v2.1.0

What's Changed

• Update to podping-schemas-python 1.0.1 (based on podping-schemas v2024.06.16.060318) which adds support for publisher and course mediums
• Update Hive node list for more stability and some new nodes
• Miscellaneous dependency updates

Full Changelog: v2.0.0...v2.1.0

v2.0.0

This is a major release that breaks the ZeroMQ server interface by utilizing a new binary format. The CLI interface has not changed.

The most important item to note is the support of Podcasting 2.0 playlists via list mediums.

What's Changed

• Add list mediums as detailed in Podcastindex-org/podcast-namespace#363 by @agates
• Add helper method PodpingHivewriter.send_podping to make it easier to integrate the server into other python projects by @agates
• Python 3.11 support by @agates
• Use plexo, new binary message formats, and zmq pair in server mode by @agates in #61
• Support writing mediums & reasons from ZMQ listener, batch up to 5 operations into each transaction by @agates in #67
• Moved schemas into separate external project podping-schemas by @agates
• Feature/return hive txid by @agates in #60
• Broadcast Hive transaction information across ZMQ listener, allowing clients such as podping.cloud to know when IRIs have been successfully completed by @agates

Full Changelog: v1.2.10...v2.0.0

v1.2.10

What's Changed

• Remove 3speak and openhive nodes for stability reasons by @agates

Full Changelog: v1.2.9...v1.2.10

v1.2.9

What's Changed

• Remove follower check in startup & re-add more hive nodes by @agates in #66

Full Changelog: v1.2.8...v1.2.9

v1.2.8

What's Changed

• Add more logic to handle bad Hive API nodes by @agates in #65

Full Changelog: v1.2.7...v1.2.8

v1.2.7

What's Changed

• Upgraded to lighthive 0.4.0. This includes a more intelligent load balancing and circuit breaking system which blacklists bad nodes for an hour before retrying again - @agates
• Properly handle resource credit exhaustion and backoff retrying requests when it occurs. In general, retries and error handling are better in combination with the lighthive upgrade - @brianoflondon and @agates
• Temporarily removed resource credit calculations on startup during --sanity-check. It will still error out on startup if the account does not have enough resource credits, but it wont try to tell the user how many posts they have capacity for - @brianoflondon
• Made startup error handling more resilient + consistent and added some more exit codes for different failure cases. These are listed in the constants.py file - @agates
• Add the ability to override the Hive operation period with --hive-operation-period. This is primarily useful for services that don't have high throughput and want to more efficiently batch podpings - @brianoflondon
• Re-added the ability to use a testnet - @brianoflondon

Full Changelog: v1.2.6...v1.2.7

v1.2.6

No-code release to fix python packaging (again).

v1.2.5

What's Changed

• Correctly handle non-standard RPC error response from Hive RPC node by @brianoflondon and @agates in #52
• Base container image on docker.io/python:3.10-bullseye instead of docker.io/pypy:3.8-bullseye. We think pypy's memory allocator is causing issues with container usage.

Full Changelog: v1.2.4...v1.2.5

v1.2.4

No-code version fix release

(because fighting with arm64 build issues in CI/CD makes one forget about everything else important)

v1.2.3

Fix CLI launch due to some dependency version mismatching.