Tim Watson

Thanks for spotting that @simonmar - it looks like a simple (ish) fix, so I'll have a look at it over the next day or so.

@qnikst @facundominguez ping - is this valid? I would've thought we should kill off both the NC threads and thereby provide a means for the hosting/containing application to know our...

And observed this evening (looking at HEAD) - ./dist/build/TestStats/TestStats (which does _barely anything_ at all) can lock up **if** you run it in a tight loop for long enough. Around...

Erm, `getNodeStats` I mean, I got confused with `getProcessInfo` - I'll update the title later...

On the point about start specs: the Supervisor process needs to spawn these other Processes, and must be able to restart them. It therefore needs to have the clousure information...

Yes, starting with the type definitions is exactly what I want to do. One of the challenges here is knowing how to distinguish between a normal and an abnormal exit...

Hi! Thanks for that feedback, it's very useful. Some thoughts inline: > First of all, each behavior could be in it's own module. This makes sense since it's one API...

@edsko - thanks for this! > Just a few brief remarks: > - You mention that "Also, in Cloud Haskell, a link between the child and supervisor will cause both...

@rodlogic - you make a good point about the gen_server basis for OTP's supervisor. But.... The gen_server concept is quite different to implement in Haskell. Differentiating between for example handle_call,...

> It's not obvious to me what the main problem is that makes it difficult to port Erlang's ideas to Cloud Haskell; is it the > problem that we need...