Something something TCP orgs vs UDP org (except former never give up on a message getting through)
-
-
Show this thread
-
Imagine a stream of flowing information that everybody needs to fully process within a fixed time, say 24h, to do a project. Now imagine 2 failure modes: a) everybody delayed to by +T on every message b) everybody in Lucy mode: keeping up X%, but dropping 100-X% the messages
Show this thread -
The +T likely compounds in dependent sequences +T, +2T, +3T etc. I gotta think about how to model this properly. The problem I’m interested in framing is: what sorts of projects degrade better under delay vs drops? And how do you mitigate each?
Show this thread -
I *think* redundant comms and embarrassingly parallel actions address droppy/lossy processing vs delayed. In human projects this means repeating yourself a lot, and redundant staffing. Fo delay, you just add time. If cost of adding enough time exceeds value of project, kill it.
Show this thread
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.