Lisanna has quit [Remote host closed the connection]
infinisil has quit [Ping timeout: 240 seconds]
infinisil has joined #nixos-dev
pxc has quit [Ping timeout: 265 seconds]
pxc has joined #nixos-dev
pie__ has quit [Ping timeout: 276 seconds]
ma27 has joined #nixos-dev
ma27 has quit [Ping timeout: 276 seconds]
MichaelRaskin has quit [Quit: MichaelRaskin]
zybell_ has quit [Ping timeout: 240 seconds]
orivej has joined #nixos-dev
zybell_ has joined #nixos-dev
orivej_ has joined #nixos-dev
orivej has quit [Ping timeout: 260 seconds]
goibhniu has joined #nixos-dev
pxc has quit [Ping timeout: 260 seconds]
pxc has joined #nixos-dev
<Mic92>
do we backport minor nodejs upgrades? (9.10.1 -> 9.11.1)
<adisbladis[m]>
Mic92: Typically we do. If you want me to I can do it (I'm the one who usually does it)
<adisbladis[m]>
Actually I'll get on that right now anyway.
<Mic92>
adisbladis[m]: ok. go ahead
pxc has quit [Ping timeout: 276 seconds]
orivej has joined #nixos-dev
orivej_ has quit [Ping timeout: 256 seconds]
__Sander__ has joined #nixos-dev
ma27 has joined #nixos-dev
ma27 has quit [Client Quit]
ma27 has joined #nixos-dev
orivej has quit [Ping timeout: 245 seconds]
orivej has joined #nixos-dev
ciil has quit [Quit: leaving]
<LnL>
urgh, how do you kill the substitutes cache again?
ciil has joined #nixos-dev
<aminechikhaoui>
kill the substitution or the disk cache ?
<LnL>
the cache
<LnL>
there's a binary-caches-*.sqlite somewhere
<aminechikhaoui>
check ~/.cache/nix/
<LnL>
yeah that's what I thought but .cache is empty
<LnL>
oh root
<aminechikhaoui>
if you're on nix master there are also few options to disable that now
pie__ has joined #nixos-dev
__Sander__ has quit [Ping timeout: 260 seconds]
__Sander__ has joined #nixos-dev
orivej_ has joined #nixos-dev
orivej has quit [Ping timeout: 256 seconds]
zybell_ has quit [Read error: Connection reset by peer]
orivej_ has quit [Ping timeout: 265 seconds]
zybell_ has joined #nixos-dev
<niksnut>
hm, why does release-18.03 have a newer version of e2fsprogs than master?
<Dezgeg>
the upgrade is still stuck in staging
jtojnar has joined #nixos-dev
<niksnut>
?
<niksnut>
oh
<niksnut>
I see it was cherry-picked in the wrong direction
jtojnar_ has joined #nixos-dev
jtojnar has quit [Ping timeout: 264 seconds]
jtojnar_ is now known as jtojnar
__Sander__ has quit [Ping timeout: 260 seconds]
kmicklas has joined #nixos-dev
<globin>
niksnut: packages-unstable.json on the website isn't updating again
<niksnut>
probably because the channel is broken?
phreedom has quit [Ping timeout: 268 seconds]
gleber_ has quit [Quit: ~]
phreedom has joined #nixos-dev
gleber_ has joined #nixos-dev
cbarrett has quit [Quit: ~]
cbarrett has joined #nixos-dev
elvishjerricco has quit [Quit: ~]
elvishjerricco has joined #nixos-dev
terrorjack has quit [Quit: ~]
vdemeester` has quit [Quit: ~]
vdemeester` has joined #nixos-dev
terrorjack has joined #nixos-dev
<globin>
niksnut: unstable updated 5 days ago, 18.03 2 days ago, but the json is from Mar 24
{^_^} has quit [Remote host closed the connection]
{^_^} has joined #nixos-dev
{^_^} has quit [Changing host]
{^_^} has joined #nixos-dev
<Sonarpulse>
Dezgeg: yeah i guess staging is jammed full of the automatic updates?
