Open JWCook opened 1 year ago
It's supposed to be pinned to ^2.8
, but it looks like I neglected to do that for the conda-forge recipe. I'll get that updated.
As for moving to pyrate-limiter v3, it seems like there are a few features missing (which the maintainer is still working on), so I'm holding off on this issue until that's sorted out.
This has been fixed on conda-forge in v0.4.2.
This has been fixed on conda-forge in v0.4.2.
Great, thanks!
It's supposed to be pinned to
^2.8
, but it looks like I neglected to do that for the conda-forge recipe. I'll get that updated.As for moving to pyrate-limiter v3, it seems like there are a few features missing (which the maintainer is still working on), so I'm holding off on this issue until that's sorted out.
Hi JWCook - Thanks for maintaining this lib. Just curious if this can be upgraded to work with pyrate-limiter v3 now? :)
Thanks!
Probably! I will take another look soon to see if pyrate-limiter v3 has re-implemented all the features I need. If not, I'll submit some issues and/or PRs for whatever is missing.
Probably! I will take another look soon to see if pyrate-limiter v3 has re-implemented all the features I need. If not, I'll submit some issues and/or PRs for whatever is missing.
Look forward to it. Thanks very much!
The main thing still missing from v3 is the multiprocess
-compatible SQLite backend I originally submitted here: https://github.com/vutran1710/PyrateLimiter/pull/68. That's a pretty important feature that I and others are using via requests-ratelimiter
. I made an issue for that upstream.
It's also missing the contextmanager features I added here: https://github.com/vutran1710/PyrateLimiter/pull/25, although that's probably something I could work around.
As far as I can tell, pyrate-limiter v3 isn't finished: https://github.com/vutran1710/PyrateLimiter/issues/118. I'm willing to help get it there, but IMO the current releases on PyPI should have been published with a pre-release tag like -alpha
. Is there a specific feature it provides that you'd like to use? Just wondering if there's a short-term workaround I could provide.
Thanks for following Jordan - much appreciated!
I've managed to get a workaround for now, so it's not a problem. I just run in to issues when upgrading packages via pip as it results in some conflicts with other packages that I use. However that's a minor issue for now as I can manually sort it out by deleting the latest pyrate-limiter package and just reinstalling requests-ratelimiter, which then installs the appropriate version of pyrate-limiter.
Hopefully the upstream issues get resolved soon!
Thanks again for following up on this!
Is there a specific feature it provides that you'd like to use? Just wondering if there's a short-term workaround I could provide.
@JWCook to add to this discussion, I'm personnally interested by the new PostgresBucket backend.
My use case is a package running API calls but used in different pods/containers: I need to track this whatever the actual number of pods currently running. I was thinking of either trying to create a new custom bucket on a S3 storage system or using the PostgresBucket (which is not supporting async, and I'm not sure about the implications yet)...
Note : if not that, the package I'm speaking of has it's own way of tracking calls which is stored in a parquet file, which could easily be stored on a S3... But I'm way more in favor of using requests-ratelimiter, though I'm not sure that's (yet) possible...
Do you have any suggestions?
Has there been any progress on this issue? I am getting this error message: