<niksnut>
worldofpeace: hm, but that will break > 20.03
<niksnut>
maybe just check for 19. ?
<worldofpeace>
a redirect would put this to rest 😺
<worldofpeace>
niksnut: pushed. my first perl if 🎊
<niksnut>
thanks :-)
<gchristensen>
afaict, after applying `patches`, patched source files in a build have timestamps different from the rest of the files. I wonder if we should reset those timestamps after the patch phase
__monty__ has quit [Quit: leaving]
<niksnut>
it should be deterministic at least, for the benefit of SOURCE_DATE_EPOCH
<gchristensen>
+1
<niksnut>
btw patches can contain timestamps, but git diffs don't
ixxie has quit [Ping timeout: 268 seconds]
<gchristensen>
right
<gchristensen>
it sounds like for that we would want to pass -Z to `patch`
ixxie has joined #nixos-dev
psyanticy has joined #nixos-dev
FRidh has joined #nixos-dev
<qyliss>
Do I need a release note for changes that are already documented with mkRenamedOptionModule/mkRemovedOptionModule?
<gchristensen>
I think it would be best, yeah, so people know to expect it
<gchristensen>
and the release note could have more context
kenjis1 has joined #nixos-dev
FRidh has quit [Quit: Konversation terminated!]
<gchristensen>
what if command-not-found outputted the package's description for each result? it might help people pick an option if there are many results for a package
<gchristensen>
for example `gcc` has many results, and `ar` has two results and neither of them are obviously right. (these are just two examples I could come up with on short notice, please don't try to pick apart the idea based on the two examples :P)
<jtojnar>
the main issue with these would be that the descriptions are likely the same
<gchristensen>
ouch
<gchristensen>
right
<clever>
ive also seen a number of people break their machine by installing busybox
<gchristensen>
maybe they shouldn't be! overridden to have a different description, describing why they exist. this sounds like a UX improvement in a lot of ways!
<clever>
it overrides a dozen system utils, some in non-compatible ways
<{^_^}>
[ "GNU Compiler Collection, version 4.8.5 (wrapper script)" "GNU Compiler Collection, version 4.9.4 (wrapper script)" "GNU Compiler Collection, version 6.5.0 (wrapper script)" "GNU Compiler Collection...
<qyliss>
Presumably it breaks activation scripts?
<jtojnar>
interesting, gcc actually has proper decriptions
<das_j>
hm, is it possible to pass multiple configurations to the <nixos/nixpkgs> configuration parameter?
<qyliss>
to have them merged together?
<das_j>
yes
<qyliss>
just import them all from one top-level configuration
<das_j>
oh yeah, that makes a lot more sense
<das_j>
thanks
Jackneill has quit [Read error: Connection reset by peer]
orivej has quit [Ping timeout: 240 seconds]
orivej has joined #nixos-dev
<aanderse>
qyliss: was it you who went through a whole postfix setup recently and mentioned as such?
orivej has quit [Ping timeout: 260 seconds]
lovesegfault has joined #nixos-dev
<qyliss>
I'm using Postfix with Mailman, and Postfix as an SMTP relay on my laptop, and have been since last year
<qyliss>
So probably not?
<aanderse>
ah ok
_ris has joined #nixos-dev
orivej has joined #nixos-dev
psyanticy has quit [Quit: Connection closed for inactivity]
<yorick>
should .a's be in .dev or .lib?
<clever>
yorick: i would put them in .dev
<clever>
if something depends on $lib/lib/libfoo.so, i dont want a dead libfoo.a hanging around eating up space
kenjis1 has quit [Remote host closed the connection]
lovesegfault has quit [Ping timeout: 248 seconds]
lovesegfault has joined #nixos-dev
<qyliss>
> Please note that NixOS assumes all over the place that shell to be Bash, so override the default setting only if you know exactly what you're doing.
<{^_^}>
error: syntax error, unexpected ',', expecting ')', at (string):277:72
<qyliss>
Is this really true? I'd be surprised if anything in NixOS used a /bin/sh shebang rather than one pointing into the store
<gchristensen>
a good question, let's try it
jonge[m] has quit [*.net *.split]
Ox4A6F has quit [*.net *.split]
domenkozar[m] has quit [*.net *.split]
ericnoan has quit [*.net *.split]
<qyliss>
I haven't been brave enough to in case I was missing something obvious
ericnoan has joined #nixos-dev
<gchristensen>
I can try it :)
<gchristensen>
this is a favorite feature of my / being erased continuously
<qyliss>
ahh of course :D
<gchristensen>
one time I turned on a lot of the systemd security features by default for every unit just to see ... it didn't work well
<samueldr>
would a `meta.vcs` attribute be a welcome change? where a url can be used to point to the source location of a package
<samueldr>
welcome addition*
<samueldr>
I've just had an annoying time tracking down where a specific software's sources were
<samueldr>
though, thinking about it I think its value would be greater if it could also encode the protocol to use, in addition to the URL, somehow
<samueldr>
(thinking about how both git and mercurial can use https:// as protocol)
kenjis1 has joined #nixos-dev
orivej has quit [Ping timeout: 260 seconds]
<jtojnar>
not sure if nixpkgs should mirror homepages
<jtojnar>
if vcs repo is hard to find, it is a website bug
<lovesegfault>
I wish `metadata` had a field for a "release feed"
<lovesegfault>
or page
<samueldr>
there is one for changelog, lovesegfault