|
| 1 | +--- |
| 2 | +tags: [meeting-notes] |
| 3 | +title: '2024-05-15' |
| 4 | +--- |
| 5 | +# conda-forge core meeting 2024-05-15 |
| 6 | + |
| 7 | +Add new agenda items under the `Your __new__() agenda items` heading |
| 8 | + |
| 9 | +- [Zoom link](https://zoom.us/j/9138593505?pwd=SWh3dE1IK05LV01Qa0FJZ1ZpMzJLZz09) |
| 10 | +- [What time is the meeting in my time zone](https://dateful.com/convert/utc?t=5pm) |
| 11 | +- [Previous meetings](https://conda-forge.org/community/minutes/) |
| 12 | + |
| 13 | +## Attendees |
| 14 | + |
| 15 | +| Name | Initials | GitHub ID | Affiliation | |
| 16 | +| ----------------------- | -------- | --------------- | --------------------------- | |
| 17 | +| Matthew R Becker | MRB | beckermr | conda-forge | |
| 18 | +| Nichita Morcotilo | NM | nichmor | prefix.dev | |
| 19 | +| Marcel Bargull | MB | mbargull | Bioconda/cf | |
| 20 | +| Filipe Fernandes | FF | ocefpaf | conda-forge | |
| 21 | +| Klaus Zimmermann | KZ | zklaus | Quansight | |
| 22 | +| Wolf Vollprecht | WV | wolfv | prefix.dev | |
| 23 | +| John Kirkham | JK | jakirkham | NVIDIA/cf | |
| 24 | +| | | | | |
| 25 | +| | | | | |
| 26 | + |
| 27 | +X people total |
| 28 | + |
| 29 | +### Standing items |
| 30 | + |
| 31 | +- [ ] |
| 32 | + |
| 33 | +### From previous meeting(s) |
| 34 | + |
| 35 | +- [ ] |
| 36 | + |
| 37 | +### Active votes |
| 38 | + |
| 39 | +- [ ] |
| 40 | + |
| 41 | +### Your __new__() agenda items |
| 42 | + |
| 43 | +- [X] (JK) Conda-build 24.5.0 |
| 44 | + - Now released. Small [hiccups](https://github.com/conda-forge/status/issues/177) in conda-forge-ci-setup with deprecated code that was already cleared up. |
| 45 | + - https://github.com/conda-forge/conda-forge-ci-setup-feedstock/issues/317 |
| 46 | + - Watch for `conda_build.config.subdir` imports and similar lines. |
| 47 | + - Could we add some testing with Conda Canary or make the warnings more visible |
| 48 | + - Perhaps https://github.com/conda-forge/cf-autotick-bot-test-package-feedstock |
| 49 | + - https://github.com/conda-forge/conda-smithy/issues/1881 |
| 50 | +- [X] Fundable tasks |
| 51 | + - https://github.com/conda-forge/conda-forge.github.io/issues/2172 |
| 52 | + - Hard to build a long list of items in HackMD or similar |
| 53 | + - Could we track this list on GH directly by labeling issues or similar |
| 54 | + - Looking for feedback/discussion on this issue |
| 55 | + - Would be nice to have this ready when someone shows up with a checkbook :) |
| 56 | + - PSF uses this approach: https://github.com/psf/fundable-packaging-improvements |
| 57 | + - Could either centralize to webpage repo (maybe with stub issues) or decentralize by just adding labels on repo issues where needed |
| 58 | + - How to discover? |
| 59 | + - Ideally want a link to find them all |
| 60 | +- [X] (JK) GLIBC 2.28 |
| 61 | + - (JK) CDT list ready? |
| 62 | + - https://github.com/conda-forge/cdt-builds/issues/66 |
| 63 | + - (IF) What to do with the ones with the `?` |
| 64 | + - (HV) Could build them for safety |
| 65 | + - (IF) Should we wait until they are requested |
| 66 | + - (HV) Concern that this could be fairly slow process |
| 67 | + - (HV) Might make it slow to upgrade |
| 68 | + - (IF) Would like to reduce these on the outset |
| 69 | + - (HV) Can we review `libglvnd`? |
| 70 | + - https://github.com/conda-forge/staged-recipes/pull/25919 |
| 71 | + - (IF) Licenses (bug?) in https://github.com/conda-forge/cdt-builds? Couldn't find them |
| 72 | + - (MRB) May be a bug |
| 73 | + - Items marked with green checks at https://github.com/conda-forge/cdt-builds/issues/66#issuecomment-1833417828 can be part of the new rebuilds |
| 74 | + - IF: Let's not add cos8 to the cdt_name, let's just use 'conda' to get rid of `cdt_name` in the pinnings and so. |
| 75 | + - (JK) Noted in issue: https://github.com/conda-forge/conda-forge.github.io/issues/1941 |
| 76 | + - (HV) Compilers already account for this |
