01:01 <jemc> SeanTAllen: I didn't think that was true...

01:02 <jemc> for primitives that is - you definitely can't define methods on a type alias

01:02 <jemc> yeah, this works:

01:02 <jemc> primitive One fun add(u: I64): I64 => u + 1

01:02 <jemc> actor Main new create(env: Env) => env.out.print((One + 1).string())

01:02 <jemc> (prints `2`)

01:03 <jemc> you could also define the infix operator named methods on a `trait` (and use that to "inject" them into a new `class` or `primitive`)

01:03 <jemc> dougmacdoug, maybe that helps you? ^

01:04 <jemc> ah, just read the rest of the backlog before the latest question - now I understand that what you basically wanted is to define those methods on a tuple, which you can't do

01:12 <dougmacdoug> right

01:14 <jemc> yeah, if you want to keep the lightweightedness of tuples, you'll have to use the more "functional" style

01:14 <dougmacdoug> I've been working out a vector/linear algebra lib with tuples, for performance.. i am making progress .. but its quite verbose.. would be really nice to have infix.. also override methods always creating a none return means i need to give each method a diff name..

01:14 <jemc> "override methods always creating a none return"?

01:15 <jemc> ah, case methods, right

01:15 <dougmacdoug> like add2() add3() add4().. if I do add() for each vec type I get a union return .. anyway, i'll just keep chipping away and get more familiar with the language

01:15 <jemc> case methods are definitely a little rough around the edges when it comes to being truly useful

01:16 <jemc> especially if you're interested mainly in performance, I'd stick to using different method names for now, as you're doing

01:16 <dougmacdoug> yeah theres a comment in the tutorial that the compiler doesnt do an exhaustive search so even if you cover every case you still get union none

01:17 <jemc> we've got a couple new features coming down the pipeline (they've cleared the RFC process, and are ready to implement) that would give you something similar to case methods, but as a zero-cost abstraction that should also be able to get rid of the union / None return issue

01:17 <dougmacdoug> yes I think stack tuples will be far better than making classes but it will require more typing :p

01:19 <jemc> these are the issues to watch for the new features I'm talking about: https://github.com/ponylang/ponyc/issues/1184 https://github.com/ponylang/ponyc/issues/1410

01:20 <jemc> still won't be able to define methods on tuples, but you can have a better solution than case methods for methods-with-the-same-name-taking-different-arguments

16:16 <chetkhatri> Hello All, I am chetan from INDIA.

17:00 <dougmacdoug> do infix ops work on primitives or only the builtin primitives

17:00 <SeanTAllen> are you having a problem getting something to work dougmacdoug ?

17:03 <dougmacdoug> no but i had named some funs on primitive infix names and I was concerned it might be an issue

17:03 <dougmacdoug> probably just confusing myself by browsing the builtin package