You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Would it be feasible to add a "publish_date ≤" filter so that mamba would effectively invoke the solver as if it were running on a date in the past?
It's a common problem that beginners forget to conda-lock or pip freeze their requirements, and their code no longer runs with up-to-date versions. In such situations, such a filter would be extremely useful, effectively acting as a time machine.
Assuming the solver is more-or-less fed the contents of repodata.json, would it simply be a matter of filtering on the timestamp property? (I checked, and 97.6% of packages in conda-forge's noarch repodata have the timestamp property. The earilest timestamp is from April 18, 2018.)
Would it be feasible to add a "publish_date ≤" filter so that mamba would effectively invoke the solver as if it were running on a date in the past?
It's a common problem that beginners forget to conda-lock or pip freeze their requirements, and their code no longer runs with up-to-date versions. In such situations, such a filter would be extremely useful, effectively acting as a time machine.
Assuming the solver is more-or-less fed the contents of
repodata.json
, would it simply be a matter of filtering on the timestamp property? (I checked, and 97.6% of packages in conda-forge's noarch repodata have the timestamp property. The earilest timestamp is from April 18, 2018.)I originally asked about this in:
conda/conda-lock#112
The text was updated successfully, but these errors were encountered: