some more information about their version of "slab allocation"
Oh the old "slab glorification". It was so much fun reading about it back then, huge deal was made out of minor detail (it's only used to manage allocation of certain sizes).
Needless to say the truth was only revealed when some actual benchmarking was made.
OS4 memory system proved to be as slow as snail.
It was adjusted later on, however, but the point is that just writing a lot of PR and using bunch of old concepts doesn't necessarily make a viable solution.
