<Enzime>
dtz: the patch also includes removing --as-needed
<Enzime>
should that stay macOS only?
<dtz>
I think so-- doesn't seem to "hurt" re:llinux so seems should leave build-system as-is
<dtz>
although there's a fair bit of laziness masquerading as wisdom in that notion ;)
<Enzime>
dtz: which notion are we talking about
<dtz>
nevermind, sorry.
<dtz>
I think the "--as-needed" part should only be changed on Darwin
<dtz>
but that's because that seems the "safe" answer re:not changing/breaking current behavior
<Enzime>
dtz: I would argue that the 'safest' behaviour is to leave as much as possible as given by the source code and change the environment we have to make it work
<dtz>
yes! That's why I'm suggesting only change it where needed
<dtz>
but if you're asking re:errno I feel very different on that front-- I'm more confident of that change and even if it somehow DID "break" something it's hard to argue against "include headers for the things you use" and AFAIK everyone agrees errno.h should be included when accessing errno xD
<dtz>
but! at the end of the day that's just my (hopefully reasonable and somewhat informed :P) opinion, it's a judgement call.
<dtz>
But I'm not sure where you stand on these things just thought I'd preempt that :)
<dtz>
(although IMO any more technical discussion and this should probably relocate to the PR for visibility0
sir_guy_carleton has joined #nixos-dev
phreedom has quit [Ping timeout: 250 seconds]
sir_guy_carleton has quit [Quit: WeeChat 2.0]
phreedom has joined #nixos-dev
sir_guy_carleton has joined #nixos-dev
orivej has quit [Ping timeout: 240 seconds]
sir_guy_carleton has quit [Quit: WeeChat 2.0]
<samueldr>
niksnut, gchristensen, on #nixos, if the user hasn't fixed their issue meanwhile, setting /mode #nixos +b seafood!*@14-203-180-240.tpgi.com.au and contacting them may be necessary, due to probable server issue / IRC client issue
<niksnut>
what does that do?
<niksnut>
oh, ban?
<samueldr>
yes
<samueldr>
which is why contacting the user through private message may be needed so they know why
<samueldr>
(you will need to either do the ban through chanserv or op yourself through chanserv beforehand)
<samueldr>
oh, this chanserv implementation doesn't seem to be able to handle setting modes / banning
<andi->
Chanserv is able to do that (here)
<niksnut>
maybe we should disable join/leave messages?
<adisbladis[m]>
That's up to each client
<niksnut>
hm, I thought there was a channel setting for that as well
<andi->
Nah, just hide them in your client.
<andi->
Weechat has nice settings for that (called smart filters) that will show/hide them depending on recent activity. You will see when someone disconnected you were just talking to but none of the zombie join/quits will show up.
goibhniu has joined #nixos-dev
obadz has joined #nixos-dev
Lisanna has quit [Quit: Lisanna]
orivej has joined #nixos-dev
Lisanna has joined #nixos-dev
jtojnar has joined #nixos-dev
orivej has quit [Ping timeout: 240 seconds]
phreedom has quit [Ping timeout: 250 seconds]
phreedom has joined #nixos-dev
__Sander__ has joined #nixos-dev
pie_ has joined #nixos-dev
Lisanna has quit [Remote host closed the connection]
<makefu>
what kind of sourcery is this!? btw, i've solved my issue with the intercepting proxy. it seemed that i have to provide the `whole certificate chain` with the latest curl instead of just providing the intermediate certificates.
<niksnut>
gchristensen: probably because of the eval out-of-memory failures...
<gchristensen>
ouch :/
Lisanna has joined #nixos-dev
<gchristensen>
it was so sad seeing this new machine sit idle so I restarted some failed builds
<gchristensen>
so the bittorrent test will either take like 60s or many minutes depending on some race condition at the start
ixxie has quit [Ping timeout: 268 seconds]
Sonarpulse has joined #nixos-dev
sir_guy_carleton has joined #nixos-dev
<gchristensen>
I don't like overlays
<niksnut>
why not?
