2021 03 29 Open NEST Developer Video Conference

previous | list | next

Agenda

1. Welcome
2. Review of NEST User Mailing List
3. Project team round
4. In-depth discussion
   • CMake targets vs. CMake flags #1905 Microhackathon
   • Discuss dump/reload of connectivity (not checkpointing!)

Mailing list

• https://nest-simulator.org/mailinglist

Project team round

Here we discuss topics that need broader attention, for example questions that came up but are outside a single project's scope, larger planned changes/PRs that affect all teams or pending work that is blocked by external factors.

1. Models / NESTML
   • nothing major to report
2. PyNEST
   • no update
3. Kernel
   • nothing major to report
4. Installation
   • see in-depth
   • PRs in review
5. Infrastructure
   • no update
6. Documentation
   • working on multi-compartment model documentation
   • doc style guide PR in review, have a look and comment
7. EBRAINS
   • no update
8. (Feature) Random number generation
   • work ongoing
9. (Feature) Automated Testing
   • nosetests to pytest, planning hackathon
10. (Feature) Extension Module System
    • no updates

In-depth discussion

CMake targets vs. CMake flags #1905 Microhackathon? -> send doodle

• use cmake flags to configure what should be built and then just the usual make && make install
  • make html vs. in-system docs nest.help()
  • CMake targets vs. CMake flags #1905 Microhackathon
    • Discuss dump/reload of connectivity (not checkpointing!)
    • separate docs from the source build?
  • coupled to Python installation, where some changes are coming in at the moment
  • usually do cmake, make && make install. The make html seems odd.
  • readthedocs builds in source tree using direct sphinx-build, needs to be checked for compatibility.
• cmake_install_prefix set: do everything as before, else build pynest into current environment. See #1959 ProcessOptions.cmake
  • does not require a separate cmake flag

Discuss dump/reload of connectivity (not checkpointing!)

• Regarding RNGs: #1919
• very large connectivity matrices (~GB), HDF5 (sparse, dense, tree...)
• More open points to consider
  • API for connectivity (e.g. like in Arbor: callbacks for GID pairs)
  • connect more after reloading?
  • re-instanciate same network with custom synapse models?
  • more than just connectivity to be loaded, e.g. from AI applications?
• need to figure out different use-cases
  • use cases for robotics?
  • AI?
  • Learning?