workspace
stringclasses 4
values | channel
stringclasses 4
values | text
stringlengths 1
3.93k
| ts
stringlengths 26
26
| user
stringlengths 2
11
|
|---|---|---|---|---|
elmlang | general | However, using `[]` to represent "waiting for data" is kind of bad form. <http://blog.jenkster.com/2016/06/how-elm-slays-a-ui-antipattern.html> | 2019-01-25T03:28:45.955500 | Nana |
elmlang | general | not excited about the asset handling | 2019-01-25T03:43:28.955800 | Desire |
elmlang | general | <@Renda> I can report that elm-ui has been really nice so far! definitely the way to go for styling in Elm | 2019-01-25T06:25:56.956200 | Nana |
elmlang | general | especially compared to the alternatives (I was pretty happy with component-scoped Stylus in Vue, but I found elm-css quite awkward in comparison, and elm-css-modules seems more like a workaround for when you already have tons of stylesheets that you don't want to rewrite) | 2019-01-25T06:29:24.956400 | Nana |
elmlang | general | encountered one limitation so far: the tables currently don't support adding styles or event handlers to rows, only cells
I was able to work around it though | 2019-01-25T06:30:52.956600 | Nana |
elmlang | general | and the row/column/padding/spacing/align system is just perfect! | 2019-01-25T06:32:09.956800 | Nana |
elmlang | general | <@Nana> hey that's great to hear! I can't wait to try it! But do you feel Elm-UI is better than SFC + stylus in Vue? (Since you said Elm-css was worse) | 2019-01-25T07:22:55.957200 | Renda |
elmlang | general | <@Renda> yeah probably | 2019-01-25T07:26:24.957400 | Nana |
elmlang | general | and definitely if you don't have years of experience with css | 2019-01-25T07:27:01.957600 | Nana |
elmlang | general | I have not looked into it but how doesnt it actually work? Is the styling done via JavaScript and absolute positionings or is the output clever css? | 2019-01-25T07:27:14.957800 | Renda |
elmlang | general | I'm definetly worst at css :stuck_out_tongue: | 2019-01-25T07:27:46.958000 | Renda |
elmlang | general | it mostly uses css flexboxes, but with a nicer API | 2019-01-25T07:29:04.958200 | Nana |
elmlang | general | Ok cool :slightly_smiling_face: | 2019-01-25T07:30:48.958400 | Renda |
elmlang | general | Hello! Is there a way to make it so Elm does not overwrite the value of inputs when `update` is called? | 2019-01-25T07:45:17.959200 | Juliann |
elmlang | general | <@Juliann> quite possibly, depending on what you mean by that. Is the problem that when you change your model, you have in your view inputs whose `value` attribute is set to the model's data? | 2019-01-25T07:46:46.961000 | Antonette |
elmlang | general | The value is unchanging | 2019-01-25T07:46:55.961400 | Juliann |
elmlang | general | Yet it is reset each time update is called | 2019-01-25T07:47:16.962100 | Juliann |
elmlang | general | I'm suddenly realising that maybe `Html.Keyed` will help here as without it Elm can't avoid re-rendering my entire view, let me quickly try that | 2019-01-25T07:48:16.964100 | Juliann |
elmlang | general | First thing that comes to mind is you could just not set the `value` attribute in your inputs. They won't change when the model changes. | 2019-01-25T07:48:19.964200 | Antonette |
elmlang | general | Why don't you want your inputs to reflect your model's data though? | 2019-01-25T07:49:09.964900 | Antonette |
elmlang | general | I need to set the initial value, can that be done without value? | 2019-01-25T07:51:07.965500 | Juliann |
elmlang | general | And because the value isn't used afterwards by Elm, it's a normal HTML form | 2019-01-25T07:51:27.966000 | Juliann |
elmlang | general | Ok. | 2019-01-25T07:52:07.966200 | Antonette |
elmlang | general | <@Juliann> all inputs in Elm are controlled inputs, the `view` is a reflection of the `model` | 2019-01-25T07:53:42.968500 | Earlean |
elmlang | general | You might have misunderstood something about the way Elm works ; the view is not just generated when you start the application. It's re-generated every time your model changes. | 2019-01-25T07:53:57.968800 | Antonette |
elmlang | general | Yes, but it does incremental rebuilding | 2019-01-25T07:54:46.970200 | Juliann |
elmlang | general | I thought that it would be smart enough to not rewrite this if the diff showed no change | 2019-01-25T07:55:10.971400 | Juliann |
elmlang | general | Which, in turns, means that if you don't change your model when you put something in your input fields, and the view sets the value from the model's data, then every time something in your model changes, your view will be updated back to the model's initial value. | 2019-01-25T07:55:13.971800 | Antonette |
elmlang | general | I'll try forcing it with lazy | 2019-01-25T07:55:22.972100 | Juliann |
elmlang | general | What problem are you seeing that you're trying to avoid? | 2019-01-25T07:55:25.972200 | Earlean |
elmlang | general | Input for which state have not changed are being re-rendered, causing their updated value to be reset | 2019-01-25T07:57:13.972900 | Juliann |
