<colemickens>
I'd love to see a writeup/summary of NixCon talks.
<colemickens>
I was excited by the various discussions about how to cleanup nixpkgs and achieve better discoverability and usability for end users.
<colemickens>
It would be useful to have a summary doc people could read. I'm thinking about hte YAML thread in Discourse... I think even a NixCon talk discussed that with a proper "package/module" system in nixpkgs, it could be possibly to do trivial package declaration in a data-only way (yaml/json) potentially.
<colemickens>
Also curious how we can elevate visibility of other repos like `nixos-hardware`.
tertl3 has quit [Quit: Connection closed for inactivity]
jackdk has quit [Disconnected by services]
jackdk_ has joined #nixos-chat
hyp3rbor3ax_ has joined #nixos-chat
hyp3rbor3ax_ has quit [Remote host closed the connection]
hyp3rbor3ax_ has joined #nixos-chat
hyp3rbor3ax has quit [Ping timeout: 268 seconds]
hyp3rbor3ax_ has quit [Remote host closed the connection]
hyp3rbor3ax_ has joined #nixos-chat
hyp3rbor3ax_ has quit [Remote host closed the connection]
<drakonis>
colemickens: nixos weekly maybe?
<colemickens>
yeah, I meant to open an issue or pr to add nixos-hardware at least.
<colemickens>
I am trying to spend some time in Nov blogging about nixos stuff (evangelizing?) and will try to touch on the things I discussed, maybe it will be useful or interesting. I know videos aren't always the best for everyone
<drakonis>
i'd suggest a book
<drakonis>
definitely need to get improved docs in general
lassulus_ has joined #nixos-chat
<colemickens>
Yes. Either upstream docs, or the wiki is where I want most of what I write to be, and then small "pointers" from a super highlevel, short blog post type thing.
lassulus has quit [Ping timeout: 245 seconds]
lassulus_ is now known as lassulus
pie__ has joined #nixos-chat
pie__ has quit [Ping timeout: 252 seconds]
snajpa has quit [Ping timeout: 264 seconds]
srk has quit [Ping timeout: 240 seconds]
snajpa has joined #nixos-chat
srk has joined #nixos-chat
snajpa has quit [Excess Flood]
snajpa has joined #nixos-chat
snajpa has quit [Excess Flood]
snajpa has joined #nixos-chat
snajpa has quit [Excess Flood]
snajpa has joined #nixos-chat
srk has quit [Ping timeout: 240 seconds]
snajpa has quit [Remote host closed the connection]
<drakonis>
a freebsd ports guy is porting nix to freebsd
<drakonis>
i assume he hasn't come to any of the irc channels to ask about it
<ekleog>
\o/
<drakonis>
its neat and surprising, but i think it'll be an uphill battle for them
<drakonis>
i'm not sure how well it works right now but the port exists
<drakonis>
oh, it doesn't even build yet
drakonis_ has quit [Ping timeout: 260 seconds]
<drakonis>
cool nevertheless
drakonis_ has joined #nixos-chat
jackdk_ has quit [Remote host closed the connection]
drakonis has quit [Quit: WeeChat 2.2]
jackdk has joined #nixos-chat
<MichaelRaskin>
Well, there was support for FreeBSD in Nix some time ago; even in Nixpkgs. Of course such things bitrot without someone who tests them and provides fixes. I think readding FreeBSD support should be feasible and hopefully the fixes are natural enough to be accepted into Nix in a reasonable timeframe.
sir_guy_carleton has quit [Quit: WeeChat 2.2]
<drakonis_>
it'll just bitrot again due to the difference in culture
<MichaelRaskin>
I think if there are a few developers who are willing to use Nix on FreeBSD, this might be enough to keep some basic support for Nix and stdenv on FreeBSD upstream. We are bad enough at consensus here that we shouldn't even have a culture that could have an idea of difference.
jasongrossman has quit [Ping timeout: 244 seconds]
<MichaelRaskin>
Of course, if in the end there are no FreeBSD users willing to send patches, adding support once won't do much good.
<drakonis_>
heh, well, i find that the average user wants ports
<drakonis_>
which is similar to portage i suppose
<MichaelRaskin>
Average computer user uses Android and doesn't understand anything they are doing. We need five enthusiasts or something like that.
<drakonis_>
i'll ask the port commiter about it
<drakonis_>
because the current state is that it doesnt build
<MichaelRaskin>
Maybe wait until it does build?
