workspace
stringclasses 4
values | channel
stringclasses 4
values | text
stringlengths 1
3.93k
| ts
stringlengths 26
26
| user
stringlengths 2
11
|
---|---|---|---|---|
elmlang | general | :wink: | 2019-02-06T04:38:46.972300 | Antonette |
elmlang | general | Looks like this is taken care of already :wink: | 2019-02-06T04:39:08.972400 | Antonette |
elmlang | general | cannot find ping example,.. | 2019-02-06T04:40:07.972800 | Liza |
elmlang | general | but my examples will be not best - im game dev - and `pings` for me is also source of data (how long singnals from server is delayed, and how much all stuff must be predicted..) | 2019-02-06T04:46:04.974100 | Liza |
elmlang | general | but not all application running 60fps | 2019-02-06T04:46:30.974500 | Liza |
elmlang | general | I'm trying to _proxy_ a union type and it's constructors from one module to another so that consumers of my module don't have to import the 3rd party module (it can stay an indirect dependency), but I'm getting a ```
-- BAD EXPORT ------ /[...]/src/Website.elm
The (..) syntax is for exposing union type constructors. It cannot be used with
a type alias like `TextVariant` though.
6| , TextVariant(..)
^^^^^^^^^^^^^^^
Remove the (..) and you should be fine!
``` The message seems perfectly clear, but is there really no way to do this? | 2019-02-06T11:40:51.979100 | Dorsey |
elmlang | general | The definition of the `TextVariant` is as follows: ```
type alias TextVariant =
Mark.Style
``` | 2019-02-06T11:42:02.979700 | Dorsey |
elmlang | general | so `type alias TextVariant = SomeUnionType`? | 2019-02-06T11:42:07.980000 | Danika |
elmlang | general | Nope, no can do | 2019-02-06T11:42:18.980200 | Huong |
elmlang | general | Ha, beat me to it | 2019-02-06T11:42:18.980300 | Danika |
elmlang | general | Ok, thanks :neutral_face: | 2019-02-06T11:48:02.981000 | Dorsey |
elmlang | general | According to <@Huong> it's not possible: <https://elmlang.slack.com/archives/C0CJ3SBBM/p1549471338980200> . I wonder if it's a technical limitation or is there some fundamental reason for that? | 2019-02-06T11:52:09.981200 | Dorsey |
elmlang | general | It's just not something the langauge supports - you can expose an alias, but an alias is just that - another name for an existing thing. Re-exposing imports is supported in some languages, but Elm has chosen not to add that. I don't know if there's a fundamental reason.
Although, perhaps, there is one case where this could get mightily confusing:
let's say I have package X, where I define `type Foo = Foo | Bar`
I now create package Y, where I "reexpose" that `type Foo`.
Now, I publish a new major version of package X, with `type Foo = Foo | Bar | Baz`.
For the sake of argument, let's say I have `"package-x": "1.0.0 <= v < 3.0.0"` in package Y: Now the values exposed by package Y depend on the version of package X you happen to have installed. I'm not sure how that should even show up on the package website :smile: | 2019-02-06T12:04:12.981500 | Huong |
elmlang | general | so, you can only expose an alias, which says nothing about the implementation details of what happens in some other module or even package | 2019-02-06T12:04:36.981700 | Huong |
elmlang | general | Thanks <@Huong>. Insightful as always :smile: | 2019-02-06T12:05:45.981900 | Dorsey |
elmlang | general | My `Url.Parser.fragment` decoder is working, however with a url like `/#foo%20bar` it’s parsing as `foo%20bar`. I’m expecting `foo bar`. How can I decode the uri component? [Http.uriDecode](<https://package.elm-lang.org/packages/evancz/elm-http/latest/Http#uriDecode>) is gone in the new elm/http | 2019-02-06T12:54:48.984000 | Myung |
elmlang | general | I could make a `port` just for the native `decodeURIComponent` but I feel like I’m just doing something wrong | 2019-02-06T12:57:24.984700 | Myung |
elmlang | general | <@Myung> You are looking for `Url.percentDecode` | 2019-02-06T12:58:54.986300 | Bert |
elmlang | general | When I use `Url.Builder.custom` to make the url, `custom Relative [] [] (Just "foo bar")` it automatically encodes the component `/#foo%20bar` | 2019-02-06T12:59:41.987900 | Myung |
elmlang | general | <@Bert> ooo checking it out, thanks | 2019-02-06T13:00:00.988800 | Myung |
elmlang | general | interesting, it says to use Url.Parser instead in the docs | 2019-02-06T13:00:38.990000 | Myung |
elmlang | general | i think Url.Parser.fragment should decode by default… this is maybe an oversight? | 2019-02-06T13:01:03.991000 | Myung |
elmlang | general | the query parameter decoders automatically do it | 2019-02-06T13:01:10.991300 | Myung |
