!djTaTBQyWEPRQxrPTb:nixos.org

Nixpkgs Architecture

216 Members
Discussions about Nixpkgs' architecture - https://github.com/NixOS/nixpkgs/labels/architecture47 Servers

Load older messages


SenderMessageTime
5 Mar 2024
@nbp:mozilla.orgnbp Robert Hensing (roberth): Thanks for opening #273815 , what you suggest about using the update operator (//) reminds me of an old proposal named S.O.S. ;) 14:19:00
@tomberek:matrix.orgtomberekI'm not available this week.14:19:44
@nbp:mozilla.orgnbp I think the mkDerivation application (calling with it) should be moved out of the fix-point if we want to keep use the update operator or any derivative of it. 14:19:52
@nbp:mozilla.orgnbp Thus one would set builder = stdenv.mkDerivation; recipe = { /* mkDerivation argument */ }; and a final phase will go after the fix-point to apply the builder to it argument. 14:21:20
@nbp:mozilla.orgnbp The same can be considered for migrating callPackage / mkScope resolution. 14:22:08
@dmills27:matrix.orgDominic Mills joined the room.14:22:19
@nbp:mozilla.orgnbpbut we might need an update operator which works across functions, which it currently does not.14:22:33
@cab404:matrix.orgcab404 hmm, still can't add https://calendar.google.com/calendar/u/0/embed?src=b9o52fobqjak8oq8lfkhg3t0qg@group.calendar.google.com&ctz=Europe/Zurich as a calendar 15:06:24
@cab404:matrix.orgcab404 Ohh, found it https://calendar.google.com/calendar/ical/b9o52fobqjak8oq8lfkhg3t0qg%40group.calendar.google.com/public/basic.ics 15:09:31
@roberthensing:matrix.orgRobert Hensing (roberth)
In reply to @nbp:mozilla.org
Thus one would set builder = stdenv.mkDerivation; recipe = { /* mkDerivation argument */ }; and a final phase will go after the fix-point to apply the builder to it argument.

Where builder and recipe are attributes of the overridable fixpoint, I suppose?
With something like lib.encapsulate, that would be

lib.encapsulate (this: {
  builder = stdenv.mkDerivation;
  recipe = { name = "hello"; ..... };
  public = this.builder this.recipe;
})
16:41:02
@roberthensing:matrix.orgRobert Hensing (roberth)I'm not sure about the names, as a derivation builder is also the process that runs as the derivation, and recipe is a bit vague, but I suppose naming isn't the primary concern16:42:03
@roberthensing:matrix.orgRobert Hensing (roberth) Also worth noting that instead of mkDerivation it should be a function that does most of those things, but no overrideAttrs. That must be handled by whatever goes into lib.encapsulate; we'll want a helper function make this kind of package. Can't ask packagers to write the above. 16:44:05
@nbp:mozilla.orgnbp Yes, where there is no function application, and where lib.encapsulate would be called on anything which is typed as a package. 17:00:55
@jade_:matrix.org@jade_:matrix.org
In reply to @philiptaron:matrix.org
At work, we use a approve-commits model, instead of an approve-PR model, which makes the review process substantially lighter. I'm sad that GitHub doesn't let that happen, since I only get the chance to review the whole squashed PR.
this kinda exists with the various pr stacking tools like pkgs.sapling and many others but erm. what if nixpkgs had gerrit, (do not use gerrithub btw it is very broken)
17:49:00
@philiptaron:matrix.orgPhilip Taron (UTC-8)
In reply to @jade_:matrix.org
this kinda exists with the various pr stacking tools like pkgs.sapling and many others but erm. what if nixpkgs had gerrit, (do not use gerrithub btw it is very broken)
We also use and pay for Graphite but it's only OK.
17:50:38
@jade_:matrix.org@jade_:matrix.orgyeah17:51:34
@jade_:matrix.org@jade_:matrix.orgit's kinda bad lol17:51:39
@jade_:matrix.org@jade_:matrix.orgi have been working on gerrit recently on a private project and it has been nice17:52:05
@philiptaron:matrix.orgPhilip Taron (UTC-8)I've used Phabricator for the last decade and TBH I love it. https://we.phorge.it/ is the successor.17:52:36
@nscnt:matrix.org@nscnt:matrix.org left the room.18:32:57
6 Mar 2024
@sammy:cherrykitten.dev@sammy:cherrykitten.dev joined the room.19:04:52
@ktemkin:katesiria.org@ktemkin:katesiria.org joined the room.22:28:08
8 Mar 2024
@infinisil:matrix.orginfinisil@room The architecture team is being dissolved, please check out https://discourse.nixos.org/t/nixpkgs-architecture-team-conclusion-and-prospective/41020 for all the details!16:55:26
@philiptaron:matrix.orgPhilip Taron (UTC-8)

at some point somebody should just be empowered to make decisions.
♥️

16:56:13
@philiptaron:matrix.orgPhilip Taron (UTC-8) *

at some point somebody should just be empowered to make decisions.
♥️ to this

16:56:30
@philiptaron:matrix.orgPhilip Taron (UTC-8)

I'm a little sad that this action is being taken:

Publicly archive the private Nixpkgs Architecture Matrix channel

... as having "direct line" to those interested in architecture and structure rather than the things being built with the structure is quite nice.

16:57:45
@nbp:mozilla.orgnbpI agree that keeping this channel might still be valuable.16:58:10
@qyliss:fairydust.spaceAlyssa RossThis room is being kept16:58:15
@qyliss:fairydust.spaceAlyssa Rossthe "private Nixpkgs Architecture Matrix channel" is not this room16:58:24
@philiptaron:matrix.orgPhilip Taron (UTC-8)
In reply to @qyliss:fairydust.space
the "private Nixpkgs Architecture Matrix channel" is not this room
👀 I see
16:58:44

Show newer messages


Back to Room ListRoom Version: 9