS today are like 20 pages of "okay that didn't work ... trying ..."
<LnL> don't recall what version but last time I tried it mostly just worked
<gchristensen> cool
<gchristensen> the other thing is I'd rather not have to use Clover, do you know if there are alternatives?
<gchristensen> my justification is Clover is difficult to build and I'd rather avoid binaries
<LnL> ah, I just used the existing opencore stuff
<gchristensen> link?
<LnL> I mean the binaries, didn't build it myself
<gchristensen> ah gotcha
<__monty__> Does this mean cache.nixos.org will stop caching builds for High Sierra systems soon?
<gchristensen> no
<gchristensen> well
<gchristensen> I don't think so
<gchristensen> LnL? :P
<LnL> no just _as_ high sierra
<LnL> bumping the min version is what's important for compatibliity, usually done together with bumping the sdk
<andi-> i tried getting to the bottom of the neovim compile error on an M1. Running the neovim binary with lldb shows "malformed mach-o file" o.O
<__monty__> Ah, but you did talk about bumping SDK to 10.14 and that would probably entail a bump to the minimum version?
eraserhd3 has joined #nix-darwin
eraserhd2 has quit [Ping timeout: 240 seconds]
<abathur> dhess: I force-pushed an update to that and just finished a test run that seems to be fine
<dhess> abathur: cool, I'll give it a try soon.
<LnL> __monty__: yeah, but that would still include it or am I getting versions mixed up?
<__monty__> High Sierra's 10.13. So if an sdk bump to 10.14 would but the minimum version to 10.14 I think it wouldn't be covered by the cache anymore?
philr has joined #nix-darwin
<LnL> ah, thought sierra was 10.13
<LnL> 10.12 is the main pain point currently so 10.13 is an option, but I think there's not much of a reason not to jump a version at this point
<__monty__> I understand. Does imply some pain for me personally though.
<LnL> the stable old stable release will still continue to work
<__monty__> Sure, for up to a year. I'm not sure keeping expressions up to date is less of a pain than building everything though : )
__monty__ has quit [Quit: leaving]