A few statuses can change (which is normal), and the system reacts, or they can ALL change, and it will react just the same (which is abnormal). We could have lots of targets suddenly disappear .... break ...
-
Show this thread
-
"I felt a great disturbance in the Force, as if millions of voices suddenly cried out in terror and were suddenly silenced. I fear something terrible has happened."
1 reply 0 retweets 7 likesShow this thread -
... end break ... and the system does not care or even know how many changed! It just works. Very very reliable, and unde-rapreciated.
1 reply 0 retweets 9 likesShow this thread -
BONUS CONTENT: O.k., what does this have to do with Maxwell's Daemon? WHAT EVEN IS THAT? Well the first kind of constant time problem, for crypto, is an example of minimizing information theoretical entropy. It's about minimizing perceptible disorder and what is computable.
1 reply 1 retweet 6 likesShow this thread -
The second is more like traditional thermodynamic entropy. With a constant-time control plane, we're trying to build a system that has a constant temperature, because it's always doing the same amount of work.
1 reply 0 retweets 9 likesShow this thread -
Ok now if I lost you with that, here's the simple version: when systems change a lot, they undergo stress, and getting stressed at the worst times, like under attacks or in the middle of outages, is really bad. At a high level these problems are the same.
1 reply 0 retweets 5 likesShow this thread -
And it turns out at a low level they are the same too! Information theoretical entropy, which is all about computability and bits, can be related to thermodynamic entropy, which is all about energy and work!
1 reply 0 retweets 4 likesShow this thread -
How? Well Maxwell proposed a thought experiment that showed a problem with thermodynamics. He imagined two rooms side by side, at even temperatures. He posited that a daemon, a ghost, could open a door between the two rooms and let faster moving molecules over to one side.
1 reply 0 retweets 3 likesShow this thread -
This sort of disproved existing thermodynamics, because the work it takes to open a door like that is less than the energy transferred due to the faster molecules changing sides. This was unresolved for a long time ... UNTIL ...
1 reply 0 retweets 4 likesShow this thread -
A bunch of folks showed that you don't just open the door, you have to observe and predict, and therefore COMPUTE, the position of the molecule. THIS TAKES ENERGY ... and they showed that the energy it takes is at least equivalent to the difference.
2 replies 1 retweet 7 likesShow this thread
So in a big old nuts, strange-loop sort of way, these completely unrelated things are totally the same at both a MACRO and MICRO level. This blows my mind .... but may also make no sense ;-)
-
-
O.k. that's it, and AMA if you want! Oh and
@threadreaderapp please unroll this mess I made.1 reply 0 retweets 11 likesShow this threadThanks. Twitter will use this to make your timeline better. UndoUndo
-
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.