-
Notifications
You must be signed in to change notification settings - Fork 1.9k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Use the same minor version number between boto3 and botocore #2702
Comments
Thanks for the feature request, @benkehoe. I opened a tracking issue to collect these types of issues around releases in the AWS Python ecosystem. |
Pinging about this, since it crops up fairly often for me. As an example, a recent change to botocore in 1.29.10 is causing issues in one of my projects benkehoe/aws-sso-util#83 and we could be living in a world where the corresponding boto3 version to communicate to that project's users was also 1.29.10, but instead it's 1.26.10. Also note the issue about improving documentation boto/botocore#2295 has been auto-closed, but I still don't think documentation is the best solution (given that boto3 1.30.0 could follow the last version of 1.26.x). |
Hi @benkehoe, wanted to provide a quick update that we've made changes to Boto3/Botocore with today's release. Both are now versioned at 1.33.0 and should remain in sync going forward. I unfortunately don't have a timeline for the CLI v1 sharing the same version number, but the CLI team is looking into supporting it. |
|
This is a regular bump of Amazon-provider related dependencies. The way how botocore releases are done, they are putting a lot of strain on `pip` to resolve the right set of dependencies, including long backtracking, when there are too man versions available. Therefore, from time to time, we are bumping minimum version of Amazon-related dependencies to limit the impact frequent releases of boto and botocore has. Also it is generally fine to update min version of dependencies for providers because at the very least users can still use previously released providers in case they have problem with those dependencies, also many of the updated dependencies contain fixes and feature we implicitly depend on and bumping them regulary is a good way to make sure all the functionalities of the Amazon provider are working as expected. Another reason for the bump is that as of 1.33 version botocore and boto version stopped being shifted by 3 (previously boto3 1.28 was the version corresponding to botocore 1.31). As of version 1.33 this problem has been solved. See boto/boto3#2702 Watchtower min version is bumped to version 3 (which is 12 months old even if before we opted for much older (more than 2 years old) and again if users want to use older version of watchtower, they can opt for previous provider version. This change saves 5-6 minutes of backtracking when `pip` try to find the right version of dependencies when upgrading to newer version. Extracted from apache#36537
This is a regular bump of Amazon-provider related dependencies. The way how botocore releases are done, they are putting a lot of strain on `pip` to resolve the right set of dependencies, including long backtracking, when there are too man versions available. Therefore, from time to time, we are bumping minimum version of Amazon-related dependencies to limit the impact frequent releases of boto and botocore has. Also it is generally fine to update min version of dependencies for providers because at the very least users can still use previously released providers in case they have problem with those dependencies, also many of the updated dependencies contain fixes and feature we implicitly depend on and bumping them regulary is a good way to make sure all the functionalities of the Amazon provider are working as expected. Another reason for the bump is that as of 1.33 version botocore and boto version stopped being shifted by 3 (previously boto3 1.28 was the version corresponding to botocore 1.31). As of version 1.33 this problem has been solved. See boto/boto3#2702 Watchtower min version is bumped to version 3 (which is 12 months old even if before we opted for much older (more than 2 years old) and again if users want to use older version of watchtower, they can opt for previous provider version. This change saves 5-6 minutes of backtracking when `pip` try to find the right version of dependencies when upgrading to newer version. Extracted from #36537
This is a regular bump of Amazon-provider related dependencies. The way how botocore releases are done, they are putting a lot of strain on `pip` to resolve the right set of dependencies, including long backtracking, when there are too man versions available. Therefore, from time to time, we are bumping minimum version of Amazon-related dependencies to limit the impact frequent releases of boto and botocore has. Also it is generally fine to update min version of dependencies for providers because at the very least users can still use previously released providers in case they have problem with those dependencies, also many of the updated dependencies contain fixes and feature we implicitly depend on and bumping them regulary is a good way to make sure all the functionalities of the Amazon provider are working as expected. Another reason for the bump is that as of 1.33 version botocore and boto version stopped being shifted by 3 (previously boto3 1.28 was the version corresponding to botocore 1.31). As of version 1.33 this problem has been solved. See boto/boto3#2702 Watchtower min version is bumped to version 3 (which is 12 months old even if before we opted for much older (more than 2 years old) and again if users want to use older version of watchtower, they can opt for previous provider version. This change saves 5-6 minutes of backtracking when `pip` try to find the right version of dependencies when upgrading to newer version. Extracted from #36537 (cherry picked from commit 298c37d)
