!aGqRytqbCECitOFhbt:nixos.org

Release Management

341 Members
25.11 "Xantusia" | https://nixos.github.io/release-wiki/Home.html94 Servers

Load older messages


SenderMessageTime
21 May 2022
@vcunat:matrix.orgVladimír ČunátAh right. I have the "Infra Team" term associated with a much narrower set of people (and corresponding private room).20:26:22
@janne.hess:helsinki-systems.deJanne Heß
In reply to @vcunat:matrix.org
Ah right. I have the "Infra Team" term associated with a much narrower set of people (and corresponding private room).
Ah, I don't think I can expect every release manager to be part of that private room ;)
20:27:14
@vcunat:matrix.orgVladimír ČunátThe room you link is much better for this.20:27:48
@vcunat:matrix.orgVladimír ČunátI'm not on the infra team itself, but I can create jobsets (in nixpkgs and nixos projects). Though I'm not sure how many other such people exist.20:28:32
@janne.hess:helsinki-systems.deJanne HeßProbably not too many20:29:11
@vcunat:matrix.orgVladimír ČunátRelease managers commonly do get restart/cancel privileges at least, right?20:31:31
@janne.hess:helsinki-systems.deJanne Heß
In reply to @vcunat:matrix.org
Release managers commonly do get restart/cancel privileges at least, right?
Yeah, I have everything except admin and create-projects
20:31:53
@janne.hess:helsinki-systems.deJanne HeßSo also the bump privilege20:32:02
@vcunat:matrix.orgVladimír ČunátRight, that's relevant.20:32:48
@janne.hess:helsinki-systems.deJanne Heß Vladimír Čunát: about your comment: is there any reason to lower the scheduling shares rather than just disabling the jobset altogether? 20:47:31
@vcunat:matrix.orgVladimír ČunátPeople still push to them sometimes.20:48:19
@janne.hess:helsinki-systems.deJanne Heßummm what?20:48:26
@vcunat:matrix.orgVladimír Čunát

Say

commit + 1a7c2f41c3bf6d7fd443e2453e7b840cdc413375 (h/release-21.05, h/nixpkgs-21.05-darwin, h/nixos-21.05-aarch64)
Author:     Robert Hensing <robert@roberthensing.nl>
AuthorDate: Thu Apr 7 11:05:28 2022 +0200
Commit:     Robert Hensing <robert@roberthensing.nl>
CommitDate: Mon May 2 09:51:01 2022 +0200

    pkgs: Add _type = "pkgs"
    
    (cherry picked from commit ad1e2500efd0aa49b0dc7427bf69d4879f3b0ff5)
    (cherry picked from commit aec730a0af4c977513ce28236cbecaca72af6901)
20:49:04
@vcunat:matrix.orgVladimír ČunátThat's to 21.05 just several weeks ago.20:49:21
@vcunat:matrix.orgVladimír Čunát * That's to 21.05 just three weeks ago.20:49:39
@vcunat:matrix.orgVladimír ČunátTo me, removal of support is just removal of commitment/guarantees about maintenance.20:50:18
@vcunat:matrix.orgVladimír ČunátBut if someone still wants to do it, I can't see why block it. But naturally it should have lower priority on the shared infra.20:50:55
@vcunat:matrix.orgVladimír Čunát * But if someone still wants to do it, I can't see why block it. Naturally it should have lower priority on the shared infra.20:51:05
@janne.hess:helsinki-systems.deJanne HeßI've had this discussion in another context, but I think changing EOL releases makes it seem a lot like it's still (at least partially) maintained20:51:49
@vcunat:matrix.orgVladimír ČunátIf they are, is it bad that they look like it?20:52:39
@vcunat:matrix.orgVladimír ČunátMaybe if it's confusing, yes.20:53:11
@janne.hess:helsinki-systems.deJanne HeßWell if they were, we should have a consistent wording. status.nixos.org clearly states these as EOL20:53:15
@janne.hess:helsinki-systems.deJanne Heßyeah20:53:16
@vcunat:matrix.orgVladimír ČunátRelease notes stress the dates visibly, more importantly.20:53:39
@vcunat:matrix.orgVladimír ČunátBut I can't estimate how well this information gets passed in practice. Our point of view is too biased to see ourselves.20:54:29
@janne.hess:helsinki-systems.deJanne HeßHm, if it was me, I'd disable the jobsets (maybe leave the previous release?) and protect the release- branch. This way we can get around all possible accidents. But that's my opinion and I know it's the strictest one and there are people who would probably disagree20:54:56
@janne.hess:helsinki-systems.deJanne Heß * Hm, if it was me, I'd disable the jobsets (maybe leave the previous release?) and protect the release- branch. This way we can get around all possible accidents. But that's my opinion and I know it's the strictest one and there are people who would disagree20:55:07
@vcunat:matrix.orgVladimír ČunátRight, e.g. for older than 21.05 right now that makes sense to me.20:57:25
@vcunat:matrix.orgVladimír ČunátThough very little impact I guess, as I'd hope noone uses such old channels/branches anymore.20:58:25
@vcunat:matrix.orgVladimír Čunát * Though very little impact I guess, as I'd hope noone follows such old channels/branches anymore.20:58:34

There are no newer messages yet.


Back to Room ListRoom Version: 6