workspace
stringclasses 4
values | channel
stringclasses 4
values | text
stringlengths 1
3.93k
| ts
stringlengths 26
26
| user
stringlengths 2
11
|
|---|---|---|---|---|
elmlang | general | The browser doesn't allow even JS to know the difference between a CORS problem and a network failure | 2019-03-25T18:38:48.718600 | Earlean |
elmlang | general | With CORS, it's up to the server to respond with headers that tell the browser it can allow the request. Inspect the headers of the response to see what the server's requirements are for the request. | 2019-03-25T18:43:57.718800 | Earlean |
elmlang | general | there are none in browser... | 2019-03-25T19:18:11.719100 | Floy |
elmlang | general | but there's a bunch of when I execute it via `httpie` | 2019-03-25T19:18:24.719300 | Floy |
elmlang | general | setting cookies etc. | 2019-03-25T19:18:50.719500 | Floy |
elmlang | general | an alternative is to talk with the API | 2019-03-25T19:20:26.719700 | Floy |
elmlang | general | nvm...I though that might be easy stuff with something obvious missing, but apparently it's not :confused: | 2019-03-25T19:22:39.720000 | Floy |
elmlang | general | anyway, thx for help | 2019-03-25T19:22:46.720200 | Floy |
elmlang | general | it's easy stuff as long as their server supports CORS | 2019-03-25T19:23:15.720400 | Earlean |
elmlang | general | if not, then it's impossible stuff. But it would be impossible with JS too. | 2019-03-25T19:23:37.720600 | Earlean |
elmlang | general | if you have JS that talks to wakatime then inspect the requests and response it gets | 2019-03-25T19:24:33.720800 | Earlean |
elmlang | general | You can also set up a cors proxy in a trusted server of yours, and query that. | 2019-03-25T20:30:13.722200 | Bebe |
elmlang | general | Don’t use third party cors servers with sensitive requests | 2019-03-25T20:30:45.723000 | Bebe |
elmlang | general | BTW <@Todd> there's a <#C7075S0GY|compile-time> that you might be interested in. | 2019-03-25T21:15:43.723200 | Carlota |
elmlang | general | Thanks a lot for the assistance, i wasn't online to continue, i m now going to try it | 2019-03-26T00:11:21.723600 | Leopoldo |
elmlang | general | Solved it. The problem was that i made a mistake creating the URL. I got a CORS error instead of a 404. A bit confusing, but there might be good reasons for doing it that way I guess. | 2019-03-26T02:49:25.724400 | Luba |
elmlang | general | why the `(Json.succeed SyncViewport)` ? | 2019-03-26T03:08:30.724600 | Leopoldo |
elmlang | general | hows does everyone manage not re-initing nested components on the OnChangeUrl? | 2019-03-26T03:30:15.725900 | Euna |
elmlang | general | I feel as though I am missing some key philosophy | 2019-03-26T03:30:31.726400 | Euna |
elmlang | general | like if you change url from /profile/something to -> /profile/anotherthing how do you maintain consistent state | 2019-03-26T03:40:14.727500 | Euna |
elmlang | general | I havent done routing in a while, but when that `OnChangeUrl` `Msg` comes in, you should be able to handle the `Route` like this:
```
Route.Profile id ->
case model.page of
Page.profile subModel ->
if subModel.id == id then
model -- Dont do anything
``` | 2019-03-26T03:45:15.729100 | Ashton |
elmlang | general | This is an extreme case where you literally dont want to do anything at all, and no state is changed, but in other cases you can do different things. | 2019-03-26T03:45:48.729700 | Ashton |
elmlang | general | wow, thank you so much for the reply, these are some very impressive numbers | 2019-03-26T04:53:05.730400 | Todd |
elmlang | general | ^.^ | 2019-03-26T05:42:11.730700 | Denae |
elmlang | general | Hi. I am working on Kite. Could you describe to me what you do with Kite? I kind of lack user data. :slightly_smiling_face: | 2019-03-26T05:52:01.730900 | Reuben |
elmlang | general | if you have a particular need (from Kite), I can increase its priority in the roadmap: <https://github.com/erkal/kite/projects/1> | 2019-03-26T05:54:20.731100 | Reuben |
elmlang | general | or maybe add it, if it is not already there | 2019-03-26T05:54:57.731300 | Reuben |
elmlang | general | Hi all, has anyone done some design on mapping Parent -> Children? | 2019-03-26T06:25:35.732000 | Zachary |
elmlang | general | you need a decoder for a message and since I don’t want to do anything with the event, the simplest way to express the fact that I want `SyncViewport` message to be generated on that event is to just `Json.succeed SyncViewport` | 2019-03-26T06:25:36.732100 | Maida |
elmlang | general | I'm currently using a `Dict Id Child.Model` to track model changes, is there something more abstract out there? | 2019-03-26T06:26:18.733100 | Zachary |
elmlang | general | I later thought about the proposed solution and realized that it is broken. | 2019-03-26T06:26:57.733500 | Maida |
elmlang | general | It works on ellie but i also noticed it only works on mousewheel | 2019-03-26T06:28:03.734700 | Leopoldo |
elmlang | general | Provisions have to be made for arrows and scrollbars dragging | 2019-03-26T06:28:28.735300 | Leopoldo |
elmlang | general | it really depends on your specific case. why `Dict Id Child.Model` doesn't fit your needs? | 2019-03-26T06:28:36.735600 | Lilli |
elmlang | general | I m yet to incorporate into my existing code and see | 2019-03-26T06:28:51.735700 | Leopoldo |
elmlang | general | it's quite some massaging to get it in | 2019-03-26T06:31:05.736100 | Zachary |
elmlang | general | but it does the job :slightly_smiling_face: | 2019-03-26T06:31:11.736300 | Zachary |
elmlang | general | questions that arise in my head after reading such a generic question:
1) mapping parent -> children. Recursive tuples? json? nested maps/dicts? List of (id, record model, nullable parent id)?
2) what is your model made of? is each element unique?
3) tracking changes... maybe create a distinct model that acts as a index? can your model be hashed?
4) are performances a deciding factor?
I don't want to appear polemic, but it feels like someone asked me how to carry "things" from a city to an other. If it's small enough, even a pigeon might do it :stuck_out_tongue: | 2019-03-26T06:42:07.745200 | Lilli |
elmlang | general | TCP/IP over pidgeon? | 2019-03-26T06:44:32.745500 | Niesha |
elmlang | general | <https://ellie-app.com/55HbSQ9mD6Na1> | 2019-03-26T06:45:57.745600 | Maida |
elmlang | general | this should work OK | 2019-03-26T06:46:02.745800 | Maida |
elmlang | general | I have moved the handling of the scroll in JS. | 2019-03-26T06:46:33.746000 | Maida |
elmlang | general | if there is no kind of electric power, it might be the only choice :smile: | 2019-03-26T06:48:49.746800 | Lilli |
elmlang | general | it's all about context | 2019-03-26T06:49:06.747000 | Lilli |
elmlang | general | I have also created a discourse topic about this:
<https://discourse.elm-lang.org/t/is-it-possible-to-handle-scroll-events-in-elm-without-the-use-of-ports/3396> | 2019-03-26T06:59:40.747200 | Maida |
elmlang | general | Regarding Parsers: How to chomp until different character than last one? | 2019-03-26T07:09:53.748500 | Miguelina |
elmlang | general | Question regarding generic record types and JSON decoding: can I? | 2019-03-26T07:32:33.749200 | Gertrude |
elmlang | general | This is not as easy as it sounds, this should work:
```
module Chomp exposing (chompWhileUnique)
import Parser exposing (..)
chompWhileUnique : Parser ()
chompWhileUnique =
loop Nothing chompWhileUniqueLoop
chompWhileUniqueLoop : Maybe Char -> Parser (Step (Maybe Char) ())
chompWhileUniqueLoop maybeChar =
oneOf
[ case maybeChar of
Just char ->
getChompedString (chompIf (\c -> char /= c))
|> andThen chompWhileUniqueNextChar
Nothing ->
getChompedString (chompIf (\c -> True))
|> andThen chompWhileUniqueNextChar
, succeed (Done ())
]
chompWhileUniqueNextChar : String -> Parser (Step (Maybe Char) ())
chompWhileUniqueNextChar str =
case String.uncons str of
Nothing ->
succeed (Done ())
Just ( c, _ ) ->
succeed (Loop (Just c))
```
For example in `elm repl`:
```
> run (getChompedString chompWhileUnique) ""
Ok "" : Result (List DeadEnd) String
> run (getChompedString chompWhileUnique) "abcefg"
Ok "abcefg" : Result (List DeadEnd) String
> run (getChompedString chompWhileUnique) "abccefg"
Ok "abc" : Result (List DeadEnd) String
``` | 2019-03-26T07:35:21.750700 | Velia |
elmlang | general | ie I have a type like this:
```
type Project task =
Project
{ id : Identifier
, name : String
, tasks : List task
, closed : Bool
}
```
Can I write a decoder in a generic way to handle `task` as one of several possible options? I've also tried instead creating a `Task` enum with all the types of task records that can be in there, but this seems to fail | 2019-03-26T07:38:34.752100 | Gertrude |
elmlang | general | <@Gertrude> it's possible to write fully generic JSON decoder, have a look at : <https://package.elm-lang.org/packages/Microsoft/elm-json-tree-view/latest/> | 2019-03-26T07:38:35.752200 | Nana |
elmlang | general | ah nvm | 2019-03-26T07:39:16.752400 | Nana |
elmlang | general | Thx... this will take a while to absorb... | 2019-03-26T07:39:42.753000 | Miguelina |
