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
@barabo standalone conda-index is fast as long as stat'ing all the files is not a bottleneck (which it absolutely was for conda-forge on slow storage). It's designed so that you can subclass the part that does a listdir() and swap it out for a database query - listdir populates the database with a list of "currently available" files, then the index joins that list against the cached metadata for those files. It has to pull the index from its database, never from an existing repodata.json.
I haven't consulted the thesaurus to find a good replacement for the "index" name currently provided by conda-build, if we want to use the new, different click CLI.
Checklist
What is the idea?
Conda has a new plugin system through which we can provide "conda index" (without using entry_points scripts). Provide one.
Will still need to figure out what to do about conda-build's "conda-index" script, which can also be called as "conda index".
Our command line arguments are necessarily different than legacy conda-index.
Why is this needed?
No response
What should happen?
No response
Additional Context
No response
The text was updated successfully, but these errors were encountered: