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
There have been slight performance improvements and newer packages provide better handling of version conflicts (this was actually an unintended side-effect in pkg_resources).
Let's go with importlib.metadata since this will be maintained as a part of Python.
* Adopt import.metadata (lablup/backend.ai#101)
* Instead of injecting "start-xxx" subcommands into the all existing
registered command sets, let CLI-enabled packages explicitly declare
their owner command set using dot-prefix.
(e.g., "ag.start-server" will be loaded into "ag" command group)
achimnol
added a commit
to lablup/backend.ai-manager
that referenced
this issue
Jan 15, 2020
Let's add a cache for entrypoints and provide an explicit refresh command to the cli package for speedup. Currently, regardless of which entrypoint loader is used, it takes about 0.8 to 1.5 seconds for the CLI to start up.
pkg_resources
is now going to be deprecated, and a recommended replacement isentrypoints
, but now the standard libraryimportlib
has adoptedimportlib_metadata
asimportlib.metadata
since Python 3.8 which is available to older Python versions as a PyPI package.There have been slight performance improvements and newer packages provide better handling of version conflicts (this was actually an unintended side-effect in
pkg_resources
).Let's go with
importlib.metadata
since this will be maintained as a part of Python.Related: #35
┆Issue is synchronized with this Asana task by Unito
The text was updated successfully, but these errors were encountered: