#elixir-lang - 06 April 2019
« Back 1 day Forward 1 day »
[07:07:53] freedrull: i'm getting this error when deploying my phoenix app to heroku https://gist.github.com/mcfiredrill/2e0c2127171f0dec82454dd8a51c317d
[07:08:21] freedrull: should this line be `supervisor(Chat.Presence)` ? https://github.com/datafruits/hotdog_lounge/blob/master/lib/chat.ex#L17
[18:25:17] starbelly: My hope was that elixir 2.0 would bring about a LOT of changes, that are ya know not backwards compat. Namely, doing away with a good bit.
[18:29:06] starbelly: nox: That's a huge list I think... but one thing would be doing away with some syntactic sugar I suppose. I would like things to be more simple. I would like to see somehow more of an impedance mismatch between getting going east/fast and being forced to learn what's going on. I would like less abstractions.
[18:29:44] starbelly: But that last one, about learning.... etc. that's a HUGE one imo... I don't know how you solve it, but yeah
[18:30:06] nox: Less abstractions is the best way to make the life of experienced programmers miserable.
[18:31:11] starbelly: whatyouhide: haha well yeah, these are so high level... they are decomp tickets 🤣
[18:33:12] nox: How are protocols more difficult to understand than any trait-like system in any programming languag?
[18:35:13] starbelly: In erlang, from where I stand anyway, you just can't get going so fast... you have to understand what's going on. Perhaps this has less to do with elixir itself and more to do with the ecosystem
[18:36:42] starbelly: I'd have to make a proper list of things I would like changed/removed... but eh... I think it would be things that would never be considered.
[18:38:17] starbelly: Keep in mind, I started in Elixir... then went to Erlang... now I bounce back and forth :)
[18:39:50] nox: tristan__: Imagine having a volunteer fixing the lexer and everything to teach the compiler to report column numbers,
[18:40:11] nox: and then be willing to implement compiler diagnostics with ranges and better error messages on top of that,
[18:51:41] starbelly: I know... but there has go to be a better way... as far as having those abstractions but also having to understand what's going on.
[18:52:25] starbelly: For every experienced engineer, there's 999 programmers making an absolute mess of the world.
[22:44:17] askadar: Newbie here. I am trying to transform [2, 2, 3, 3, 2, 3, 3, 3, 2, 2] into [, [2, 3, 3], [2, 3, 3, 3], , ]. Couldn't manage to do so. I tried chunk to no avail. Maybe chunk_while...Any hint?
[22:47:25] askadar: Enum.chunk_by(list, fn x -> x == 2 end) produces [[2, 2], [3, 3], , [3, 3, 3], [2, 2]]
[23:06:21] edw: OK. I'm going to be trusting and assume I'm not frustrating someone's pedagogic ends.
[23:07:07] askadar: I am trying to create a ToC through heading tags (h2 and h3) collected by Floki.find
[23:16:13] askadar: I cannot grasp the accumulator mechanism. It's just too complex to me. Love pattern matching tho. Thanks @ariedler btw
[23:16:42] ariedler: no problem; yeah the accumulator mechanism is something you get use to with time. hence why I said the fold/reduce is much simpler
[23:18:23] edw: The cognitive load of remembering all these fancy special case fns is often heavier than just pounding out a reduce once you get the hang of it.
[23:50:41] edw: ariedler: In much of Manhattan above 14th St, yes. I live on Charles St the West Village, Manhattan. To the south is 10th St, Christopher St, Barrow St. To the north is Perry St, then 11th. West: Hudson St, Greenwich St, Washington St. East: Bleecker St, West 4th St, 7th Avenue. Other boroughs offer differing levels of craziness. E.g. In Brooklyn, there's a N 2nd St, a S 2nd St, and a plain old 2nd St. None of them touch.