MicroHs: Haskell Implemented with Combinators | Hacker News Reader

MicroHs: Haskell Implemented with Combinators

github.com

38 pointstodsacerdoti20 comments

This looks great! Note that Lennart was a member of the original Haskell committee in the 80's and has written numerous compilers (including one still used at Standard Chartered). The combinator approach was originally used by David Turner's Miranda in the 80's and performed well on hardware of that era. Presumably embedded CPUs also favour the combinator approach.

I don't know how up to date it is, but the author of this article claims that the combinator-based solution is among the fastest:

https://willamette.edu/~fruehr/haskell/evolution.html?#combi...

Could someone explain the problem being addressed, and the solution, please?

It's the Hashell equivalent of MicroPython. A lean Haskell designed for embedded and other constrained systems.

That explains neither the problem nor the solution!

That should be enough information for you to find out the details yourself. I suggest you take a look at the MicroPython web site.

Hi there grumpyprole, so, this is a site where we interact as a community, and have discussions. You're replying as if you think it's StackOverflow. What I did was read TFA (the README) and determine that I could not understand from that resource what problem was being addressed or what the solution being proposed was. But I didn't need to start this thread at all if I was simply trying to find out what TFA was doing; I am capable of researching that myself.

Perhaps you're now thinking "well isn't it lazy of you to ask in this thread instead of researching it yourself?" And if you're thinking that then you misapprehend what this site is about. By asking the question, I was hoping that someone knowledgable who likes sharing their knowledge would contribute an explanation, and in the discussion thus-generated perhaps interesting things would occur or be learned, or at any rate it might be enjoyable to participants.

Hopefully you can see now that your replies are way off-base. If you don't want to explain this subject then that is perfectly fine; you don't need to tell us that.

To be honest I think you were the one originally off base here. If someone doesn't answer the question to your satisfaction then it would be more effective for you to rephrase the question rather than telling that person that they answered wrongly. The original author of the software has also failed to answer the question to your satisfaction, which is further evidence that it could have been phrased more clearly.

I don't think that I can get much clearer than my original phrasing. If someone doesn't know how to explain a piece of software in terms of what problem it's addressing and how it solves it then that person isn't the type of person I'm looking to discuss things with.

If you had put just half as much effort into articulating exactly what you did and didn't understand as you have in bashing me, then there might have been a discussion. Rather than lecturing me on how to participate in the community, you should perhaps reflect on how you yourself come across.

I was genuinely interested in this topic. I'm pretty disappointed that my straightforward request

> Could someone explain the problem being addressed, and the solution, please?

generated such a silly meta-conversation instead of a conversation about the technical matter at hand, and that the only person who responded to it was someone who was incapable of explaining technical concepts.

> the only person who responded to it was someone who was incapable of explaining technical concepts

More childish insults. You know absolutely nothing about me. At least I have contributed something in this thread. Again, reflect on your own behaviour and why you didn't get the outcome you wanted.

I know that this is how you communicate with people:

> That should be enough information for you to find out the details yourself. I suggest you take a look at the MicroPython web site.

So, I infer that you are patronizing, arrogant, and high-handed.

> If you had put just half as much effort into articulating exactly what you did and didn't understand

That's one of the several things you don't understand. My question

> Could someone explain the problem being addressed, and the solution, please?

was clear and polite. The fact that you didn't understand it reflects poorly on you. The fact that, given that you didn't understand it, you still chose to make noise reflects even more poorly on you.

> was clear and polite

It was a short open-ended question with no additional context. That is not a clear requirement for an answer. Did you even read the README? Do you know anything about Haskell or embedded systems? You wrote a single sentence then complained when I wrote just a few sentences back. That wasn't polite, so my subsequent response wasn't either.

Why do you think it necessarily addresses a problem? It's just a fun project. If someone finds it useful, so much the better.

Ah I just meant "the problem" in an academic sense. I.e. what is the thing which it is trying to do.

The thing that it is trying to do is to be a fun thing for me to hack on. This includes running it on microcontrollers, which is not easy with THC.

Thanks! What is it about your implementation that makes it more suitable for running on microcontrollers than GHC?

The runtime system is small and has very few dependencies. Binaries are quite small (when compressed).

Reply on news.ycombinator.com