This is a regular bump of Amazon-provider related dependencies. The way how botocore releases are done, they are putting a lot of strain on `pip` to resolve the right set of dependencies, including long backtracking, when there are too man versions available. Therefore, from time to time, we are bumping minimum version of Amazon-related dependencies to limit the impact frequent releases of boto and botocore has. Also it is generally fine to update min version of dependencies for providers because at the very least users can still use previously released providers in case they have problem with those dependencies, also many of the updated dependencies contain fixes and feature we implicitly depend on and bumping them regulary is a good way to make sure all the functionalities of the Amazon provider are working as expected. Another reason for the bump is that as of 1.33 version botocore and boto version stopped being shifted by 3 (previously boto3 1.28 was the version corresponding to botocore 1.31). As of version 1.33 this problem has been solved. See boto/boto3#2702 Watchtower min version is bumped to version 3 (which is 12 months old even if before we opted for much older (more than 2 years old) and again if users want to use older version of watchtower, they can opt for previous provider version. This change saves 5-6 minutes of backtracking when `pip` try to find the right version of dependencies when upgrading to newer version. Extracted from #36537 GitOrigin-RevId: 298c37d355eeadfccbd655efb2922d39ba17052c
This is a regular bump of Amazon-provider related dependencies. The way how botocore releases are done, they are putting a lot of strain on `pip` to resolve the right set of dependencies, including long backtracking, when there are too man versions available. Therefore, from time to time, we are bumping minimum version of Amazon-related dependencies to limit the impact frequent releases of boto and botocore has. Also it is generally fine to update min version of dependencies for providers because at the very least users can still use previously released providers in case they have problem with those dependencies, also many of the updated dependencies contain fixes and feature we implicitly depend on and bumping them regulary is a good way to make sure all the functionalities of the Amazon provider are working as expected. Another reason for the bump is that as of 1.33 version botocore and boto version stopped being shifted by 3 (previously boto3 1.28 was the version corresponding to botocore 1.31). As of version 1.33 this problem has been solved. See boto/boto3#2702 Watchtower min version is bumped to version 3 (which is 12 months old even if before we opted for much older (more than 2 years old) and again if users want to use older version of watchtower, they can opt for previous provider version. This change saves 5-6 minutes of backtracking when `pip` try to find the right version of dependencies when upgrading to newer version. Extracted from #36537 GitOrigin-RevId: 298c37d355eeadfccbd655efb2922d39ba17052c
This is a regular bump of Amazon-provider related dependencies. The way how botocore releases are done, they are putting a lot of strain on `pip` to resolve the right set of dependencies, including long backtracking, when there are too man versions available. Therefore, from time to time, we are bumping minimum version of Amazon-related dependencies to limit the impact frequent releases of boto and botocore has. Also it is generally fine to update min version of dependencies for providers because at the very least users can still use previously released providers in case they have problem with those dependencies, also many of the updated dependencies contain fixes and feature we implicitly depend on and bumping them regulary is a good way to make sure all the functionalities of the Amazon provider are working as expected. Another reason for the bump is that as of 1.33 version botocore and boto version stopped being shifted by 3 (previously boto3 1.28 was the version corresponding to botocore 1.31). As of version 1.33 this problem has been solved. See boto/boto3#2702 Watchtower min version is bumped to version 3 (which is 12 months old even if before we opted for much older (more than 2 years old) and again if users want to use older version of watchtower, they can opt for previous provider version. This change saves 5-6 minutes of backtracking when `pip` try to find the right version of dependencies when upgrading to newer version. Extracted from #36537 GitOrigin-RevId: 298c37d355eeadfccbd655efb2922d39ba17052c
Currently, boto3 version
1.16.x
depends on botocore version1.19.y
. Every time the botocore minor version is bumped, the boto3 minor version is bumped as well. Could the next minor version of boto3 be1.20.0
to depend on botocore version1.20.0
? For those of us who deal in the details inside of the SDK, figuring out what boto3 version to support based on a feature in the botocore code would become simpler.The text was updated successfully, but these errors were encountered: