# AiiDA Team Meeting 2021-05-05

###### tags: `team meetings`
###### time: 4pm CET

[TOC]

### Present

* Chris
* Marnik
* Leo
* Louis
* Seb
* Simon
* Flaviano
* Casper
* Sasha
* Gio

### Catch-up round

*Max. 3 minutes each*

Marnik:
* QE school AiiDA tutorial:
* Date set: Friday 28th of May 2021 - Only in morning (see [Program](http://indico.ictp.it/event/9616/other-view?view=ictptimetable)).
* Tutors: Myself (presenter), Francisco, Chris, Flaviano (?) - Confirm
* AiiDA July tutorial:
* Dates: 5-9 July 2021 ([Event site is online](https://www.aiida.net/aiida-virtual-tutorial-july-2021/)).
* Docs: [Advertisement](https://docs.google.com/document/d/1tk90RDwBGYXYNio2zeSceARWWwHFEwdlUppoJGK7Ktk/edit#) ; [Application form](https://docs.google.com/forms/d/1jM6KoBQpbFjZTeXUMmatQICwrnsUuAqnjqGxu7wgUs8/edit) ; [Schedule](https://docs.google.com/spreadsheets/d/1taYiFCfnmBtuydd-8Yc_ifSuyOZ-Om4OcbnBlwastPU/edit#gid=64673141) ; [Website](https://docs.google.com/document/d/1vit5svzvh3Oo6zo8TbYLEBFCm2V6Fi87Wb43qtzgdIg/edit#)
* Giovanni might only be partially available, anyone else who is unavailable? @Louis: are you available this week?
* Working on the material:
* 19th of May: Francisco and I will reorganise the material into modules and create/assign issues by this date.
* 19 May - 7 June: Time for everyone to test/update their section.
* Week of 7th of June: organise 2 coding days. [FILL IN DOODLE](https://doodle.com/poll/ngwc6me3f93hizrr)
**Note:** PR for your section is merged? No coding days for you! :tada:
* New AiiDA team member checklist (Work in progress): [document](https://docs.google.com/document/d/1l-ulmWOCprSfavja5WwD8jW7a43hn1CMusrFJzafdSk/edit#) - Additions/Comments welcome!

Simon:
* _Not much to report, mainly MarketPlace / AiiDAlab work._
* _On vacation as of May 13th (next Thursday) until May 31st._
* Reviewed PR on query how-to [#4903](https://github.com/aiidateam/aiida-core/pull/4903)
* In progress: PyYAML update (security report); related: we should be using [safe_load()](https://pyyaml.org/wiki/PyYAMLDocumentation)
* [NF updates]
* _Requests now send out at the beginning of each month, apparently managed by Arliss Collins._
* v1.6.3 release
* Tutorial - Take from twitter

Leopold:
* Worked on updating plugins for AiiDA 2
* done: aiida-zeopp, aiida-cp2k
* in progress: aiida-ddec
* on hold: aiida-raspa
* Have seen your comments on the GIN index for process state - will give a look
* In case you're also on Apple Silicon and considering to switch to dockerized rabbitmq+postgresql [gist](https://gist.github.com/ltalirz/fda14abd28afa2a63f5fc0fe6e782c96)
* Ninad [opened AEP](https://github.com/aiidateam/AEP/pull/24) on workflow extension for AiiDA REST API - will review later today, your reviews are welcome as well
* Do we need to have a final vote on how to specify verbosity?
`-v`, `-vv`, ... vs `-v DEBUG`, `-v INFO`, `-v WARNING` , `-v CRITICAL`, ...

Flaviano:

* aiida-core: Support for magnetic structure [#4866](https://github.com/aiidateam/aiida-core/issues/4866). Long discussion, no clear conclusions. What to do and where?
* aiida-quantumespresso: Support for noncollinear calculation in the get_builder_from_protocol [#679](https://github.com/aiidateam/aiida-core/issues/4866). Temperarily solution to facilitate handling complex magnetic materials. To implement: Pass a list of vectors instead of a list of scalers.

Sebastiaan:
* **PR**_[merged]_ Fix: respect nested output namespaces in Process.exposed_outputs [[#4863]](https://github.com/aiidateam/aiida-core/pull/4863)
* **PR**_[merged]_ Fix aiida.cmdline.utils.decorators.with_dbenv always loading the db [[#4865]](https://github.com/aiidateam/aiida-core/pull/4865)
* **Release v1.6.2**
* **PR**_[merged]_ Fix IndexError in override_log_formatter_context utility [[#4873]](https://github.com/aiidateam/aiida-core/pull/4873)
* **PR**_[merged]_ Fix aiida.cmdline.utils.decorators.load_backend_if_not_loaded [[#4878]](https://github.com/aiidateam/aiida-core/pull/4878)
* **Release v1.6.3**
* **PR**_[merged]_ Remove all deprecated resources for v2.0 [[#4881]](https://github.com/aiidateam/aiida-core/pull/4881)
* **PR**_[merged]_ Rename name column of Computer database table to label [[#4882]](https://github.com/aiidateam/aiida-core/pull/4882)
* **PR**_[merged]_ verdi status: add version and repository UUID to the output [[#4883]](https://github.com/aiidateam/aiida-core/pull/4883)
* **PR**_[merged]_ Profile: do not leak repository implementation in interface [[#4891]](https://github.com/aiidateam/aiida-core/pull/4891)
* **PR**_[merged]_ Reinstate aiida.manage.tests.unittest_classes [[#4894]](https://github.com/aiidateam/aiida-core/pull/4894)
* **PR**_[merged]_ Config: warn when the configuration directory is created [[#4896]](https://github.com/aiidateam/aiida-core/pull/4896)
* **PR**_[merged]_ Database schema: set {} as default for DbNode.repository_metadata [[#4898]](https://github.com/aiidateam/aiida-core/pull/4898)
* **PR**_[merged]_ Profile: only initialise the repository during the migration [[#4900]](https://github.com/aiidateam/aiida-core/pull/4900)
* **PR**_[merged]_ CLI: set localhost as default for database hostname in verdi setup [[#4908]](https://github.com/aiidateam/aiida-core/pull/4908)
* **PR**_[open]_ Config: add the delete_profile method [[#4901]](https://github.com/aiidateam/aiida-core/pull/4901)
* **PR**_[open]_ ORM: replace InputValidationError with ValueError and TypeError [[#4888]](https://github.com/aiidateam/aiida-core/pull/4888)

Casper:
- Tested the repository for migrations with the new repository.
Found some issues that have been fixed (cf. Seb).
Will still conduct a last test for the large `li-ion-conductors` (Leonid Kahle) database, but I will wait for the final issues found to be fixed (if some are still not fixed).
Note, for the OPTIMADE test databases it seems to work fine with the [AiiDA-OPTIMADE](https://github.com/aiidateam/aiida-optimade) package that implements an OPTIMADE REST API server.
- Opened issue [#4910](https://github.com/aiidateam/aiida-core/issues/4910) to follow up on the `path` vs. `raw_input` folders during repository migration.
- Discussed the work that has to be done for a possible CUDS/OWL plug-in for AiiDA with Louis Ponet (this is a key part of the OpenModel EU project, as well as a large part of the INTERSECT and MarketPlace projects).

Giovanni:
* Happy birthday Sasha! 🎉🎂
* tested the repository; all seems fine except for strange error on multi-threaded SQLite access via SQLAlchemy when running the daemon ([[#4899]](https://github.com/aiidateam/aiida-core/issues/4899)). Not clear if it leads to data corruption, but we should undertsand why (any idea on how to understand who's triggering it?)
* Martin would be available for discussions on the design of the engine when needed (e.g. number of slots, avoid duplicate subscribers errors, ...)
* Various benchmarking done on querying performance for large DBs (especially `verdi process list`). See e.g. [[#2770]](https://github.com/aiidateam/aiida-core/issues/2770). I describe some tests to be done by people with large DBs (>~100'000 calculations; at least Leo, Seb, Marnik).
* Support for SQLAlchemy 1.4 requires work, especially on QueryBuilder [[#4813]](https://github.com/aiidateam/aiida-core/issues/4813). Note that SQLA 1.4 prepares for the 2.0 release, so we should directly prepare/change the syntax for the 2.0 version I think. Disk-Objecstore is essentially ready to at least work fine with SQLA 1.4 ([[#108]](https://github.com/aiidateam/disk-objectstore/pull/108)) even if I think it would be better to revisit all calls to conform with the new SQLA 2.0 syntax.
* Reminder: meeting later today (3PM) on open issues for (our) 2.0 release.

Sasha:
* Tested new repository for the common workflows, found a couple of issues:
* BigDFT (Excepted <AttributeError: 'Repository' object has no attribute '_get_base_folder'>). Created [issue](https://github.com/BigDFT-group/aiida-bigdft-plugin/issues/6).
* Fleur (Excepted <TypeError: put_object_from_filelike() got an unexpected keyword argument 'mode'>). Created [issue](https://github.com/JuDFTteam/aiida-fleur/issues/120).
* verdi data psml uploadfamily (set_file() got multiple values for argument 'filename'). Created [issue](https://github.com/aiidateam/aiida-core/issues/4932).
* Tested new repository on AiiDAlab:
* Migrated my database on AiiDAlab (~23000 nodes) - all fine.
* QE app: seems to work fine.
* LSMO app: in progress.
* Nanribbon app: in progress.
* Issues:
* Computers: name is deprecated, to be replaced with label
* Transport.get_valid_transports() to be replaced with get_entry_point_names("aiida.transports").
* Scheduler.get_valid_schedulers() to be replaced with get_entry_point_names("aiida.schedulers")
* Found an issue in QE (that has been fixed in QE 6.6): fermi energy is not reported for the fixed occupation. Open issue to fix [here](https://github.com/aiidateam/aiida-quantumespresso/issues/681).

Chris:
- Released 1.6.2 (yanked!), released 1.6.3 on Conda
- wanted to add pip check but issue with psycopg2-binary
- PR reviewing and testing (looking in to multi-threaded SQLite)
- Started working on repository CLI command and new archive format
- `verdi node delete` no longer reclaims disk space
- Updates to aiida-lammps

### Specific Agenda Items