workspace
stringclasses 4
values | channel
stringclasses 4
values | text
stringlengths 1
3.93k
| ts
stringlengths 26
26
| user
stringlengths 2
11
|
|---|---|---|---|---|
elmlang | general | Hi, has anybody somewhere worked with Processes? I mean, is it possible to apply them as workers for heavy elm functions? | 2019-01-14T03:20:23.298500 | Victorina |
elmlang | general | I’ll check these packages, thanks all for your helps. | 2019-01-14T03:22:10.298600 | Raymonde |
elmlang | general | I think those are still on the main thread (as opposed to a background worker) | 2019-01-14T03:25:33.302000 | Jake |
elmlang | general | Not directly. I guess it's planned for the future. Currently you can make separate Elm programs and wire them up using ports. You will have to do encoding and decoding. To start a headless Elm process use `Platform.worker` as described here: <https://package.elm-lang.org/packages/elm/core/latest/Platform#worker> | 2019-01-14T03:42:41.303000 | Dorsey |
elmlang | general | You will have to manage the threads from JS code. | 2019-01-14T03:43:24.303200 | Dorsey |
elmlang | general | Is anyone else having problems editing Elm code with the latest version of linter in Atom (2.3.0)? Since upgrading today, the linter no longer signals Elm compiler messages. If I downgrade Atom's linter to 2.2.0 (`apm install [email protected]`) things work again. Looking at the release notes of linter ("Remove support for legacy linter APIs"), I wonder if this is a problem with language-elm relying on deprecated APIs | 2019-01-14T12:11:04.307100 | Tynisha |
elmlang | general | Yes, same here. I didn't notice the linter being updated though, but that would probably explain it :+1: | 2019-01-14T12:12:17.307200 | Elyse |
elmlang | general | Suppose I had a function which throws away data, such that I can do
`tossA : Foo A -> Foo a`
What is the term for converting a type to a type variable (A -> a) ? Is there one?
If I want to discuss or or remind a reader that A -> a, how do I state that clearly?
“Note that the type parameter whent from … to a type variable?” | 2019-01-14T12:38:04.309100 | Leoma |
elmlang | general | shot in the dark, but maybe you'd call that "unbinding" | 2019-01-14T12:42:10.309700 | Elina |
elmlang | general | Most of the time, such a function forces you to return an "empty state" version of `Foo` (e.g. `List Int -> List a` must return `[]`) | 2019-01-14T13:45:05.311500 | Carman |
elmlang | general | I wonder if describing that behavior might be helpful when trying to explain the type change | 2019-01-14T13:46:43.312100 | Carman |
elmlang | general | You could destructure the record in a function in such a way that the fields become required, but that's not quite the same.
doThing : MyRecord -> Result
doThing { field1, field2 } =
{{function code here}} | 2019-01-14T13:52:45.312300 | Hyacinth |
elmlang | general | Eh, but that's not the same as requiring every field in the record. | 2019-01-14T13:53:20.312500 | Hyacinth |
elmlang | general | Could you use a custom type for the use case that you're trying to do? | 2019-01-14T13:54:08.312700 | Hyacinth |
elmlang | general | Currently in pieces across a number of my tabs. | 2019-01-14T17:49:59.313600 | Grisel |
elmlang | general | I’ll spin up a private Gist and DM you. | 2019-01-14T17:50:21.314400 | Grisel |
elmlang | general | Does elm note allow apostrophies at the end of function names anymore -> `name'` ? | 2019-01-14T18:15:23.315200 | Danika |
elmlang | general | <@Danika> yep, it was removed in 0.18 | 2019-01-14T18:15:50.315500 | Earlean |
elmlang | general | Ah okay | 2019-01-14T18:17:52.317100 | Danika |
elmlang | general | `'` is an uncommonly allowed in variable names in most languages that people come to Elm from, it's common use lead people to ask whether there was something special about it which put people off. | 2019-01-14T18:18:18.317200 | Earlean |
elmlang | general | Removing it also encourages people to think of better names | 2019-01-14T18:18:39.317700 | Earlean |
elmlang | general | or replace it with `_` :angel: | 2019-01-14T18:19:42.318000 | Danika |
elmlang | general | well perhaps you could help me with a naming problem then ^^ | 2019-01-14T18:19:52.318400 | Danika |
elmlang | general | yep, elm-upgrade did that automatically for the 0.18 upgrade | 2019-01-14T18:20:08.318800 | Earlean |
elmlang | general | but `_` is very commonly allowed in variable names across languages so people don't have the same reaction to it | 2019-01-14T18:20:47.319800 | Earlean |
elmlang | general | I have a type `Param = Note Int | Frequency Float | ...` for midi note and note frequency and some other things. I'd like to have the very standard `mtof` (midi to frequency) and `ftom` functions as utils for a package I'm working on.
