Now that Artiq 8 has been released, are the Artiq 7 conda packages (specifically the Win-64 ones) no longer available on the m-labs conda channel?

As a note for anyone who might find this message due to issues with installing Artiq 7 via Conda, the packages at the moment aren't installed properly into the environment. This can be fixed by copying the artiq, artiq_comtools, pythonparser, and sipyco folders from <env>\site-packages to <env>\Lib\site-packages

Github issue for this bug: https://github.com/m-labs/artiq/issues/2398

4 days later

there are some extra constraints to take into account when installing artiq with conda because there is no version pinning. for artiq 7, we match the python version in nix, make sure numpy stays constrained, and constrain pip too due to enforcement of version number formatting that artiq 7 does not adhere by. but yeah, this is conda, so this list of constraints can get longer any moment...

channels:
  - https://conda.m-labs.hk/artiq-legacy
  - conda-forge
dependencies:
  - python=3.9
  - artiq=7
  - numpy<2 # Constrain version for potential compatibility issues
  - pip=24.0 # Added for package install through pip, constraint version due to compatibility issue
    6 days later

    lriesebos constrain pip too due to enforcement of version number formatting that artiq 7 does not adhere

    Is that still an issue? I remember fixing a bunch of silly pip complaints relatively recently.

      sb10q you did, but that was for ARTIQ 8. ARTIQ 7 still has a pip-incompatible version number, and the recent pip update now errors because of that. since pip is not constrained by the ARTIQ 7 package in conda, conda chooses the latest version of pip which causes issues.

      interesting enough, I think ARTIQ 7 and the latest pip still install fine with conda in a single environment. though if you then invoke pip in that environment, it errors because of the ARTIQ 7 version number.

      anyway, this does not need to be fixed for me, but might be good to know for ARTIQ 7 + conda users. without the extra constraint they might run into issues when invoking pip in their environment.