elmlang | general | I noticed the same thing with my app a little while ago, and Evan is looking for evidence at this point. So if you know a popular framework that automatically does URI decoding for fragments, feel free to respond in this thread. | 2019-02-06T13:01:27.991800 | Bert |
elmlang | general | i don’t use other frameworks | 2019-02-06T13:02:01.992100 | Myung |
elmlang | general | i think the behavior should be consistent though | 2019-02-06T13:02:10.992400 | Myung |
elmlang | general | 1) Url.Builder automatically encodes the fragment, so 2) I would expect Url.Parser to automatically decode as well | 2019-02-06T13:02:48.993600 | Myung |
elmlang | general | Yeah. I actually don't know if path sections are decoded or not. | 2019-02-06T13:03:01.994500 | Bert |
elmlang | general | and 3) if the query builder/parser already has this behaviour, i think the fragment builder/parser should have the same behavior | 2019-02-06T13:03:12.994800 | Myung |
elmlang | general | so those are my official responses :smile: | 2019-02-06T13:03:24.995200 | Myung |
elmlang | general | <@Bert> i lied, `Url.Builder` does not automatically encode | 2019-02-06T13:20:55.996700 | Myung |
elmlang | general | `Url.Builder.custom Url.Builder.Relative [] [] (Just "foo bar")` yields `"#foo bar"` | 2019-02-06T13:21:27.997200 | Myung |
elmlang | general | `Url.Builder.custom Url.Builder.Relative [] [] (Just (Url.percentEncode "foo bar"))` yields the intended `"#foo%20bar"` | 2019-02-06T13:21:53.997600 | Myung |
elmlang | general | so this is okay, albeit a pain in the ass | 2019-02-06T13:22:11.998100 | Myung |
elmlang | general | can’t imagine why the _url builder_ doesn’t do this for us | 2019-02-06T13:23:07.999000 | Myung |
elmlang | general | but my builder/parser is working now . thank you | 2019-02-06T13:23:27.999500 | Myung |
elmlang | general | Is it possible to create a `Url.Parser` that consumes arbitrarily deep paths and return them as a `String`? Like `/a` or `/a/b` or `/a/b/c`. I thought `Url.Parser.custom` will do that, but apparently not. | 2019-02-06T15:29:30.002100 | Dorsey |
elmlang | general | <@Dorsey> I imagine you’d do that with something like `</> string` to match `/a/b/c/*` and then do additional parsing on the total match of `a/b/c*` | 2019-02-06T15:37:24.003500 | Myung |
elmlang | general | if you want to match a variadic path, you’d have to capture all of the paths and then handle case for 0, 1, or more | 2019-02-06T15:38:24.004800 | Myung |
elmlang | general | :thinking_face: but wouldn't `</> string` capture only `a`? | 2019-02-06T15:38:34.005000 | Dorsey |
elmlang | general | <@Dorsey> ah sure enough, it says it only matches “one segment” | 2019-02-06T15:39:43.005500 | Myung |
elmlang | general | never tried to do something like this | 2019-02-06T15:41:36.006600 | Myung |
elmlang | general | Yes. I thought to do some magic with `List.foldl`, but I would need `andThen` to carry result of one parser to another. | 2019-02-06T15:42:03.007800 | Dorsey |
elmlang | general | i’ll think about it over lunch :green_salad: | 2019-02-06T15:42:06.007900 | Myung |
elmlang | general | yeah i was thinking about a recursive or folding parser too.. | 2019-02-06T15:42:26.008500 | Myung |
elmlang | general | Enjoy your lunch. I just ate supper :stuck_out_tongue: | 2019-02-06T15:42:38.009000 | Dorsey |
elmlang | general | i’m thinking the solution will surprise us | 2019-02-06T15:42:45.009200 | Myung |
elmlang | general | it _should_ be simple | 2019-02-06T15:43:10.009800 | Myung |
elmlang | general | Let's hope it will be a happy surprise. | 2019-02-06T15:43:19.010100 | Dorsey |
elmlang | general | haha yeah | 2019-02-06T15:43:24.010400 | Myung |
elmlang | general | The only example where multiple segments are captured is for `map`. Maybe that's the key. | 2019-02-06T15:45:10.011000 | Dorsey |
elmlang | general | hmm, what's the point of parsing an url and then returning it as a string? why not just use the original string then? | 2019-02-06T15:51:44.011900 | Nana |
elmlang | general | To match against arbitrary paths, e.g. pointing to files in a filesystem. There are certain rules about what paths are valid, but the depth of the path is variadic. | 2019-02-06T16:00:08.013700 | Dorsey |
elmlang | general | Also, I want to have the part "below" a certain path. Like: `/file/**/*.txt` to give me `File "whatever/path/was/matched.txt"`. | 2019-02-06T16:01:49.015300 | Dorsey |
elmlang | general | :wave: Where did you learn the term variadic? | 2019-02-06T16:35:14.015500 | Agustin |
elmlang | general | From <@Myung> <https://elmlang.slack.com/archives/C0CJ3SBBM/p1549485504004800> | 2019-02-06T16:48:12.015700 | Dorsey |
elmlang | general | :smile: | 2019-02-06T16:48:16.016000 | Dorsey |
