Merge fix for node stall ready search

[BLuedtke]

Closes #28

Draft because tests still have to be run on how this affects overall operation. More info available in issue #28

[BLuedtke]

Looking at it again, it is now clear to me why this worked before the fix.

The point of searching the stall_affected_nodes_queue was to see if a node was already in there - if not, add it. Later, in bidib_node_update_stall, when the node is not stalled anymore, all entries in the stall_affected_nodes_queue are popped and for each entry, we try to send any queued messages.

With the old code where the search function for the stall_affected_nodes_queue did not work, it would always return false -> so the node might be added to the stall affected nodes queue multiple times. But that just means that when the node is not stalled anymore, and it has multiple identical entries in the stall_affected_nodes_queue, it just tries to send queued messages multiple times, which is no problem.

So the fix only prevents the same node from being added to the stall affected nodes queue multiple times, nothing else.

[eyip002]

Changes look good. Minor suggestion for a comment.

[BLuedtke]

Huh? Not sure why it says I dismissed the review. Let me see if I can revert that.

Edit: Can't see where I can revert it. Makes no sense, not sure how this happened. GitHub is acting weird wrt. merge requests recently, sometimes it links completely unrelated issues automatically.