But it also seems reasonable to have the same functions for generic Int/Float types. Do you think `mtof_` would be fine for the generic function? | 2019-01-14T18:23:47.322100 | Danika |
elmlang | general | Personally I don't like having raw numbers in my programs. I tend to wrap everything in a custom type that represents the unit | 2019-01-14T19:05:33.324800 | Carman |
elmlang | general | So I'd probably have a custom type that looks something like
```
type Frequency = Hertz Float
-- OR
type Hertz = Hertz Float
``` | 2019-01-14T19:07:11.325700 | Carman |
elmlang | general | I usually put these in their own modules, often a with a function to unwrap the raw value although that's not needed as much as you might expect | 2019-01-14T19:08:04.326400 | Carman |
elmlang | general | I don't see much utility in `Frequency (Hertz Float)` compared to `Frequency Float`, the Float is already wrapped why wrap it again? | 2019-01-14T19:10:42.327400 | Danika |
elmlang | general | That first one isn't double wrapping. It just has a different name for the type and the constructor, allowing you to add other units in the same category | 2019-01-14T19:12:21.328200 | Carman |
elmlang | general | e.g.
```
type Frequency = Hertz Float | MegaHertz Float | ...
``` | 2019-01-14T19:12:40.328900 | Carman |
elmlang | general | I usually don't do that though | 2019-01-14T19:12:52.329300 | Carman |
elmlang | general | I stick with `type Hertz = Hertz Float` | 2019-01-14T19:13:09.329800 | Carman |
elmlang | general | I understand that, but I already have `Frequency Float` which is part of
```type Params
= Value Float
| Note Int
| Frequency Float
| ... ```
So what you propose would be
```type Hertz = Hertz Float
type Params
= Value Float
| Frequency (Hertz Float)
| ... ```
no? | 2019-01-14T19:15:20.331900 | Danika |
elmlang | general | In my particular case I think `Frequency Float` already gives enough information that this is in hz (given the context of the rest of the package) and any more specificity doesn't really bring any extra benefit. | 2019-01-14T19:16:52.333100 | Danika |
elmlang | general | What does the `Params` type represent? | 2019-01-14T19:19:29.334300 | Carman |
elmlang | general | are these URL parameters? Inputs from a form? | 2019-01-14T19:19:52.334700 | Carman |
elmlang | general | I think we're probably going way beyond the scope of what I originally asked but..
```type Node
= Node { id : NodeID, type_ : NodeType, params : Dict String Param }```
`Param` represents some parameter of an audio node on a graph. | 2019-01-14T19:21:13.335600 | Danika |
elmlang | general | Do you do any math on param values? Or just output them directly? | 2019-01-14T19:24:18.336400 | Carman |
elmlang | general | There will inevitably be math on param values yes | 2019-01-14T19:25:53.336700 | Danika |
elmlang | general | I assume some operations will only work on frequencies, not on any param type right? | 2019-01-14T19:27:26.337600 | Carman |
elmlang | general | indeed | 2019-01-14T19:27:42.337800 | Danika |
elmlang | general | For those functions, I really like to be able to have signatures like `freqOp : Hertz -> Hertz` rather than `freqOp : Float -> Float` | 2019-01-14T19:29:45.339400 | Carman |
elmlang | general | That is a valid point | 2019-01-14T19:31:29.339700 | Danika |
elmlang | general | Depending on how the program is structured, maybe there's a conversion step somewhere between a `Params` and `Hertz`. Alternatively maybe `Params` wraps a `Hertz` like you showed earlier. | 2019-01-14T19:35:16.340800 | Carman |
elmlang | general | Definitely something to think about. In any case its nearly 1 so I should sleep! Thanks for the input :slightly_smiling_face: | 2019-01-14T19:40:42.341400 | Danika |
elmlang | general | ls | 2019-01-14T19:42:35.341500 | Ruthann |
elmlang | general | bloody slack always takes window focus when it loads :stuck_out_tongue: <@Earlean> several times! | 2019-01-14T21:51:33.343600 | Ruthann |
elmlang | general | It's a classic meme from the IRC days | 2019-01-14T21:54:14.343800 | Earlean |