elmlang | general | <@Gertrude> Elm cannot generically derive decoders, no. However, you can ask it to try a few different decoders. | 2019-03-26T07:39:50.753400 | Niesha |
elmlang | general | you could write `projectDecoder : Decoder a -> Decoder (Project a)`
just like `Decode.list`etc. work | 2019-03-26T07:40:37.754300 | Nana |
elmlang | general | ~~Hmm. I'm wondering if the issue is how to pass the right constructor to Decode.succeed. I guess enum record types don't create the same kind of constructor as a type alias~~ (wrong avenue) | 2019-03-26T07:42:01.755100 | Gertrude |
elmlang | general | Ms doing elm dev?? | 2019-03-26T07:42:41.755500 | Danika |
elmlang | general | Well the use of `loop` makes it look more complicated than it is, but this avoids stack issues. It would look simpler by using a recursive parser. | 2019-03-26T07:42:55.755600 | Velia |
elmlang | general | <@Danika> yup it's <@Carl> at MS | 2019-03-26T07:43:44.756300 | Nana |
elmlang | general | you don't have to use a constructor if it doesn't fit. you can write your own function there that takes the parameters and composes a value of the right type | 2019-03-26T07:44:36.757300 | Sharon |
elmlang | general | So I have my `Project` decoder defined like this:
```
projectDecoder : Decoder (Project task)
projectDecoder =
Decode.succeed mkProject
|> required "id" int
|> required "name" string
|> required "tasks" (Decode.oneOf [list reportableTaskDecoder, list markableTaskDecoder])
|> required "closed" bool
```
And it complains about that `oneOf`, because the result is a mismatched list | 2019-03-26T07:45:37.758300 | Gertrude |
elmlang | general | No, no... loop is totally fine! I just used it for building a list. It's just that I'm very new (2h) to parsers. | 2019-03-26T07:47:15.758600 | Miguelina |
elmlang | general | Ok, anyway here is the recursive version without tail-call elimination:
```
chompWhileUnique : Parser ()
chompWhileUnique =
chompWhileUniqueHelp Nothing
chompWhileUniqueHelp : Maybe Char -> Parser ()
chompWhileUniqueHelp maybeChar =
oneOf
[ case maybeChar of
Just char ->
getChompedString (chompIf (\c -> char /= c))
|> andThen chompWhileUniqueNextChar
Nothing ->
getChompedString (chompIf (\c -> True))
|> andThen chompWhileUniqueNextChar
, succeed ()
]
chompWhileUniqueNextChar : String -> Parser ()
chompWhileUniqueNextChar str =
case String.uncons str of
Nothing ->
succeed ()
Just ( c, _ ) ->
chompWhileUniqueHelp (Just c)
``` | 2019-03-26T07:49:30.758800 | Velia |
elmlang | general | `reportableTaskDecoder` and `markableTaskDecoder` both have their own task-type instead of the generic `task` ? | 2019-03-26T07:51:16.760500 | Sharon |
elmlang | general | Yes. They're both record types, though similar, and the API is guaranteed to only ever have same type lists (ie. all projects have a list of only markable, or only reportable tasks) | 2019-03-26T07:52:39.761300 | Gertrude |
elmlang | general | Thx!! | 2019-03-26T07:53:27.762200 | Miguelina |
elmlang | general | I can express this potentially with an enum as well, but this gets cumbersome because I need to write constructor functions for each, which is a lot of code for only two record types possible, and also wrong, because a `type Task = Reportable | Markable` or similar enum would allow mixed type lists of tasks | 2019-03-26T07:54:13.763100 | Gertrude |
elmlang | general | Well your function is trying to implement both `Decoder (Project ReportableTask)` and `Decoder (Project MarkableTask)` at once, but you can't do that because they are different types.
