Welcome to MicroHs Discussions!

[augustss]

👋 Welcome!

We’re using Discussions as a place to connect with other members of our community. We hope that you:

• Ask questions you’re wondering about.
• Share ideas.
• Engage with other community members.
• Welcome others and are open-minded. Remember that this is a community we
  build together 💪.

To get started, comment below with an introduction of yourself and tell us about what you do with this community.

[Ebmtranceboy]

I'm looking for ways of communication between mHs and the HTML of the browser. I have mongoose in mind, and at the moment, I'm testing mHs FFI to call mongoose API. Zero or one argument works fine, even with structures and/or several layers of pointers. Yet, I'm wondering if that is possible since I can't even call a simple function with more than one argument (like int add(int a, int b)).

[konsumlamm]

I think the problem is that you're using Int instead of CInt (from the Foreign.C.Types module) in the foreign import signature. Haskell's Int does not correspond to C's int.

[Ebmtranceboy]

I've tried with CInt, with Ptr Int, Ptr CInt, Int8 and probably some other arguments but I've never had a successful build.

[konsumlamm]

Ptr Int, Ptr CInt and Int8 are definitely wrong. Are there any warnings? Unfortunately, I can't reproduce the issue, which leads me to believe that this is architecture specific. Have you tried using ccall instead of capi? Either way, I suggest opening an issue.

[augustss]

Please open an issue.

[Ebmtranceboy]

I've tried Ptr a with a modified version of mySrc.c of course and the use of alloca in the calling hs file. I generally have no warning except at the moment of showing the result when I'm using for instance Int64. I have also tried to use ccall. Thanks, I've opened an issue (edit: issue closed)

[konsumlamm]

Hi, I'm konsumlamm. I think it's nice to have some competition to GHC and I have fun contributing to MicroHs. Currently, I'm trying to get more packages to work with MicroHs and fix/report all the bugs I find on the way ;)

[HeinrichApfelmus]

Hi! 👋 I'm Heinrich. My main interest in MicroHs also comes from JavaScript/WebAssembly as compilation target. At the same time, I like the competition to GHC: MicroHs is (still?) simple and small enough that it seems reasonable for me to modify the source code; it's the principle of the thing.

[arossato]

Hi, my name is Andrea. I don’t have a CS background: professionally I do legal research and teaching in what used to be called cyberlaw. I fell in love with Haskell many years ago, and was lucky enough to collaborate with some exciting projects along the way, which became a great learning experience.

When I discovered mhs I was literally blown away: a bidirectional type-checker in ~4k LOC, which opens up the possibility, for someone like me, to seriously engage with literature I had always thought was out of reach. So far my contribution is almost non-existent — I’m very much still in learning mode.

I’d like to thank @augustss for opening GitHub Discussions: it feels like exactly the right place to talk about design ideas and questions without turning everything into an issue or a PR. Lately I’ve been particularly intrigued by the recent higher-rank work, the removal of MonoLocalBinds, the monomorphism restriction, and the design choices behind those changes.

More practically, I see this implementation as a breath of fresh air for Haskell in general, especially given the current GHC monoculture. My plan is to use mhs as the primary platform for the only project I’m presently (slowly) developing: Haskell bindings to the Arcan display server shared-memory and text-UI interfaces.

[augustss]

The change from MonoLocalBinds to MonomorphismRestriction is just to be compatible with GHC. I think the former is the better choice, but I want more packages to compile. So I made the change.

If you don't have either of them you are in a bit of trouble. Any overloaded local let then risks being evaluated multiple times.