Russ Bradberry
Russ Bradberry
its hard to tell from this single file, but it might be something in `Vtap`? it seems you reference it in the second post but it doesn't seem to be...
this shouldn't be the case. the api for all calls should be the same.
What is the main concern here? While it binds on all ports by default, it is only [sending out on localhost](https://github.com/sivy/node-statsd/blob/master/lib/statsd.js#L211). Additionally, would you be willing to create a pull...
hmm, I don't think it should just be "handled" the only thing that would make sense would be to make `StatsD` an emitter and bubble the error emission up. Although...
yeah, it's bad API design on my part for forcing the breakage of LoD ... and then blatantly documenting it. heh. If I have some time in the next week...
What is the purpose of this? UDP doesn't do any handshaking, so there is no performance gain by batching these calls. From what I can tell it does not much...
Fair enough. It looks like the `.close` method clears the timeout for sending batches, however, does not check to see if there are messages waiting to be sent and send...
I think so, having those be configurable would be optimal.
this looks good to me @sivy thoughts?
if your goal is to come in less than the MTU then shouldn't the packet header lengths be taken into account?