Remove the rule chains?

[Sean Kauffman]

Right now we store two linked lists of rules in a specification to speed up access to the rules that subscribe to a certain event label. This probably improves performance a tiny bit over just iterating over all the rules each time, but it also adds significant complexity and the difference is likely virtually nil. Removing the linked list, and just iterating over all the rules, is probably a better choice.

When trying this, though, we should definitely test it with large specifications and see what the difference is in performance.

[Sean Kauffman]

This is really not worth the effort. There's nothing inherently wrong with how it's done and changing it might slow things down slightly.

[Sean Kauffman]

I ended up doing this while fixing the exponential blowup bug.

[Sean Kauffman]

This is nice because it simplifies a lot of things but hopefully it doesn't end up hurting performance for large specs too much. Small testing I performed seems to show it doesn't matter much.