In alpakka every connector a separate entity and implementation details varies from one to another.
In particular blocking vs nonblocking io and thread pool usage varies greatly between them.
When someone (like me) evaluates whether a particular connector fits requirements such information would be helpful. Some connectors, like s3 uses non-blocking io, whereas others like awslambda or sqs use bloking underneath.
Akka itself is very clear on dispatchers usage, and having similar information in alpakka could be beneficial as well - maybe adding a note in documentation related to connectors about underlying implementation is a good idea ?
Yeah it’s a good point.
Since Alpakka very much is a community driven and maintained thing at this point (we’re looking to get “full time” people looking after it soon though), it would be best if you can open a ticket about this and try getting such markers started in a few places, inviting other maintainers to mark their modules as well.
WDYT?