<{^_^}>
#57770 (by shirona, 2 days ago, open): Vulkan applications no longer work
<arianvp>
I'd love to get vulkan working again but not sure where to go from here
<arianvp>
Something regressed in MESA
init_6 has joined #nixos-dev
jtojnar has quit [Remote host closed the connection]
timokau has quit [Ping timeout: 245 seconds]
drakonis1 has quit [Ping timeout: 246 seconds]
drakonis1 has joined #nixos-dev
jtojnar has joined #nixos-dev
johanot has joined #nixos-dev
orivej has quit [Ping timeout: 272 seconds]
johanot has quit [Quit: WeeChat 2.4]
ajs124 has joined #nixos-dev
jtojnar has quit [Read error: Connection reset by peer]
jtojnar has joined #nixos-dev
pbogdan has left #nixos-dev ["ERC (IRC client for Emacs 26.1)"]
<gchristensen>
FYI: RFC 34, Expression Integrity has moved in to Final Comment Period with a recommendation of Rejection (with many more specifics -- please read this reply before replying on IRC) https://github.com/NixOS/rfcs/pull/34#issuecomment-474404580
init_6 has quit [Ping timeout: 244 seconds]
init_6 has joined #nixos-dev
asymmetric has joined #nixos-dev
orivej has joined #nixos-dev
asymmetric has quit [Remote host closed the connection]
asymmetric has joined #nixos-dev
timokau has joined #nixos-dev
<simpson>
gchristensen: Since nobody else so far has said it: Thank you for a thoughtful endcap to what was overall a remarkably productive discussion.
<simpson>
gchristensen++
<{^_^}>
gchristensen's karma got increased to 80
init_6 has quit []
<gchristensen>
thank you, simpson :) I was anxious about how this RFC would resolve, I hope to find the FCP is indeed quiet with general agreement
timokau has quit [Ping timeout: 258 seconds]
asymmetric has quit [Ping timeout: 268 seconds]
drakonis has joined #nixos-dev
timokau has joined #nixos-dev
timokau has quit [Ping timeout: 246 seconds]
asymmetric has joined #nixos-dev
asymmetric has quit [Remote host closed the connection]
asymmetric has joined #nixos-dev
orivej has quit [Ping timeout: 245 seconds]
Jackneill has quit [Ping timeout: 255 seconds]
asymmetric_ has joined #nixos-dev
asymmetric_ has quit [Read error: Connection reset by peer]
asymmetric has quit [Ping timeout: 264 seconds]
Jackneill has joined #nixos-dev
jtojnar_ has joined #nixos-dev
jtojnar has quit [Ping timeout: 268 seconds]
jtojnar_ is now known as jtojnar
orivej has joined #nixos-dev
timokau has joined #nixos-dev
timokau has quit [Ping timeout: 272 seconds]
timokau has joined #nixos-dev
jtojnar has quit [Ping timeout: 250 seconds]
timokau has quit [Ping timeout: 244 seconds]
JosW has joined #nixos-dev
timokau has joined #nixos-dev
jtojnar has joined #nixos-dev
timokau has quit [Ping timeout: 268 seconds]
timokau has joined #nixos-dev
jtojnar_ has joined #nixos-dev
jtojnar has quit [Ping timeout: 246 seconds]
jtojnar_ is now known as jtojnar
timokau has quit [Ping timeout: 246 seconds]
jtojnar has quit [Read error: Connection reset by peer]
timokau has joined #nixos-dev
timokau has quit [Ping timeout: 246 seconds]
<tilpner>
Any opinions on a libfaketime LD_PRELOAD hook that use SOURCE_DATE_EPOCH?
<tilpner>
That seems easier to do than finding where man-db sets the date
<gchristensen>
is this for r13y?
<tilpner>
Yes
<tilpner>
I'm about to give up if I'm not allowed to use faketime though
<gchristensen>
hmm
disasm has quit [Quit: WeeChat 2.0]
<tilpner>
It also seems more robust than hunting and patching every last date substitution
<tilpner>
So I would have expected this to already exist
<gchristensen>
interesting -- debian's is reproducible already but don't have a patch
<tilpner>
Their default build process could fake the system time
<tilpner>
Which IMO stdenv should do too, if feasible
<gchristensen>
it might be ok on linux, but wouldn't work for macos
* gchristensen
pokes around the man-db source
<tilpner>
Hmm, right
<tilpner>
But this is for a NixOS ISO
<tilpner>
While I agree that it would be nice if it could be built deterministically from Darwin, it's of far lower priority to me
<simpson>
tilpner, gchristensen: Noticing a pattern in testing (including r13y) where forcing a value to be always 1 is nice, but being able to sweep it through a range of values means that you've created a parameterized property test, ala Hypothesis.
<gchristensen>
one technique I should write up is diffing two build's tmp dirs
<gchristensen>
ie: `nix-build '<nixpkgs>' -A man_db --check -K; nix-build '<nixpkgs>' -A man_db --check -K` then `diffoscope /tmp/nix-build-man-db-2.7.5.drv-1 0tmp/nix-build-man-db-2.7.5.drv-1`
<simpson>
gchristensen: To make this a secant, one high-level example (that I think you've already got) is sweeping through multiple SOURCE_DATE_EPOCH choices and marking out where things change, giving a basic map of one kind of unr13y.
<gchristensen>
ah!
<tilpner>
simpson: That only works if things respect SOURCE_DATE_EPOCH, right?
<tilpner>
Which in this case they don't
<simpson>
Or more generally viewing diffoscope as an explorer for a multicolored map of the different varieties of unr13'able behavior within a package.
<simpson>
tilpner: Yeah. I'm saying that we can map out *how* they don't.
<simpson>
Anyway, just a high-level thought that I felt that I had to share in the moment.