Kevin Amado added the comment:
Yeah definitely it must be workers
I've experimented a lot about it and finally found something with an interface
similar to asyncio.as_completed
- You control concurrency with `workers` parameter
- You upper-bound memory usage with `worker_greed
Change by Kevin Amado :
Removed file: https://bugs.python.org/file49377/materialize-implementation.py
___
Python tracker
<https://bugs.python.org/issue41505>
___
___
New submission from Kevin Amado :
Sometimes when dealing with high concurrency systems developers face the
problem of executing concurrently a large number of tasks while taking care of
a finite pool of resources
Just to mention some examples:
- reading asynchronously a lot of files without
