VOTE: Use "CI Fix" label for merging into continuous integration runs

[David Roe]

The following proposal has been made several times the last few weeks: in PR #37428, in this thread and then in this thread.  It is orthogonal to the ongoing vote in this thread.  With no further discussion, I'm calling a vote.

Background

Starting in Sage 10.2, PRs with the Blocker label have been merged into all other PRs before running CI; see the changelog and this post for more details.  This has led to disagreements about whether this label should be applied.

Proposal

We use "CI Fix" rather than Blocker to determine whether an open PR should be merged before running CI.  Blocker will retain its previous meaning of a PR that should be merged before the next release is finished.  The process below describes how to resolve disagreements about whether the "CI Fix" label should be applied.

a. Only PRs with positive review should be marked with the "CI Fix" label.  This should be done if both author and reviewer agree that it is appropriate, and a rationale should be given in a comment on the ticket.

b. If a PR becomes disputed (as described in this proposal), the "CI Fix" status can be voted on separately upon request; otherwise it should be applied if and only if positive review is applied.

Voting will be open until Wednesday, March 13.

David

[Kwankyu Lee]

+1

[Dima Pasechnik]

+1

--
You received this message because you are subscribed to the Google Groups "sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CAChs6_mYLUWXMU6AZKJGPKd2oz0AC_qAUjnGoD9Q9yixzNBC2w%40mail.gmail.com.

[Giacomo Pope]

+1

[Edgar Costa]

+1

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/77cc05b2-1c52-4aae-80ca-4d0ec0830e2dn%40googlegroups.com.

[William Stein]

+1

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CA%2BiQ7x40QryVx%2BPDDcGB4bJwrd%3D5NqbT94Xq5nSb0wxncwvGZQ%40mail.gmail.com.

--

William (http://wstein.org)

[Matthias Koeppe]

+1

[G. M.-S.]

+1

Guillermo

[Vincent Delecroix]

+1

> --
> You received this message because you are subscribed to the Google Groups "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CANnG189Rx%3DfXHDCQvDiTS0f_qHwRG9LbfiKKqph6DXuF%3Dd1zcw%40mail.gmail.com.

[John H Palmieri]

+1

[Nathan Dunfield]

+1

[Travis Scrimshaw]

+1

[David Lowry-Duda]

+1

- DLD

On 03:43 Mon 04 Mar 2024, David Roe wrote:
>The following proposal has been made several times the last few weeks: in

>PR #37428 <https://github.com/sagemath/sage/issues/37428>, in this thread
><https://groups.google.com/g/sage-devel/c/uWHaWK7b6H4> and then in this
>thread <https://groups.google.com/g/sage-devel/c/XDvKkMRoDk4>. It is

>orthogonal to the ongoing vote in this thread

><https://groups.google.com/g/sage-devel/c/E0qfJTMETDk>. With no further

>discussion, I'm calling a vote.
>

>*Background*

>
>Starting in Sage 10.2, PRs with the Blocker label have been merged into all
>other PRs before running CI; see the changelog

><https://github.com/sagemath/sage/wiki/Sage-10.2-Release-Tour#open-blocker-prs-are-applied-automatically-in-ci-workflows>
>and this post
><https://groups.google.com/g/sage-devel/c/uWHaWK7b6H4/m/cM77D8E5AAAJ> for

>more details. This has led to disagreements about whether this label
>should be applied.
>

>*Proposal*

>We use "CI Fix" rather than Blocker to determine whether an open PR should
>be merged before running CI. Blocker will retain its previous meaning of a
>PR that should be merged before the next release is finished. The process
>below describes how to resolve disagreements about whether the "CI Fix"
>label should be applied.
>a. Only PRs with positive review should be marked with the "CI Fix" label.
>This should be done if both author and reviewer agree that it is
>appropriate, and a rationale should be given in a comment on the ticket.
>b. If a PR becomes disputed (as described in this proposal

><https://groups.google.com/g/sage-devel/c/IgBYUJl33SQ>), the "CI Fix"

>status can be voted on separately upon request; otherwise it should be
>applied if and only if positive review is applied.
>
>Voting will be open until Wednesday, March 13.
>David
>

>--
>You received this message because you are subscribed to the Google Groups "sage-devel" group.
>To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.

>To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CAChs6_mYLUWXMU6AZKJGPKd2oz0AC_qAUjnGoD9Q9yixzNBC2w%40mail.gmail.com.

--
David Lowry-Duda <da...@lowryduda.com> <davidlowryduda.com>

[Eric Gourgoulhon]

+1

Eric.

[seb....@gmail.com]

+1

Eric Gourgoulhon schrieb am Sonntag, 10. März 2024 um 16:58:42 UTC+1:

+1

Eric.

[David Roe]

This vote has passed, and I've added a "CI Fix" label on github.  I am not going to have time to implement changes to our CI infrastructure in the near future, but others are welcome to do so.

David

--
You received this message because you are subscribed to the Google Groups "sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/d35c59ad-b5a6-4fb7-9a2f-8d0adc238ce3n%40googlegroups.com.

[Matthias Koeppe]

I have implemented this change in https://github.com/sagemath/sage/pull/37950, needs review.

[Matthias Koeppe]

This was merged in 10.4.beta6, so the "CI Fix" label can now be used.