<Sonarpulse>
shlevy: so at the very least we need the intensional store because Tweag using Nix + Bazel is rightfully an embarrassment for Nix :D
<shlevy>
:P
jtojnar_ has joined #nixos-dev
<Sonarpulse>
shlevy: and sometime (after I get lunch at least) I need to explain my very convoluted idea for drvs for delayed eval for inport for derivation along with drvs than produce drvs
<Sonarpulse>
for perfect --dry-run
<shlevy>
Sonarpulse: :) We should probably have a call at some point to sync up convoluted ideas
<Sonarpulse>
hehe yeah
<Sonarpulse>
drvs that produce drvs is sort of my idea for a recursive nix alternative
jtojnar has quit [Ping timeout: 256 seconds]
jtojnar_ is now known as jtojnar
<shlevy>
You should talk to taktoa. I tried to convince him we couldn't avoid recursive nix altogether and I think he's now on board after a summer trying to work around it :P
<shlevy>
But yeah drvs that produce drvs (and then rewrite up the chain) is fundamentally my idea for recursive nix
<shlevy>
Where you split the phases can be configurable in principle
<Sonarpulse>
shlevy: oh good
<Sonarpulse>
yeah drvs that produce drvs mean no blocking on other builds
<Sonarpulse>
a blocking which I'd fine very conceptually suspect and probably not good for our infra anyways
<shlevy>
Hmm say more?
<shlevy>
What do you mean by blocking here?
<Sonarpulse>
if you nix-build in a nix-build
<Sonarpulse>
then you are like a non-top call stack frame waiting for that build to finish
<Sonarpulse>
drvs producing drvs is more cps
<Sonarpulse>
you make your continuation drv
<shlevy>
Ah yeah. My design won't disallow that in principle but it's not a nice way to do it :)
<shlevy>
And you don't get the nice rewriting properties
<Sonarpulse>
rewriting?
<shlevy>
So, obviously we're not to far away from a world where glibc uses recursive nix instead of Make. Should you rebuild the world because someone added a comment in nixpkgs?
<Sonarpulse>
oh yeah intensional store plus drv->drv helps with that
<shlevy>
Don't need intensional store
<Sonarpulse>
for me they are very related
<Sonarpulse>
in that "forwarding" one drv to another
<Sonarpulse>
is just like substituting drvs hash of deps with actual builds
<Sonarpulse>
ok g2g food
<shlevy>
a.drv has b.drv as an input. b.drv replaces itself with b'.drv. a.drv gets rewritten to a'.drv that depends on b'.drv. Everything's already built, everyone is happy
jtojnar has quit [Quit: jtojnar]
pie__ has quit [Read error: Connection reset by peer]
pie_ has joined #nixos-dev
goibhniu has quit [Ping timeout: 264 seconds]
pxc has joined #nixos-dev
pxc has quit [Ping timeout: 260 seconds]
<Sonarpulse>
shlevy: right that sort of input substitution is shared with intensional store
<zybell_>
in postgresql.nix you generate a file and mark it as Generated File (not your change, was before) On #guix a user couldn't figure out why edits were missing. So I think: while your at it, could you please insert that the file is Generated by postgresql.nix (with full path)? That could help such users.
<Sonarpulse>
cstrahan: it seems odd to me that we are subtracting lists twice?
<Sonarpulse>
I mean probably not a big deal
<cstrahan>
Sonarpulse: in one case we're checking if the flags aren't in the set of those we expect, and in the other we're removing the disabled flags from the (default + explicitly enabled) flags.
<cstrahan>
If we can do that in one go, that would be cool, but I don't see how we can do that.
<cstrahan>
At any rate, 99% of the time a package will leave things at their defaults, in which case both expressions are never evaluated.
<Sonarpulse>
cstrahan: yeah
<Sonarpulse>
I think I'm about to say worry about it post merge :)
<Sonarpulse>
let me read once more for <1 minute :)
<cstrahan>
Correction: erroneousHardeningFlags will still be evaluated, but it'll be O(1) (or at least it should be -- if the second operand is [], it should immediately return [])
<Sonarpulse>
cstrahan: oops somebody else pinged me
<Sonarpulse>
just merge it :)
<Sonarpulse>
I'll approve
<Sonarpulse>
cstrahan: btw, and now worries if you are totally exhausted with *-wrapper, but would you be interested in helping abstract things?
<Sonarpulse>
I don't regret splitting cc-wrapper and binutils-wrapper
<Sonarpulse>
but I also want to add a pkg-config-wrapper
<Sonarpulse>
and well there'd be decent duplication of things then (though not hardening)
<Sonarpulse>
shlevy: yeah especially with all these automatic updates I really think we need separate staging for refactors and version bumps
<Sonarpulse>
refactors are more likely to cause superficial breakage, but version bumps cause deeper breakage when they do
<Sonarpulse>
cstrahan: you're approved, but I can merge too if you like
<zybell_>
in postgresql.xml a comment start is removed, but not the end (can be missing from start,can be missed by me)
<zybell_>
that was all I found
<Sonarpulse>
cstrahan: oh wait found another thing