2025-08-20

2025-08-20

Apologies

  • Alin Elena

Attending

  • Bjørk Hammer (BH)
  • Adam Jackson (AJJ)
  • Ask Hjorth Larsen (AHL)
  • Karsten Wedel Jacobsen (KWJ, chair initial 5 mins of meeting)
  • Yuji Ikeda (YI)
  • Paul Erhart (PE)
  • Xing Wang (XW)
  • JJ Mortensen (JJM)
  • James Kermode (JK, appointed chair)
  • Michele Ceriotti (MC, left early)
  • Hanna Türk (HT)
  • Gabor Csanyi (GC)
  • Michael Herbst (MH)
  • Lars Pastewka (LP)
  • Andrew Rosen (AR)
  • Elena Gelžinytė (EG)
  • Kristian Thygesen (KT)
  • Andrew Peterson (AP, arrived late)

AOB / Agenda suggestions

  • BH would like to discuss pre-releases
  • AHL invited JJM to attend this meeting

Appointment of chair

  • James Kermode was unanimously approved as chair

Review draft governance document

  • There was a discussion on whether members could send deputies to the steering group. The concensus was that since meetings will only take place every 6-12 months this would not be necessary.
  • Meeting frequency of every 6 months initialy and every 12 months in steady state was agreeed.
  • Process for rotating Chair and membership were discussed. Principle of rotation was agreed but details need to be refined. 2-3 year term limits for both chair and members were viewed as broadly appropriate, perhaps with a maximum of two 3-year terms. It was noted that there is a need for a process to bring in new Steering Commmittee members from the broasder community before term limits can be imposed.
  • ACTION JK to draft a proposal for next meeting
  • The commitee also agreeed the governance document could be clearer about what interests the committee is supposed to represent. This would help inform decisions about committee membership process.
  • ACTION JK to develop the document and circulate prior to next meeting, considering questions such as:
    • How does the steering committee relate to big technical decisions?
    • Does it “shield” maintainers from community?

ASE workshop

  • CECAM Report circulated. It was suggested this should be made availble more widely to the ASE community.
  • ACTION JK Circulate CECAM report via ASE users webpage and/or put in ASE-SC Meeting Minutes GitLab repo once this exists (see below)
  • New proposal submitted for CECAM/Psi-k workshop to take place in 2026 led by EG, HT and others. The committee thanked them for taking this forward.

ASE Technical subcommittee (ASE-TSC) update

General updates

  • A summary of ASE-TSC progress since the CECAM workshop was pre-circulated.
  • The web site has been moved to ase-lib.org and is now built with GitLab CI. This was welcomed by the committee, with thanks to the Technical Subcommitee members and to PE for purchasing the domain. Currently the documentation is shown on the top level domain, but in future it could be moved to docs.ase-lib.org with other subdomains used e.g. for plugins, etc. There was broad support for adding a new landing page that is less dynamic than the Sphinx docs. This could include a short overview of the project, names and affiliations of Steering Committee members, etc.
  • ACTION ALL Please update links!
  • ACTION ASE-TSC Setup top-level landing page. PE is willing to provide guidance on easiest way to do this.
  • Release 3.26.0 of ASE has been made.
  • This includes the new Optimizable interface.
  • It was agreed a Steering Commmitee email list should be setup on the DTU listserv. This will provide a means for ASE-TSC to contact the full ASE-SC for advice/support in between meetings.
  • ACTION AHL to setup committee email list and add members to it.
  • It was agreed that summary meeting minutes will be made public. Agreed outcomes will be listed along with summaries of discussions. Members will be given an opportunity to review minutes before they are made public. A GitLab repo will be setup to host meeting minutes within the ase GitLab organisation.
  • ACTION AJJ/JK to draft and circulate meeting minutes
  • ACTION AHL to setup a GitLab repo to host minutes

Documentation working group

  • We discussed setting up a Documentation working group to build on the outcomes of the breakout group at the CEAM workshop. It was propose that progress could be made via remote documentation sprints. This will need volunteers from community. Maintainers would be available for tech support but delegate the sprint organisation to others.
  • ACTION HT Assemble documentation group and select date for first sprint *Update since meeting: dates of 17 and 18th September 2025 have been chosen.
  • ACTION ALL publicise ASE doc sprint amongst your local ASE users. Consider booking a seminar room or similar to host a local group taking part in the sprint.
  • It was discussed that it would be good to create new issues for sub tasks. But there are so many open issues Important that they are high quality and easy to pick up. All are encouraged to use labels when creating/triaging issues.
  • ACTION ASE-TSC In the run up to code sprints, create a “milestone” for sprint items, so the milestone page can be used to track sprint effort without other clutter.

