Louis Jenkins

Hey @e-kayrakli , I found another bug that relates to type inference from the compiler... Although to be sure, I'd need to figure out how to query more information about...

> Are you sure you don't have a race in trunk as of now. I checked the line initialize a CCQueue. When I run it on my machine within a...

For emphasis, if you see deadlock, try it both when you change the type of the queue to `Queue` and then to `CCQueue`, furthermore when its `Queue` and `writeln(retval)` is...

Also, I realized now that `cclock` is still using first class functions. I never changed it because I need to hard-code it to a specific use-case (like CCQueue) and it...

NQueens itself does not contain a deadlock, as it only spins like that *inside* of `DistributedFIFOQueue`, meaning the deadlock would be inside `DistributedFIFOQueue`. I do see now that there is...

Hey Engin, in terms of removing first class functions in it's entirety (and to get it out of the way, with hopeful benefit of eliminating that undefined behavior bug), I...

Hm, actually perhaps I should just go ahead and use sync vars for now. CCLock would be an optimization in and of itself, so I can insert/inject that in *after*...

~~Seems the issue has been fixed by changing the domain to be 0-based, Thanks Engin!~~ Hm, nevermind, race condition is still there... Edit: So, now I get a *compiler* internal...

I'm thinking... I know you mentioned a lot I should try to keep everything self-contained (especially for `CCLock`, of which I'll agree 100%), and I'm thinking that... maybe I should...

So, I was messing around with a prototype that just checks if the domains match at _runtime_ and then indexes via domain based on this, and I'm getting some weird...