Memory allocator behavior (i.e. small objects grow, do they get copied to anon etc)
How to properly reference count using the kdb+ C API
Why are lambdas limited the way they are (96 constants etc)
Any strategies pertaining to minimize copies of data in q
How to write a language in k (i.e. walk us through s.k)
A detailed walk through of q.k code for executing a partitioned/segmented query showing the map reduce etc
Coverage of sd0/1 and best use of mainloop reads (I assume FH).
How does Arthur calculate his #s (i.e. http://www.kx.com/q/d/a/q.htm#Performance)
Behavior of SIGINT on -s or neg port?
Semantics of visibility of objects in namespaces - kind of know them but there seem to be some edge cases
Reminders for nuances like indenting to avoid line continuation etc (I'm always reminded when I'm declaring functions and I accidentally add white space to a new line and get a projection).
Misc esoteric performance tricks (i.e. {x*x*x}, max/min instead of any/all, avoiding implicit casts)
