Peter Dillinger

This area of the language is rare enough not to warrant significant extra complexity just for some code brevity. Brevity, particularly in rare contexts, can of course be the enemy...

> a general concern with lambdas in Pact is the possibility of forming a Y-combinator, does this "immediate application" address that? Yes it does. You have to allow binding symbols...

Proposal: In order to cleanly support `do` and to clean up support for a number of other natives/special forms, I recommend adding essentially varargs support to the function type system....

> Yeah I'd go with new functions with a "v2" suffix for fixing up APIs like this (adding `char** errptr` to APIs that can fail). cc @pdillinger, @siying, @ltamasi in...

Old implementation with this bug has been deleted. HyperClockCache is now available as a much faster alternative (see API comments for caveats). I still need to update wiki and write...

Doing a MultiGet benchmark with db_bench is user-hostile, because defaults for --multiread_batched and --batch_size work against you.

I don't think we should introduce small _passive_ performance regressions (affecting performance when new feature is not enabled) for features that haven't been proven necessary. One possibility that I'm not...

> Personally, the change doesn't strike me as something that would introduce a significant overhead; then again, I might be turn out be wrong on this one, in which case...

My test shows an unexplained small improvement (~2% ops/sec on a benchmark with 50% CPU in LRUCache) with the feature present but disabled, and ~9% overhead with the feature enabled....

@mrambacher Is this something you can look into? The race condition seems rather "baked in" to the Configurable interface with GetOptionsPtr.