API changes

  • The most significant workshop outcome was a proposal to split Atoms and Calculators. An epic issue describing the proposed plans was created by JK and an initial draft of the new calculator API has been made by EG. An ase._4 subpackage has been created as a workspace for experimental changes, where MRs can be made to build things gradually. A full ASE 4 branch would be unmergeable over timescale required
  • ACTION AJJ arrange meeting EG, AHL, JK, YI, HT to seed an ASEv4 working group

Plugins

  • AJJ reported that progress on the plugins outcome from the ASE workshop is bering made in a separate repo as work in MRs got very messy. This will also soon be moved to ase._4 to start testing.

  • Items that will move to plugins will include:

    • untestable calculators
    • odd requirements
    • other “heavyweight” things like calculators that support opensource DFT codes, along with associated IO functions.
    • But still some calculators should be kept in core ASE so testing is meaningful
  • Documentation will be updated to point to more external calculators we don’t currently list
  • There was a discussion around performance impact which have been taken into consideration (along with the ASEv4 API changes above).

AOB for ASE-TSC

  • It was agreed a list of what ASE will never include would be useful, to help potential contributors know whether an MR is likely to be acceptable at an early stage.
  • Relatedly, there was a discussion in response to comments from some contributors who don’t like being asked to make big far-reaching changes for their small MR. It was agreed that while it is good to aspire to the ‘Boy Scout’ principle of always leaving the code in a better place than you found it, this needs to be kept proportionate.
  • ACTION ASE-TSC to draft some guidance on what makes a good/acceptable MR, but it is fundamentally difficult
  • There was a discussion around mutable/immutable atoms, optimisation, inconsistency - all items which the API changes above intended for ASEv4 are intended to fix.
  • Currently the pyproject.toml gives ase-users email address as contact which was agreed not to be ideal. It was agreed this could be switched to the committee email list once this has been setup.
  • ACTION AHL Change contact email once ase-committee setup.
  • It was agreed the GitLab issue tracker should not be used for basic user support, only for development. Help requests should be firmly redirected to the MatSci.org forum and closed. The issue template needs to be amended to make this policy clear.
  • ACTION ASE-TSC Technical subcommittee will implement this policy.
  • There was also a Discussion of reducing usage of the ase-users mailing list; in future it could be used for broadcasting information only. No action will be taken on this.

Pre-releases

  • A downstream code hit a breaking change in last release. It was suggested pre-released would help to catch this sooner. There was a consensus that the released frequency could be increased with more automation, but there is insufficient resource to support pre-releases as well. We recommend downstream codes should pin their ASE dependencies and/or test against master.
  • ACTION ASE-TSC Improve docs to recommend pinning dependency on ASE and testing against master

ASE foundation

  • There was strong consensus behind setting up an ASE Foundation which can receive money. Initially it is envisaged this would be small amount of money to support events, website expenses etc, and not to fund salary costs.
  • The challenge of donating public money without an obvious return was noted, along with the recognition we are not in a position to commit extra technical support.
  • Non-profit organisations like that run by the FHI-aims team, which charges a voluntary license fee, or charities such as Psi-k which receive donations provide possible templates. Charities can send ‘requests for payment’ for donations which are very similar to invoice.
  • Post meeting update: AP noted that the NumFocus organisation provides an umbrella non-profit that helps to support a number of high-profile scientific computing projects, including the big ones like numpy, scipy and julia, plus more ASE-sized ones like FEniCS, PyMC, Cantera, etc. They care of all the administrative work of running a non-profit, and take something like a 15% cut on donations to projects. There are small-scale grants available for things like running workshops, etc. This may well be easier than trying to setup our own non-profit, and could save us a lot of hassle.
  • ACTION JK and KWJ to look into options and propose a concrete route forward to the committee before the next meeting

Next meeting

  • Jan/Feb, following documentation hackathon and before in-person CECAM.
  • ACTION JK Send out a Doodle poll to schedule the next meeting.

Updated: