# AiiDA Team Meeting 2023-06-1 ###### tags: `team meetings` ###### time: 14:00 CET [TOC] ### Present * Giovanni * Xing * Marnik * Jason * Edan * Francisco * Kristjan * Chris * Sebastiaan ### Catch-up round *Max. 3 minutes each* Marnik: * ✅ **Merged PRs**: * `ase`: [Parsing: Fix `pbc` parsing for `extxyz` format](https://gitlab.com/ase/ase/-/merge_requests/2876) - related to [[#5967](https://github.com/aiidateam/aiida-core/issues/5967)]: Waiting for release, but last release of ASE on PyPI was December 1st 2021... * Replace flask `JSONEncoder` by `JSONProvider` [[#5987](https://github.com/aiidateam/aiida-core/pull/5987)]. I mean kind of merged. Seb had to clean it up because I went too far with the dependency changes. 😇 * ↩ī¸ **Open PRs**: * `Node`: Validate nested attribute and extra keys [[#5971](https://github.com/aiidateam/aiida-core/pull/5971)]. Already reviewed one round, ready for round two @7wMTWWsmS1i2a1QmzsgXgg (that's apparently you, Seb). * 📋 **Projects / Road map** - Same as last meeting * Organised ["User-friendliness" project](https://github.com/orgs/aiidateam/projects/4/views/1) a bit. Still a lot of triaging to be done. * Regarding [moving the "materials science" data types out of AiiDA-core](https://github.com/aiidateam/team-compass/issues/21), I plan to start the `aiida-atomistic` repo soon, and start moving issues related to the materials-science types _that require backwards-incompatible changes_ there. * đŸĢĩ Appoint a new dependency manager (DM), see [AEP 002](https://github.com/aiidateam/AEP/tree/master/002_dependency_management). * 🙋‍♂ī¸ Self-appointed myself as mailing list manager (or whatever we use next). * 🙏 Have a `type/user-friendliness` to distinguish from other "usability" problems. * ❓ Coding week: dedicated session on protocols/overrides? Jason: * Who can access to [aiida-bot](https://github.com/aiida-bot) account? We need create a token for `aiida-registry` for GSoC. * To support smart caching, I need the changes of https://github.com/aiidateam/aiida-core/pull/6039. It looks a pattern for the similar workflow, I have it for the PH and BandWorkchain. If this is useful, we can have another "base" workchain for this logic. * Is that possible to remove the dependency of pymatgen from aiida-quantumespresso? * Import archive has no hash for all nodes, is this expected? I use master branch with some changes for caching. Giovanni: * Schedule already (every ~1.5months?) dedicated coding two-days? E.g. Mon/Tue. To help convert issues into PRs. We can rotate who organizes them (i.e. ensures we identify 1 issue per person to work on that is high priority and doable in 2 days, typically on bug fixes or usability/documentation issues) * Plans/timeline for the infinite recursion error fix (Chris) * AiiDA 3.0: roadmap? What are the breaking changes that we want to introduce and are realistic in the next ~6 months? Probalby we don't needed it - for now we just check if there are things we *might* want to do * Knowledge transfer from Chris to Edan on a number of AiiDA-related topics * Discussion and KT from Chris to Jason for Quantum Mobile * Coding week: let's fix place and time and people schedule trips etc. (also fix room - CoViz2?) + activities * Firecrest roadmap * Discussed with Jason about caching and RemoteData nodes. Possibly something for Jason to write down, and we have a focused meeting to decide what to implenent in AiiDA core (if anything, or just docs) Xing: * I have actively used aiida for the XPS calculation in the last month. Reported a few issues ([#937](https://github.com/aiidateam/aiida-quantumespresso/issues/937), [#938](https://github.com/aiidateam/aiida-quantumespresso/issues/938)) and made PRs ([#6032](https://github.com/aiidateam/aiida-core/pull/6032), [#936](https://github.com/aiidateam/aiida-quantumespresso/pull/936)). Will report more this month. * In PwCalculation, `_default_symlink_usage` default is set to `False` to allow multiple nscf to run on top of the same scf. Are many case like this? If not, better to set it to `True`. In of my calculations (restart from failed one), it took 2 h for upload (copy around 150 files). Kristjan: * Nothing aiida-related Sebastiaan: * [Add the `SqliteDosStorage` backend and `verdi profile setup`](https://github.com/aiidateam/aiida-core/pull/6023): New backend that uses disk-object store and sqlite. Variant of existing backends, but this is "temporary" but mutable. Very useful for tutorials, demos etc. where you don't care too much about performance, but setup is extremely easy. Especially with the new command `verdi profile setup` that generates CLI options dynamically for the backend. `verdi profile setup core.sqlite_dos -n --profile sqlite-dos` automatically creates a new profile with functioning backend. * [`CalcJob`: Add support for nested targets in remote_symlink_list](https://github.com/aiidateam/aiida-core/pull/5974) * [`BaseRestartWorkChain`: Factor out attachment of outputs](https://github.com/aiidateam/aiida-core/pull/5983) * [Repository: Cache the calculation of the hash](https://github.com/aiidateam/aiida-core/pull/5840) ### Specific Agenda Items