This is a version bump and should not adversely affect existing applications that use the hive-ruby gem. Changes are focused on performance improvements to Hive::Stream subroutines.
The main change is that we no longer have to rely on json-rpc-batch to stream blocks. Prior to this change, if the API node supported batched requests, it would ask for up to 50 blocks using this published standard.
But not all API endpoints supported the standard. In HF24, we got the block_api.get_block_range method. And although not all endpoints support this method, it's much more supported than json-rpc-batch.
Hive Ruby will try whatever the endpoint will provide, starting with get_block_range, then json-rpc-batch, and then just fall back on non-batched requests.
In addition to this new method, if you area streaming virtual operations, Hive Ruby will now use account_history_api.enum_virtual_ops (if supported). This will improve performance and reliability, even when streaming virtual operations from the head block.
See rev.: 94142bc
To update your application:
bundle update hive-ruby
See previous hive-ruby posts: peakd.com/c/hive-139531/search?q=hive-ruby