elmlang | general | :joy: | 2019-02-06T16:49:00.016200 | Agustin |
elmlang | general | <@Myung> were did you learn that term? | 2019-02-06T16:49:10.016400 | Agustin |
elmlang | general | I think we used it correctly, didn't we? | 2019-02-06T16:49:34.016600 | Dorsey |
elmlang | general | I ask as I’ve only seen it once before, in a Ruby PR, and the author was like “its obvious, everyone knows this” and myself and another Ruby dev both with over 10 years of experience had never heard of it before | 2019-02-06T16:50:02.016800 | Agustin |
elmlang | general | So I’m keen to learn were it’s been introduced to others, feel like I’m missing out | 2019-02-06T16:50:24.017000 | Agustin |
elmlang | general | Always be learning etc | 2019-02-06T16:50:52.017200 | Agustin |
elmlang | general | I think I heard / saw the word before, but it probably wouldn't come to my mind without a prompt from Ohko. I probably couldn't explain it's meaning without some lookup. | 2019-02-06T16:51:34.017400 | Dorsey |
elmlang | general | Once you are familiar with some term, it's difficult to asses how well known it is in general. | 2019-02-06T16:52:42.017600 | Dorsey |
elmlang | general | <@Dorsey> i thought about it a little more, but i’m sad to say i don’t have a great recommendation | 2019-02-06T18:02:45.018500 | Myung |
elmlang | general | if i had your problem i would use either a query param or the fragment | 2019-02-06T18:03:12.019000 | Myung |
elmlang | general | but only because i don’t know of a “better” way to do it | 2019-02-06T18:03:27.019300 | Myung |
elmlang | general | `/download?p=/foo/bar/file` or `/download#/foo/bar/file` | 2019-02-06T18:04:20.020600 | Myung |
elmlang | general | Yeah, that won't work for me. I'm thinking about generating a long list of parsers using `map`. | 2019-02-06T18:05:12.021600 | Dorsey |
elmlang | general | I'm not there yet, but I think I'm getting close. | 2019-02-06T18:05:33.022400 | Dorsey |
elmlang | general | yeah you could have parsers that match 1 segment, 2, 3, up to N segments | 2019-02-06T18:05:34.022500 | Myung |
elmlang | general | and then error out after that | 2019-02-06T18:05:37.022700 | Myung |
elmlang | general | Now that I say this out loud, it sounds like the Elm way to do things | 2019-02-06T18:06:11.023400 | Myung |
elmlang | general | kinda like you see with map map3 map4 etc | 2019-02-06T18:06:24.023800 | Myung |
elmlang | general | Mhm. | 2019-02-06T18:06:35.024100 | Dorsey |
elmlang | general | <@Dorsey> thanks for sharing your issue. would you mind pinging me if you come up with something else? | 2019-02-06T18:07:14.024800 | Myung |
elmlang | general | Sure. | 2019-02-06T18:07:47.025000 | Dorsey |
elmlang | general | <@Myung> are you affiliated with the elm/url somehow? | 2019-02-06T18:11:04.025400 | Dorsey |
elmlang | general | Is there a way to figure out why `elm install elm/http` wants to install 1.0.0 if 2.0.0 is the latest? | 2019-02-06T18:29:11.026000 | Kimiko |
elmlang | general | check your elm.json? | 2019-02-06T18:29:56.026200 | Ruthann |
elmlang | general | What do I check for, exactly? `elm/http` isn't mentioned explicitly in it | 2019-02-06T18:33:18.026800 | Kimiko |
elmlang | general | <@Kimiko> it's probably because of other packages depending on `1.0.0`. Try removing and installing everything again. | 2019-02-06T18:34:28.027800 | Dorsey |
elmlang | general | E.g. you may have older version of `elm/core` and it depends on `http 1.0.0`. | 2019-02-06T18:35:11.028400 | Dorsey |
elmlang | general | Ah, yes, I guess it was `elm/core` pinning the dependency. Thanks. And I guess from your reply there's no automatic way to find which package is responsible, I just have to bisect my dependencies or something? | 2019-02-06T18:36:41.029400 | Kimiko |
elmlang | general | I don't think there is. Dependency management needs some more love, I guess :wink: | 2019-02-06T18:39:43.030200 | Dorsey |
elmlang | general | Thanks! | 2019-02-06T18:39:56.030400 | Kimiko |
elmlang | general | There’s a third-party project that can help with it. I’ll see if I can find it! | 2019-02-06T18:42:17.030500 | Lorilee |
elmlang | general | <https://www.markuslaire.com/github/elm-dependencies-analyzer/> | 2019-02-06T18:45:11.030700 | Lorilee |
elmlang | general | This is a pretty helpful project. I just used it to find updates etc | 2019-02-06T18:45:25.030900 | Lorilee |
elmlang | general | Very cool, thanks! | 2019-02-06T18:48:31.031100 | Kimiko |
elmlang | general | Hey, I almost got it. Actually I got it if I hardcode the type of the parser, but I want it to be parametrized. Here is the whole code (not working): ```
module Routes exposing
( Route(..)