elmlang | general | I'm trying to figure out elm/parse but I think I'm finding the types a little confusing. I've written up a (hopefully) clear gist with my problem at <https://gist.github.com/bstro/2436e7d3214e4f22544f872a56ad0512> … hoping someone with some experience with `elm/parse` might be able to take a look? It seems like a relatively simple problem to solve overall, I'm just unfamiliar with types in general, let alone parser combinators. | 2019-01-14T22:22:15.345600 | Santina |
elmlang | general | `<http://Parser.int|Parser.int>` will fail on dot. Instead of consuming digits one by one until it faces something else it just fails if complete string cannot be converted to integer | 2019-01-15T01:34:47.345900 | Lynne |
elmlang | general | Strategy here is to use `chompWhile` and then combine chomped characters into a number | 2019-01-15T01:35:22.346100 | Lynne |
elmlang | general | So basically your complete parser has to be built around your `dotDigit` | 2019-01-15T01:36:55.346300 | Lynne |
elmlang | general | Regarding the compilation errors. The first one says you are trying to return parsers which are capable of decoding into different types (from `Parser Int` to `Index` and from `Parser (Int, List Int)` to `Index`). All elements of a list should be of the same type | 2019-01-15T02:04:17.346500 | Lynne |
elmlang | general | Second one says that you are providing wrong value at right hand side of `|.`: it is expecting `Parser ignore` but you are giving it `Parser.andThen <function>` | 2019-01-15T02:05:34.346700 | Lynne |
elmlang | general | You should use `|>` when piping into `andThen` | 2019-01-15T02:05:48.346900 | Lynne |
elmlang | general | <https://ellie-app.com/4stpQV6wb8xa1> | 2019-01-15T02:16:32.347100 | Lynne |
elmlang | general | hmm, `String.toTitleCase` doesn't capitalize åäö :disappointed: | 2019-01-15T03:09:32.347900 | Nana |
elmlang | general | I guess a PR would be in order? | 2019-01-15T03:09:58.348300 | Nana |
elmlang | general | `String.toUpper` works correctly though | 2019-01-15T03:10:58.348600 | Nana |
elmlang | general | <@Nana> As a temp workaround does CSS `text-transform: capitalize;` work for you? | 2019-01-15T03:16:12.349400 | Hoa |
elmlang | general | aha, here's the source:
```
toTitleCase : String -> String
toTitleCase ws =
let
uppercaseMatch =
Regex.replace (regexFromString "\\w+") (.match >> toSentenceCase)
in
ws
|> Regex.replace
(regexFromString "^([a-z])|\\s+([a-z])")
(.match >> uppercaseMatch)
```
`[a-z]` only matches precisely a - z | 2019-01-15T03:25:30.351000 | Nana |
elmlang | general | hmm weird, even replacing it with `(.)` won't capitalize åäö! so there must be another problem as well | 2019-01-15T03:54:58.352000 | Nana |
elmlang | general | \w does not match those, so that needs to change as well. | 2019-01-15T03:55:43.352400 | Timika |
elmlang | general | aah right | 2019-01-15T03:56:00.352600 | Nana |
elmlang | general | ah, finally got it working. didn't quite understand how the original worked so I made this instead:
```
toTitleCase : String -> String
toTitleCase =
Regex.replace (Regex.fromString "^(.)| (.)" |> Maybe.withDefault Regex.never) (.match >> String.toUpper)
```
seems good enough? | 2019-01-15T04:13:13.353600 | Nana |
elmlang | general | Maybe `\s` instead of just a space? Alternatively, `\S+` to match everything that is not a white space, and transform those chunks, similar to the way it was before? | 2019-01-15T04:15:10.354100 | Timika |
elmlang | general | aha! here's another version then, which more closely follows the behavior of the original, but works for extended alphabet:
```
toTitleCase : String -> String
toTitleCase ws =
let
uppercaseMatch =
Regex.replace (Regex.fromString "\\S+" |> Maybe.withDefault Regex.never) (.match >> String.toSentenceCase)
in
ws
|> Regex.replace
(Regex.fromString "^([a-zà-ÿ])|\\s+([a-zà-ÿ])" |> Maybe.withDefault Regex.never)
(.match >> uppercaseMatch)
``` | 2019-01-15T04:22:07.355500 | Nana |
elmlang | general | (I found `[a-zà-ÿ]` was suggested to match all characters when googling) | 2019-01-15T04:22:58.356200 | Nana |
elmlang | general | hmm `[a-zà-ÿ]` doesn't work for Cyrillic though | 2019-01-15T04:29:05.356800 | Nana |
elmlang | general | it's quite a quest to find the correct Regex selector :stuck_out_tongue: | 2019-01-15T04:29:47.357800 | Nana |
