# AiiDA core coding days 21-22 Aug 2025

###### tags: `event`

[toc]

## General Notes

- What about the PRs from the GSOC time?
- Closed some stale PRs

## Edan

* https://github.com/aiidateam/aiida-core/issues/6976
- Looking into a potential bug with the Pydantic model system
- `ProcessNode` instances cannot be created from `from_serialized` due to mismatch between `Model` fields and backend constructor arguments
- This might be a larger issue (the mismatch) beyond just `ProcessNode`
- [jg] Use cases? Data exchange -> REST API, interaction with GUI
- [jg] serialization for storage in db? -> How is it right now serialized for that
- Look together with [xw] and [ke] and [dh]?
* JG: @Edan, do you think there will be time also for [#6198 Migrate `Transport` to use pydantic for configuration](https://github.com/aiidateam/aiida-core/pull/6198)?
* EB: @Julian, hopefully, though that one is awaiting some feedback from Ali/Alex. But yes, we will work towards wrapping all pydantic PRs.

## Marnik

* Reviewed my boi's JG's PR: https://github.com/aiidateam/aiida-core/pull/6946
* @giovannipizzi: I think you originally suggested that the `tmp_dir` should be configurable via an option, but I'm not sure that has any real use case, see [my comment here](https://github.com/aiidateam/aiida-core/pull/6946#pullrequestreview-3138924398).
* Raised new issue related to archive creation: https://github.com/aiidateam/aiida-core/issues/6982
* I feel I've seen this before, wasn't this solved? Also I think I did a similar large export on the same profile at EPFL (have to check `aiida-core` version) and it worked. ➡️ Will look into this further.
[JG] I feel the same, seems familiar.
* [Improve docstrings of `chmod` method of transports](https://github.com/aiidateam/aiida-core/issues/6974)
-> **PR open**: https://github.com/aiidateam/aiida-core/pull/6988
* Look into failing nightly build of `aiida-quantumespresso`
* Review Daniel's `pytest` fixtures tests: https://github.com/aiidateam/aiida-core/pull/6882

## Julian

**Notes**
* This appeared again, and is critical:
* https://github.com/aiidateam/aiida-core/issues/6545
* Archive creation breaks for as low as 3.5k PwBaseWorkChains
* Might be related to changes of bulk add_node PR
* Creating nodes and adding them to DB as slow as ~1 min per 1k Int nodes??
* Should we continue to assign people on GH? Or just request review?
* Closed a few stale PRs (require discussion)

**PRs**
* #6967: Cleanup (almost) all warnings in the test suite - PR by danielhollas - approved & merged
* #6966: Improve typing of QueryBuilder.all() - PR by danielhollas - approved & merged
* #6929: Fix sqlite_zip profile deletion with missing aiida archive - merged -> Good discussion with agoscinski (see also https://github.com/aiidateam/aiida-core/issues/6984)
* #6946: Add --tmp-dir option for archive creation - Discussion ongoing, do we even need the --tmp-dir option? (thanks MBx for the review)
* #6985: Remove ci section from .pre-commit-config.yaml -> merged
* #6963: Prior check if migration needed on SqliteZipBackend initialisation - [jg] still need to fix review comments, thanks, XW!
* #6733: Add ArithmeticAdd CJ Node fixture without run or submit - [jg] ~3-times speedup, need to see where to replace

## PRs

* Assignees: GeigerJ2
- #6967: Cleanup (almost) all warnings in the test suite (by danielhollas) -> approved & merged
- #6966: Improve typing of QueryBuilder.all() (by danielhollas) -> approved & merged
- #6829: Adding additional information to be shown with `verdi storage info --detailed` (by ayushjariyal) -> Waiting for tests
- #6819: Add dedicated path type throughout the codebase (by ayushjariyal) -> todo

* Assignees: agoscinski
- #6929: Fix `sqlite_zip` profile deletion with missing aiida archive (by GeigerJ2) -> Merged
- #6981: Bump the gha-dependencies group across 1 directory with 3 updates (by dependabot) -> todo

* Assignees: danielhollas
- #6820: Bug fix of get angle (by Muhammad-Rebaal) -> todo
- #6780: `CI`: remove matrix redundancy between `test-install` and `ci` (by khsrali) -> discussion
- #6760: `orm/utils/node.py`: Bug fix, `removeprefix` was being applied to `'data'` instead of `'process'` (by Muhammad-Rebaal) -> Will be superseeded by #6983

* Assignees: edan-bainglass
- #6198: Migrate `Transport` to use pydantic for configuration (by sphuber)

* Assignees: khsrali
- #6847: Use TextIOWrapper for opening file from repository in text mode (by ahkole) -> todo
- #6434: CLI: Add color flag to `verdi process list` (by agoscinski) -> closed as stale

* Assignees: mbercx
- #6946: Add `--tmp-dir` option for archive creation (by GeigerJ2) -> Discussion ongoing, do we even need the `--tmp-dir` option

* Assignees: superstar54
- #6963: Prior check if migration needed on SqliteZipBackend initialisation (by GeigerJ2) -> [jg] still need to fix review comments, thanks, xw!

* Assignees: unkcpz
- #6954: Typecheck aiida.cmdline.utils.{ascii_vis,common,echo}.py modules (by danielhollas) -> review
- #6736: xy data documentation (by ayushjariyal)
- #6733: Add `ArithmeticAdd` CJ Node fixture without `run` or `submit` (by GeigerJ2) -> [jg] ~3-times speedup, need to see where to replace

* Assignees: Unassigned
- #6922: Remove the publish-testpypi job in release workflow (by agoscinski) -> DH?
- #6882: Add more pytest fixture tests (by danielhollas) -> MBx?
- #6879: Filter active processes for the default user when running `verdi process repair` (by edan-bainglass)
- #6828: fix(Group Node): rename `time` field to `ctime` for consistency with other node types (by superstar54) -> db-migration
- #6811: issue#6393 Fixed User unfriendly behavior of attributes.all when stored (by Karanveer266) -> Discussion?
- #6741: Enforce context usage of Container within migrator (by agoscinski)
- #6722: Move all 2nd layer import to 1st layer to form pub APIs (by unkcpz) -> aiida v3
- #6659: Storage: Add `extend_json` Tag for Enhanced JSON Field Handling in `bulk_update` Operations (by rabbull) -> db-migration
- #6636: Mark private functionalities with an underscore to separate better from public API (by agoscinski) -> aiida v3
- #6632: Providing support for atomistic StructureData (by mikibonacci)
- #6578: Add `--also-remote` option to `verdi process dump` (by GeigerJ2) -> refactored
- #5971: `Node`: Validate nested attribute and extra keys (by mbercx)
- #5840: `Repository`: Cache the calculation of the hash (by sphuber)

gh pr list --limit 100 --state open --draft=false --json number,title,assignees,author,url | jq -r '
  group_by(.assignees | map(.login) | sort | join(",")) |
  map({
    assignees: (.[0].assignees | map(.login) | join(", ")),
    prs: map({number, title, author: .author.login, url})
  }) |
  .[] |
  "Assignees: " + (if .assignees == "" then "Unassigned" else .assignees end) + "\n" +
  (.prs | map("  #\(.number): \(.title) (by \(.author))") | join("\n")) + "\n"'