I think the easiest is to just make 2 different functions. Then maybe see if you can extract common logic afterwards | 2019-03-26T07:56:12.763500 | Sharon |
elmlang | general | <@Gertrude>
```
projectDecoder : Decoder task -> Decoder (Project task)
projectDecoder taskDecoder =
Decode.succeed mkProject
|> required "id" int
|> required "name" string
|> required "tasks" (list taskDecoder)
|> required "closed" bool
``` | 2019-03-26T07:56:19.763600 | Nana |
elmlang | general | Yea, like that. But the problem of the mixed types comes from the `oneOf` (which is extracted out here), that will need to be separated anyway | 2019-03-26T07:59:33.765800 | Sharon |
elmlang | general | Yeah. Extracting that out to a different function still has the same problem. I think maybe this is not expressable how I think | 2019-03-26T08:03:41.767400 | Gertrude |
elmlang | general | It's not lists' fault, it's that they types make no sense. a non-list decoder would have to look like `or : Parser a -> Parser b -> Parser ?` which makes no sense. | 2019-03-26T08:05:13.769000 | Gertrude |
elmlang | general | Ah! I get it now. I missed the point of <@Nana>'s code. We make taskDecoder an argument, and pass the right decoder when I need it, which thankfully works for this. | 2019-03-26T08:09:32.770200 | Gertrude |
elmlang | general | Good will read. I m trying to avoid the use of ports though | 2019-03-26T08:31:39.771100 | Leopoldo |
elmlang | general | That was the point of the discourse post, to find a solution without ports. In any case, if ports is the way, then use ports. | 2019-03-26T08:55:16.771400 | Maida |
elmlang | general | based on my experience, extreme avoidance of ports is not a good thing. Minimize the use of ports but if the ports solve your problem, use them. | 2019-03-26T08:56:36.771600 | Maida |
elmlang | general | Ok | 2019-03-26T09:03:04.771800 | Leopoldo |
elmlang | general | can anyone recommend me a way to cast a 1 len String to Char? | 2019-03-26T09:54:40.772500 | Daysi |
elmlang | general | <@Daysi> `String.toList x |> List.head` | 2019-03-26T09:57:52.773100 | Nga |
elmlang | general | `String.uncons >> Maybe.map Tuple.first` gives you a `Maybe Char` representing the first unicode codepoint | 2019-03-26T09:57:54.773300 | Huong |
elmlang | general | would give you a Maybe char | 2019-03-26T09:57:56.773500 | Nga |
elmlang | general | `toList >> head` would also work, inded | 2019-03-26T09:58:10.773700 | Huong |
elmlang | general | there's also uncons | 2019-03-26T09:58:25.774500 | Nga |
elmlang | general | what are uncons? | 2019-03-26T09:58:37.774900 | Daysi |
elmlang | general | It's a function `String -> Maybe (Char, String)` that takes the first character off | 2019-03-26T09:59:02.775600 | Huong |
elmlang | general | (if the string isn't empty) | 2019-03-26T09:59:16.775900 | Huong |
elmlang | general | Ok | 2019-03-26T09:59:38.776300 | Daysi |
elmlang | general | You could also write a `NonEmptyString` newtype | 2019-03-26T09:59:47.776600 | Niesha |
elmlang | general | Which would be `(Char, String)` | 2019-03-26T09:59:55.776900 | Niesha |
elmlang | general | Anyone know how I can refactor this into a single line?
```let
( newSession, newCmd ) =
Session.redirectWithFlash session
in
( { model | divShown = True }, newSession, newCmd )
``` | 2019-03-26T10:00:57.777600 | Nga |
elmlang | general | <https://klaftertief.github.io/elm-search/?q=c%20-%3E%20(a%2C%20b)%20-%3E%20(c%2C%20a%2C%20b)> | 2019-03-26T10:01:58.778000 | Niesha |
elmlang | general | Oh yeah :stuck_out_tongue: That'll do it. Thanks. | 2019-03-26T10:03:29.778300 | Nga |
elmlang | general | `Session.redirectWithFlash session |> \(newSession, newCmd) -> ({model | divShown = True}, newSession, newCmd)` | 2019-03-26T10:03:53.778400 | Lewis |
elmlang | general | Thanks, that also works :slightly_smiling_face: | 2019-03-26T10:04:25.778600 | Nga |
elmlang | general | Elm is supposed to return all headers from a HTTP request right? For some reason the only header I can see is `content-type`, however I can see in my browser console that it has lots of headers like content-length, etc. I'm outputting the headers like so: `Debug.log (String.concat <| Dict.keys metadata.headers) x`, and the only output I get is: `content-type: <function>`. I know it's not a problem with the debug line, because I can do `String.concat ["one","two"]` in my debug line and I'll see "onetwo" in the console, so it seems Elm just isn't getting all headers in its headers dict? Unless I'm missing something weird :confused:. This is for a `Http.BadStatus_`, inside the metadata it has. | 2019-03-26T11:00:36.782500 | Nga |
elmlang | general | Are you doing a CORS request? | 2019-03-26T11:10:53.783000 | Jin |
elmlang | general | Yes I am. | 2019-03-26T11:11:02.783200 | Nga |
elmlang | general | You then have to expose additional header explicitely, see <https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Access-Control-Expose-Headers> | 2019-03-26T11:12:12.784100 | Jin |
elmlang | general | Thanks. Man, I've been bitten by CORS so many times! | 2019-03-26T11:12:37.784500 | Nga |
elmlang | general | So why can the browser see those headers but Elm can't? Hmm. | 2019-03-26T11:14:09.785100 | Nga |
elmlang | general | Elm will only allow you to access headers exposed with Expose-Headers, but the browser doesn't care? | 2019-03-26T11:15:56.786100 | Nga |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.