, parse
, parser
)
import Url exposing (Url)
import Url.Parser as Parser exposing (..)
type Route
= Home
| Content String
| NotFound
parser : Parser (Route -> Route) Route
parser =
let
content =
Parser.custom "Content"
(\path ->
case String.split "." path of
[ base, "html" ] ->
Just base
_ ->
Nothing
)
variadic last =
10
|> List.range 1
|> List.foldl
(\_ parser_ ->
Parser.oneOf
[ parser_
, Parser.map
(\parent path ->
parent ++ "/" ++ path
)
(Parser.string </> parser_)
]
)
last
in
Parser.oneOf
[ Parser.map Home <http://Parser.top|Parser.top>
, Parser.map Content (variadic content)
]
parse : Url -> Route
parse url =
url
|> Parser.parse parser
|> Maybe.withDefault NotFound
``` | 2019-02-06T19:19:27.031700 | Dorsey |
elmlang | general | And here is a working example (with hardcoded type): ```
module Routes exposing
( Route(..)
, parse
, parser
)
import Url exposing (Url)
import Url.Parser as Parser exposing (..)
type Route
= Home
| Content String
| NotFound
parser : Parser (Route -> Route) Route
parser =
let
content =
Parser.custom "Content"
(\path ->
case String.split "." path of
[ base, "html" ] ->
Just (Content base)
_ ->
Nothing
)
variadic last =
10
|> List.range 1
|> List.foldl
(\_ parser_ ->
Parser.oneOf
[ parser_
, Parser.map
(\parent child ->
case child of
Content path ->
Content (parent ++ "/" ++ path)
_ ->
NotFound
)
(Parser.string </> parser_)
]
)
last
in
Parser.oneOf
[ Parser.map Home <http://Parser.top|Parser.top>
, variadic content
]
parse : Url -> Route
parse url =
url
|> Parser.parse parser
|> Maybe.withDefault NotFound
``` | 2019-02-06T19:21:31.031900 | Dorsey |
elmlang | general | Thanks for sharing <@Lorilee> | 2019-02-06T19:23:36.032100 | Dorsey |
elmlang | general | The signature of `Parser.map` is really bizarre. I'm trying to wrap my head around it. | 2019-02-06T19:24:51.032400 | Dorsey |
elmlang | general | ```
-- CORRUPT BINARY - /Users/…/Lookup.elmo
The binary data at
/Users/…/elm-stuff/0.19.0/Lookup.elmo is corrupt.
```
Couldn’t find an issue on elm/compiler about it | 2019-02-06T19:45:57.033100 | Ruthann |
elmlang | general | nuking elm-stuff works, wonder if this is a known/tracked issue | 2019-02-06T19:46:29.033600 | Ruthann |
elmlang | general | i take that back, mentioned here: <https://github.com/elm/compiler/issues/1853#issuecomment-440618037> | 2019-02-06T19:47:06.033800 | Ruthann |
elmlang | general | I have a form with (so far) 5 text input fields. Initially I had one `Msg` for each field, e.g. `EmailUpdated String`. However, I realized I could get away with a single message that packs a closure, `UpdateInput (String -> Model -> Model) String`. The closure then gets defined right at the relevant input, instead of over in the update function. Are there reasons this might be a bad idea? | 2019-02-06T22:29:12.036100 | Dede |
elmlang | general | Snippet:
```type alias Model =
{ username : String
, full_name : String
, display_name : String
, email : String
, password : String
}
type Msg
= SignupPressed
| UpdateInput (String -> Model -> Model) String
view model =
div []
[ input
[ placeholder "Full name"
, value model.full_name
, onInput
UpdateInput
(\full_name model -> { model | full_name = full_name })
]
, ...
]
update model msg =
case msg of
UpdateInput updater value ->
( model |> updater value, Cmd.none )
SignupPressed ->
( model, Cmd.none )
...``` | 2019-02-06T22:29:27.036200 | Dede |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.