<drakonis_>
good idea
<MichaelRaskin>
I think the history teaches us that there is no willingness in Nix project to care about platform coverage without a few users willing to use this support and pushing specific working patches.
<drakonis_>
true
<MichaelRaskin>
Yes, there still will be some pushback, but «more platforms help keep assumptions in check» crowd can be gathered quickly around a clean-ish working patch.
<drakonis_>
freebsd has extremellt poor support
<drakonis_>
lots of packages would need to be modified for support
<drakonis_>
which is a frequent occurence there
<drakonis_>
might seriously dirty things up
<MichaelRaskin>
Well, how many of the changes are explaining to the package that gmake is the make you need?
<drakonis_>
most of them
<MichaelRaskin>
Note that we already force nontrivial FS layout on all the packages
<drakonis_>
yup
<MichaelRaskin>
Upstream Nixpkgs support for FreeBSD could be «things that are easy to build with glibc-on-FreeBSD». This can be added with reasonable «splash damage»
<MichaelRaskin>
And this can be useful for some FreeBSD users.
<drakonis_>
language specific packages
<MichaelRaskin>
And until we have tested code for that minimal base, there is no point to plan further, I think.
<drakonis_>
fair enough
<MichaelRaskin>
There is a useful MVP and it _is_ a blocker for anything larger.
<drakonis_>
nothing we can do until the first hurdle is cleared, as i dont know how advanced the port dev is on using nix
<drakonis_>
there are nontrivial compiler patches
<drakonis_>
rust has a a cool 30
MichaelRaskin has quit [Quit: MichaelRaskin]
jackdk has quit [Ping timeout: 272 seconds]
jD91mZM2 has joined #nixos-chat
pie_ has joined #nixos-chat
Synthetica has quit [Quit: Connection closed for inactivity]
pie__ has joined #nixos-chat
pie_ has quit [Ping timeout: 272 seconds]
jasongrossman has joined #nixos-chat
<srhb>
Hum.. Is there no way to spawn a one-shot jobset that disables if it fails to evaluate?
<srhb>
I suppose I can disable polling and command it to evaluate right after creating it..
<infinisil>
srhb: You could just make any exit code indicate success
<infinisil>
Along with oneshot and RemainAfterExit or so
<srhb>
infinisil: *boggle*
<srhb>
I didn't understand that at all. :D
<infinisil>
srhb: Oh, a hydra jobset?
<infinisil>
I thought you meant systemd
<srhb>
Ah sorry! My bad. Hydra, yes :)
<srhb>
There's something weird about oneshot being none-shot.
<srhb>
Workaround appears to be to set the poll frequency to something insanely high. :P
<srhb>
It won't disable it, but at least it won't keep evaluating either..
LnL has quit [Ping timeout: 245 seconds]
LnL has joined #nixos-chat
Synthetica has joined #nixos-chat
drakonis has joined #nixos-chat
__Sander__ has joined #nixos-chat
LnL has quit [Ping timeout: 252 seconds]
<elvishjerricco>
Thought I'd take a peak at bootstrapping GHC from the very beginning. Seems like the easiest way is to compile HBC and use that to compile early versions of GHC. But it seems quite hard to get HBC to build. Seems to need a pretty darn old C compiler.
<maurer>
elvishjerricco: I thought hugs might have a C bootstrap you could use?
<maurer>
Yeah, hugs is entirely in C, and supports Haskell98
<elvishjerricco>
Or at least, nhc98 isn't a good place to start
LnL has joined #nixos-chat
<maurer>
Huh, I didn't know that early ghc depended on recursive modules
<maurer>
Although, you could probably patch early ghc to not
<elvishjerricco>
But HBC is attractive because it's the actual compiler used to build the first GHC. It just needs a C compiler to bootstrap it's LML compiler.
<elvishjerricco>
So there's a chance all I need is to find some old GCC and glibc.
<maurer>
What is it doing that needs old gcc out of curiosity?
<elvishjerricco>
I'm assuming more than one thing, but the one I've hit so far is that varargs.h is no longer implemented. I could probably patch that, but I'm particularly attracted to using unmodified sources and reproducing the actual lineage
<maurer>
varargs.h was legacy before I started coding...
<elvishjerricco>
Lol yea this stuff is from like 1988
LnL has quit [Ping timeout: 252 seconds]
LnL has joined #nixos-chat
<joepie91>
help
<joepie91>
one of my old blog posts is at the top of HN again :(