Conversation
hartmans
left a comment
hartmans
left a comment
There was a problem hiding this comment.
I don't think this is right. There are cases where a message will be sent
with _no_resp_for and no _sync_type. I.E. a metadata only message.
There are likely future cases where you will get a message without a sync_type, for example as we're adding more advanced bandwidth management.
I guess a valid check here might be whether there are any keys that don't start with underscore or something. or probably more correct and more efficiently have handle_meta return whether it did anything.
hartmans
left a comment
hartmans
left a comment
There was a problem hiding this comment.
When do we end up without a _sync_owner that we're trying to send?
Are we confident that introducing this change does not create a security problem with the checks in operations.py?
Also, i don't understand the diff to the setter; is it some sort of whitespace change, and if so, what's going on?
2 participants
No description provided.