elmlang | general | This is super tricky stuff, even though it does not look like it on the surface. I wonder if locales play a role here too. Different languages might capitalize differently, even though it’s the same character. | 2019-01-15T04:29:55.358000 | Timika |
elmlang | general | Would not it be easier to split string by space and then manually capitalize each word with `String.toUpper`? | 2019-01-15T04:35:54.359500 | Lynne |
elmlang | general | Related: <https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/String/toLocaleUpperCase> | 2019-01-15T04:36:13.359900 | Timika |
elmlang | general | <@Lynne> that's basically what my simpler version does | 2019-01-15T04:36:21.360200 | Nana |
elmlang | general | But you want to harness regular expressions? :slightly_smiling_face: | 2019-01-15T04:36:45.360500 | Lynne |
elmlang | general | > Also notice that conversion is not necessarily a 1:1 character mapping, as some characters might result in two (or even more) characters when transformed to upper-case. Therefore the length of the result string can differ from the input length. | 2019-01-15T04:37:07.361000 | Timika |
elmlang | general | I guess I could have done it with String.split yeah, it was just that the original used Regex :stuck_out_tongue: | 2019-01-15T04:37:17.361300 | Nana |
elmlang | general | I can see the benefit of it... if only it worked | 2019-01-15T04:37:39.361600 | Lynne |
elmlang | general | <@Lynne> this works pretty well:
```
toTitleCase : String -> String
toTitleCase =
Regex.replace (Regex.fromString "^(.)|\\s(.)" |> Maybe.withDefault Regex.never) (.match >> String.toUpper)
```
It uppercases any character which is after line-start or a space, and should work for all alphabets (depending on how `String.toUpper` works) | 2019-01-15T04:39:58.363400 | Nana |
elmlang | general | I would go mad though if I saw such code in project :smile: | 2019-01-15T04:42:37.364800 | Lynne |
elmlang | general | <@Lynne> yeah regex is weird, how's this then? :slightly_smiling_face:
```
toTitleCase : String -> String
toTitleCase phrase =
phrase |> String.split " " |> List.map String.toSentenceCase |> String.join " "
```
good enough for PR? | 2019-01-15T04:48:06.366000 | Nana |
elmlang | general | I am not maintainer of `String.Extra`, can't say if it is good or not :slightly_smiling_face: | 2019-01-15T04:48:39.366600 | Lynne |
elmlang | general | or maybe the Regex version is faster | 2019-01-15T04:48:42.366700 | Nana |
elmlang | general | or Regex might be slow, no idea | 2019-01-15T04:48:56.366900 | Nana |
elmlang | general | Well, unless you run on 1k+ lists every other second it should not be a big deal | 2019-01-15T04:49:25.367400 | Lynne |
elmlang | general | Hello folks | 2019-01-15T05:34:10.367800 | Sadie |
elmlang | general | Once again I reach my limitation with Json decoders | 2019-01-15T05:34:25.368100 | Sadie |
elmlang | general | I would like to handle the case where a JSON property can be a nullable or an object or missing | 2019-01-15T05:34:46.368600 | Sadie |
elmlang | general | I started with something like this: ` |> optional "test" (nullable (Decode.map Just decodeTestInfo)) Nothing` | 2019-01-15T05:34:58.369000 | Sadie |
elmlang | general | (elm-json-decode-pipeline) | 2019-01-15T05:35:06.369300 | Sadie |
elmlang | general | Maybe `optional "test" (nullable decodeTestInfo) Nothing` is sufficient in that case? | 2019-01-15T05:36:04.369700 | Sadie |
elmlang | general | I think even `optional "test" (map Just decodeTestInfo) Nothing` should be sufficient | 2019-01-15T05:38:53.370200 | Lynne |
elmlang | general | `optional` already handles null and missing fields | 2019-01-15T05:39:06.370600 | Lynne |
elmlang | general | Unless you want null and missing cases to be represented differently in the model | 2019-01-15T05:42:50.371200 | Lynne |
elmlang | general | Thanks <@Lynne> i am going to try that | 2019-01-15T06:19:56.372300 | Sadie |
elmlang | general | It works | 2019-01-15T06:37:55.372500 | Sadie |
elmlang | general | Damn. Thank you <@Lynne>, that was incredibly helpful and above and beyond what I expected. I really appreciate your help. | 2019-01-15T07:34:16.372800 | Santina |
elmlang | general | You are welcome. Writing parsers is not an easy thing. | 2019-01-15T07:47:22.373000 | Lynne |
Subsets and Splits