Comparing the performance of brake with rates 10.000 and 100.000 there isn't much of a difference in throughput. I wonder why that is...
I'm implementing an internatl length option so that more than one byte is sent per data-event, if rate is too big and/or period too small.
Super old issue, but works well now with big data™:
$ brake 1000000 /dev/urandom | pv > /dev/null
^C75MB 0:00:08 [ 1MB/s] [ <=> ]
cool! and learned about pv