01:28
phreedom has quit [Ping timeout: 250 seconds]
01:29
phreedom has joined #nixos-dev
01:56
pie_ has quit [Ping timeout: 276 seconds]
02:55
lassulus_ has joined #nixos-dev
02:58
lassulus has quit [Ping timeout: 256 seconds]
02:58
lassulus_ is now known as lassulus
05:07
lopsided98 has quit [Ping timeout: 276 seconds]
05:14
lopsided98 has joined #nixos-dev
07:19
__Sander__ has joined #nixos-dev
09:49
jtojnar has quit [Read error: Connection reset by peer]
09:51
jtojnar has joined #nixos-dev
10:26
jtojnar has quit [Quit: jtojnar]
10:28
pie_ has joined #nixos-dev
11:19
jtojnar has joined #nixos-dev
11:32
pie__ has joined #nixos-dev
11:36
pie_ has quit [Ping timeout: 256 seconds]
12:40
<
infinisil >
gchristensen: git diff outputs non-zero when there's changes? Can't reproduce
12:41
<
infinisil >
Even with -q
12:41
<
infinisil >
Does it work for you?
12:42
<
gchristensen >
it seems to work ...
12:42
<
gchristensen >
one sec
12:42
<
infinisil >
$? is always zero for me, with or without changes
12:44
<
gchristensen >
oops
12:44
<
gchristensen >
somebug there, -q implies --exit-code
12:44
<
gchristensen >
but it isn't working
12:45
<
infinisil >
--exit-code does work for me
12:45
<
gchristensen >
fixed :)
12:45
<
gchristensen >
I think git has a bug then?
12:47
<
infinisil >
Oh right, I often have a dirty submodule, which I can't commit (git add -A wouldn't), but your way would still error out
13:23
Sonarpulse has quit [Ping timeout: 276 seconds]
13:31
mpickering has joined #nixos-dev
13:32
<
gchristensen >
I find myself habitually never committing to my config repos, so hopefully this fixes it:)
13:43
<
gchristensen >
haha
13:55
<
gchristensen >
I have a derivation which, when running `time nix-shell --run true` takes 5min04s. what sort of things should I be doing to figure out why it took so long? I'm not familiar at all with examining the performance of Nix stuff
13:56
<
gchristensen >
s/derivation/shell.nix/
13:56
<
niksnut >
wow that's a lot
13:56
<
clever >
gchristensen: how long does it take to nix-instantiate the shell.nix?
13:56
<
niksnut >
do you have a giant source tree that's copied every time?
13:59
<
gchristensen >
(checking on both questions) in other shell.nix's which take a long time (30s-1min), no I'm not copying a giant source tree but it is possible this one is
14:00
<
niksnut >
also try adding -vvvvv to the nix-instantiate to see what it's doing
14:03
<
gchristensen >
I hate having to use two laptops :')
14:04
<
orivej >
"NIX_DEBUG=6 nix-shell" might also tell something
14:06
<
niksnut >
what is NIX_DEBUG=6?
14:07
<
clever >
if the problem is on the nix side, that will have no effect
14:07
<
clever >
nix-instantiate alone confirms if the problem is on the nix side
14:10
guibou has joined #nixos-dev
14:10
<
fadenb >
don't act so surprised. In the nix world everything is perfectly documented ;)
14:10
<
gchristensen >
fadenb++
14:10
<
{^_^} >
fadenb's karma got increased to 1
14:11
<
gchristensen >
dang, I can't replicate the problem :/
14:12
<
Dezgeg >
maybe you just had a slow binary cache connection or something?
14:12
<
gchristensen >
I replicated it once, and another coworker can replicate it easier
14:13
<
gchristensen >
maybe so
14:13
<
gchristensen >
I'll put -vvvvvv and NIX_DEBUG=6 in to my back pocket for when they come online again. thanks niksnut / clever / orivej / Dezgeg
14:14
<
clever >
,profiling
14:14
<
{^_^} >
Use NIX_COUNT_CALLS=1 and/or NIX_SHOW_STATS=1 to profile Nix evaluation
14:14
<
clever >
gchristensen: and this
14:15
<
gchristensen >
oh nice
14:16
<
gchristensen >
1,126 calls to scopedImport!
14:17
<
clever >
oh, of note, nix will cache any call to normal import
14:17
<
clever >
it assumes that import returns the value value, given the same path
14:17
<
clever >
but scopedImport disables that cache
14:17
<
gchristensen >
I'm not using scopedImport! `rg scopedImport` returns no results except within Nix
14:17
<
clever >
oh, and also, import is just an alias for `scopedImport {}`
14:18
<
gchristensen >
It Would Be Cool (tm) if there was a builtins.performanceTick thing which would let me add an impure performance ticker for profiling
14:18
<
clever >
gchristensen: line 2152-2153 creates an empty attr set, line 2154 applies it to scopedImport, and then saves the partially applied in import
14:21
* gchristensen
searches for how to do things he has no business doing
14:34
<
gchristensen >
ah, here is one which takes 41s
14:58
ixxie has quit [Ping timeout: 248 seconds]
15:16
__Sander__ has quit [Ping timeout: 255 seconds]
15:17
__Sander__ has joined #nixos-dev
15:29
__Sander__ has quit [Quit: Konversation terminated!]
15:52
bgamari has quit [Ping timeout: 240 seconds]
16:00
bgamari has joined #nixos-dev
16:15
Lisanna has quit [Quit: Lisanna]
16:32
Lisanna has joined #nixos-dev
17:01
Lisanna has quit [Quit: Lisanna]
17:03
Lisanna has joined #nixos-dev
17:06
Lisanna has quit [Client Quit]
17:30
phreedom has quit [Remote host closed the connection]
17:31
phreedom has joined #nixos-dev
17:35
Sonarpulse has joined #nixos-dev
17:36
goibhniu has quit [Ping timeout: 268 seconds]
17:52
goibhniu has joined #nixos-dev
18:17
Lisanna has joined #nixos-dev
18:17
Sonarpulse has quit [Ping timeout: 265 seconds]
19:00
bgamari has joined #nixos-dev
19:30
lassulus has quit [Ping timeout: 256 seconds]
19:40
Sonarpulse has joined #nixos-dev
19:43
lassulus has joined #nixos-dev
20:42
jtojnar_ has joined #nixos-dev
20:43
jtojnar has quit [Ping timeout: 240 seconds]
20:43
jtojnar_ is now known as jtojnar
22:19
orivej has quit [Ping timeout: 256 seconds]