| 77 | + - (MRB) sysroot repodata hack can be dropped |
| 78 | +- [X] (JK) NumPy 2 |
| 79 | + - (HV) Currently best [option](https://github.com/conda-forge/conda-forge-pinning-feedstock/pull/5851#issuecomment-2111433502) seems to "soft-close" 3.12 migration; avoids a lot of PR churn, but keeps status page and bot running. |
| 80 | + - Adding NumPy 2.0 bump in Python 3.12 migrator would break that migrator |
| 81 | + - Build with Python 3.12 (lots of churn) |
| 82 | + - (Soft-)Close Python 3.12 |
| 83 | + - Def of Soft |
| 84 | + - Apply migration to global pinnings |
| 85 | + - Keep migrator open |
| 86 | + - Benefit is tracking status of migrator |
| 87 | + - (MRB) In favor of soft-closing |
| 88 | + - (HV) Have typical run until the end |
| 89 | + - (MB) Time a migrator stays open is not defined anywhere, so it's ok to override the "normal time other migrators have stayed open" |
| 90 | + - (IF) Possible technical solution. Is there a key ignore option? |
| 91 | + - (MRB) It is undocumented |
| 92 | + - (MRB) [`override_cbc_keys`](https://github.com/regro/cf-scripts/blob/master/conda_forge_tick/make_migrators.py#L387) |
| 93 | + - (MRB) Allows to override what keys are used from `conda-build-config` |
| 94 | + - (MRB) Link: https://github.com/regro/cf-scripts/blob/master/conda_forge_tick/make_migrators.py#L387 |
| 95 | + - Consensus is soft-closing Python 3.12 migration to unblock NumPy 2 |
| 96 | + - (MB) Recommend adding a note that is a pragmatic decision. Not based on a subset of the community's recommendation |
| 97 | +- [X] (JK) CUDA ppc64le support |
| 98 | + - https://github.com/conda-forge/cuda-feedstock/issues/27 |
| 99 | + - Noted in CUDA 12.4 next release won't have ppc64le support in CTK |
| 100 | + - Still building with CUDA 12.0 in conda-forge, which still has ppc64le support |
| 101 | + - Also existing packages through CUDA 12.4 are still available |
| 102 | +- [X] (HV) NumPy [considering](https://github.com/numpy/numpy/pull/26430#issuecomment-2108839048) to move to VS2022 baseline |
| 103 | + - Probably far-reaching enough that we'd have to upgrade too |
| 104 | + - VS2019 is EOL since a few weeks, but not yet being removed from various CI providers |
| 105 | + - JRG: Azure has no mentions of future vs2019 deprecations at https://learn.microsoft.com/en-us/azure/devops/pipelines/agents/hosted?view=azure-devops&tabs=yaml |
| 106 | +- [x] (WV) Neighbourhoodie SOW accepted :tada: |
| 107 | +- [x] (WV / NM) rattler-build support in conda-smithy |
| 108 | + - https://github.com/conda-forge/conda-smithy/pull/1876 |
| 109 | + - Any questions? |
| 110 | + - (MRB) reviewing currently :) |
| 111 | + - (WV) Label handling |
| 112 | + - (IF) Handled in conda-forge-ci-setup |
| 113 | + - (IF) Please send a PR :) |
| 114 | +- [X] (MRB) lots of recent bot work |
| 115 | + - bot now carefully uses docker containers for isolation of recipe / user code from bot info for most steps |
| 116 | + - migrations as objects can now be serialized to JSON |
| 117 | + - documentation has moved to the README in the bot repo (no more GH pages, all links should be updated) |
| 118 | + - cleaned out old code, env vars, etc. |
| 119 | + - logs should be a lot easier to parse now |
| 120 | +- [ ] (IF) Mingw support |
| 121 | + - Cross compiler works now (`host=linux-64` -> `target=win-64`) |
| 122 | + - Need native compiler to work (`host=win-64` -> `target=win-64`) |
| 123 | + - Needed for R builds |
| 124 | + - https://github.com/conda-forge/ctng-compilers-feedstock/pull/134 |
| 125 | + - Building this with `build=linux-64`. conda-build has issues here https://github.com/conda/conda-build/issues/5341 |
| 126 | + - conda-libmamba-solver also has to fix some things |
| 127 | + - |
| 128 | + |
| 129 | +### Pushed to next meeting |
| 130 | + |
| 131 | +- [ ] (JK) CUDA 12 |
| 132 | + |
| 133 | +### CFEPs |
| 134 | + |
| 135 | +- [ ] |
0 commit comments