The language also has a pretty good manual: https://hobbes.readthedocs.io/en/latest/ that also give a bit more background on why this was developed in the first place. Imagine being able to answer real time what is going inside an application engine, even in cases that may involve complicated state. Then doing this without affecting the latency constraints of the critical path. Or being able to change the runtime behavior of an application without sacrificing latency at all while still having type level guarantees that what you are doing is safe. There are quite a few solutions that allow this two requirements but hardly any in a domain where zero overhead really must be zero and not just a vague marketing term. For those that ask KDB does not even qualify on the second requirement while hardly meeting the real time for the monitoring requirement.
The language also has a pretty good manual: https://hobbes.readthedocs.io/en/latest/ that also give a bit more background on why this was developed in the first place. Imagine being able to answer real time what is going inside an application engine, even in cases that may involve complicated state. Then doing this without affecting the latency constraints of the critical path. Or being able to change the runtime behavior of an application without sacrificing latency at all while still having type level guarantees that what you are doing is safe. There are quite a few solutions that allow this two requirements but hardly any in a domain where zero overhead really must be zero and not just a vague marketing term. For those that ask KDB does not even qualify on the second requirement while hardly meeting the real time for the monitoring requirement.
For those interested, previous discussions here:
https://news.ycombinator.com/item?id=14783539
https://news.ycombinator.com/item?id=21892432