Time  Nick   Message
19:23 hmmmm  right, so as i was saying
19:24 hmmmm  (timestamp - block.timestamp - (server_start - server_last_start)) / (interval * chance) times
19:25 hmmmm  obviously this would create a lag on loading a block
19:25 hmmmm  from disk
19:25 hmmmm  so perhaps there ought to be an upper bound to the amount of times the ABM is called
19:25 hmmmm  especially if it's Lua
19:26 hmmmm  do mobs depend on ABMs to function?  I am enitrely unknowledgable of how mobs work currently
19:26 thexyz why do we need to call it several times at server startup?
19:27 hmmmm  noo, noo, not on server startup, on block load
19:27 hmmmm  the whole reasoning behind this is so that the node timers can catch up
19:27 thexyz oh
19:28 thexyz well, then there should definetly be some upper limit
19:28 hmmmm  you lay down a bunch of dirt, go do something else, that block unloads, you come back a bit later, and grass actually 'grew' in between the time the chunk was unloaded and you came back
19:28 thexyz better to make it configurable
19:28 hmmmm  of course
19:28 hmmmm  but this is all just my idea
19:28 hmmmm  i'm proposing it, i'm not saying "this is what i'm adding"
19:28 hmmmm  i'm sure certain people would have objections
19:29 thexyz that's why it should be configurable
19:29 thexyz so they'd have nothing to object about
19:30 hmmmm  a feature being configurable or disable-able is not an excuse for it to be added though
19:30 hmmmm  it might require changes in code structure or assumptions that one can make
19:30 hmmmm  which would piss off other coders
19:31 hmmmm  but anyway for the upper limit of number of calls, i propose that instead it'll keep track of time deltas in between calls to the ABM so that it'll tally it up
19:31 hmmmm  so you'd be be able to configure it by saying, "leaf ABM gets 500ms of cpu time max"
19:32 hmmmm  it'd have to be a bit more complicated than a time delta though, since you'd have to take into account for context switches
19:32 hmmmm  or not, i don't know