<gchristensen>
it is difficult-to-impossible to add them to my system and trust they're only modifying a specific subset of the packages
<gchristensen>
in other words, all the problems of the lack of POLA for the module system, but for packages
<niksnut>
like a NixOS module :-)
<gchristensen>
7:D
<gchristensen>
exactly
<gchristensen>
so, now, there is talk about removing packageOverrides and that seems bad
<gchristensen>
also -- it is impossible to extend an existing imported nixpkgs with an overlay, you must re-import nixpkgs to do so
<niksnut>
well, doesn't packageOverrides have the same problem?
<gchristensen>
hmm yeah I suppose if I was adding potentially arbitrary repos to it
<shlevy>
gchristensen: If you could extend an existing nixpkgs with an overlay, wouldn't that obviate your first concern?
<shlevy>
Since you can just locally override it for whatever package the overlay provides?
<gchristensen>
sure, that would make me happy enough, since I could do an overlay with { garbas-pythonpkgs-overlay = pkgs.extend_with_overlays [ .. ]; } and then pkgs.garbas-pythonpkgs-overlay.foobar...
<shlevy>
That would be trivial to add, though I'm not suree off-hand if it would be painfully inefficient or anything
<gchristensen>
my understanding from pierron is that it isn't trivial to add
<shlevy>
gchristensen: Sure it is. You can just have it do the reimport for you
<shlevy>
all-packages.nix knows the original arguments you passed and the path to itself, that's all you need
<gchristensen>
ah, I meant, not trivial to add without a large increase in memory usage :)
<shlevy>
I wonder if it would help to somehow stage nixpkgs, so you could get an uncomposed set evaluated once then apply arbitrary composition only where it's needed
<shlevy>
Part of the problem is on the current idioms we have absolutely no way to distinguish inputs that are parameters from inputs that are dependencies, and it only gets worse where we have multiple layers of functions, each layer having mixed inputs and parameters
<shlevy>
So it's really hard to do better
<domenkozar>
similarly pierron wanted to fix memory usage problems, callPackage would store initial filename and then .overide would reimport it with correct inputs
<domenkozar>
instead of stacking overrides on top of each other
<domenkozar>
also helps with tracking what did what
<Profpatsch>
Ah, I see. You have a pull request. :P
<gchristensen>
cool
<Profpatsch>
I can’t select a certain platform, right?
<Profpatsch>
But it should be a no-op on the other platforms anyway.
<gchristensen>
right, you can't
orivej has quit [Ping timeout: 244 seconds]
<Enzime>
LnL: is the correct way to fix "fatal error: 'CoreServices/CoreServices.h' file not found" to add darwin.apple_sdk.frameworks.CoreServices to buildInputs?
erickomoto has joined #nixos-dev
pie_ has quit [Ping timeout: 256 seconds]
<gchristensen>
the macs have churned through like 8k builds today, so ... thats good
pie_ has joined #nixos-dev
<dtz>
\o/
sir_guy_carleton has quit [Quit: WeeChat 2.0]
pie_ has quit [Ping timeout: 260 seconds]
erickomoto has quit [Ping timeout: 240 seconds]
cransom has quit [Quit: WeeChat 2.0]
cransom has joined #nixos-dev
FRidh has quit [Quit: Konversation terminated!]
pie_ has joined #nixos-dev
erickomoto has joined #nixos-dev
erickomoto has quit [Client Quit]
erickomoto has joined #nixos-dev
goibhniu1 has joined #nixos-dev
goibhniu has quit [Ping timeout: 265 seconds]
Lisanna has quit [Remote host closed the connection]
erickomoto has quit [Ping timeout: 265 seconds]
Lisanna has joined #nixos-dev
erickomoto has joined #nixos-dev
phreedom has quit [Ping timeout: 250 seconds]
phreedom has joined #nixos-dev
erickomoto has quit [Ping timeout: 240 seconds]
mbrgm has joined #nixos-dev
ixxie has quit [Ping timeout: 268 seconds]
mbrgm has left #nixos-dev ["WeeChat 2.1"]
orivej has joined #nixos-dev
infinisil has quit [Quit: Configuring ZNC, sorry for the joins/quits!]
infinisil has joined #nixos-dev
Sonarpulse has quit [Ping timeout: 240 seconds]
genesis has quit [Read error: Connection reset by peer]
erickomoto has joined #nixos-dev
erickomoto has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]