elmlang | general | if they're unchanged, aren't they at their reset value? | 2019-01-25T07:58:19.973400 | Earlean |
elmlang | general | No, because inputs can be updated outside of Elm | 2019-01-25T07:59:00.973700 | Juliann |
elmlang | general | inputs being updated will produce a `input` event that you can listen for in Elm | 2019-01-25T08:00:00.974500 | Earlean |
elmlang | general | And then I need to manage the state of this entire form, which would double the size of this program | 2019-01-25T08:00:41.975200 | Juliann |
elmlang | general | Elm used to support this -> <https://github.com/elm/html/commit/4912804dadbb2e7cdb6effa8ee737d2e08a96247> | 2019-01-25T08:00:48.975600 | Juliann |
elmlang | general | Yes, if you have a form that you're rendering from Elm you should be storing it's state in the Elm model | 2019-01-25T08:01:09.975900 | Earlean |
elmlang | general | I would agree if Elm did anything with it | 2019-01-25T08:06:43.978200 | Juliann |
elmlang | general | Elm is responsible for perhaps 5% of the state here, the rest is plain old HTML | 2019-01-25T08:06:57.978700 | Juliann |
elmlang | general | It's reinventing the wheel | 2019-01-25T08:07:03.978900 | Juliann |
elmlang | general | And until 0.19 this was supported | 2019-01-25T08:07:17.979200 | Juliann |
elmlang | general | you can recreate the defaultValue property yourself | 2019-01-25T08:07:40.979500 | Virgie |
elmlang | general | on the other hand: if elm renders it, then elm does do something with it right? | 2019-01-25T08:09:20.980400 | Virgie |
elmlang | general | It's a javascript element property, not a html property, so you cannot | 2019-01-25T08:09:43.980800 | Juliann |
elmlang | general | I may be stuck here, sadly | 2019-01-25T08:10:27.981300 | Juliann |
elmlang | general | I'm confused as to why `lazy` doesn't work | 2019-01-25T08:10:40.981900 | Juliann |
elmlang | general | I'm not sure what you mean by that? you can implement the `defaultValue` as it existed in 0.18 yourself in 0.19 | 2019-01-25T08:10:56.982300 | Virgie |
elmlang | general | You cannot, it uses a Kernel module to set javascript properties on the virtual DOM. Only `elm/*` packages can do this | 2019-01-25T08:12:26.983400 | Juliann |
elmlang | general | e.g.
```
import VirtualDom
defaultValue : String -> Attribute msg
defaultValue v =
VirtualDom.property "defaultValue" (Encode.string v)
``` | 2019-01-25T08:13:00.983900 | Virgie |
elmlang | general | yes but the kernel functionality is exposed, with exactly this sort of use case in mind | 2019-01-25T08:13:48.984400 | Virgie |
elmlang | general | ooh, that's different from the way it is defined in core | 2019-01-25T08:14:13.984800 | Juliann |
elmlang | general | let's try that :slightly_smiling_face: | 2019-01-25T08:14:16.985100 | Juliann |
elmlang | general | it should be similar to the way it was defined in 0.18 | 2019-01-25T08:14:36.985400 | Virgie |
elmlang | general | In 0.18 it uses the Kernel module while other the other properties use the method you've shown | 2019-01-25T08:15:50.986100 | Juliann |
elmlang | general | It works, wonderful | 2019-01-25T08:16:22.986500 | Juliann |
elmlang | general | Thank you so much | 2019-01-25T08:16:24.986700 | Juliann |
elmlang | general | :heart: | 2019-01-25T08:16:44.987100 | Juliann |
elmlang | general | And now also this:
<https://package.elm-lang.org/packages/pzp1997/assoc-list/1.0.0/>
which works with any type, without a `k -> comparable` function. Performance is different though. | 2019-01-25T08:58:28.987400 | Velia |
elmlang | general | I'm trying to use elm-upgrade, but I get the following error message:
```
{ MaxRedirectsError: Redirected 10 times. Aborting.
at ClientRequest.fn.request.res (/home/fredrik/.npm-global/lib/node_modules/elm-upgrade/node_modules/got/index.js:40:23)
at Object.onceWrapper (events.js:315:30)
at emitOne (events.js:116:13)
at ClientRequest.emit (events.js:211:7)
at HTTPParser.parserOnIncomingClient (_http_client.js:551:21)
at HTTPParser.parserOnHeadersComplete (_http_common.js:115:23)
at Socket.socketOnData (_http_client.js:440:20)
at emitOne (events.js:116:13)
at Socket.emit (events.js:211:7)
at addChunk (_stream_readable.js:263:12)
message: 'Redirected 10 times. Aborting.',
host: '<http://package.elm-lang.org|package.elm-lang.org>',
hostname: '<http://package.elm-lang.org|package.elm-lang.org>',
method: 'GET',
path: '/search.json',
statusCode: 301,
statusMessage: 'Moved Permanently' }
ERROR: Unable to connect to <https://package.elm-lang.org>. Please try again later.
```
I am behind a corporate proxy, but `elm install` etc works fine. | 2019-01-25T10:31:02.988500 | Chaya |
elmlang | general | Try installing a package you never installed with `elm install` to verify it is not the proxy's issue. | 2019-01-25T10:32:20.989200 | Lynne |
elmlang | general | As of 0.19 `elm install` takes packages from local cache so it may not even need to go to Internet if package is cached already | 2019-01-25T10:33:08.990100 | Lynne |
elmlang | general | I can open `<https://package.elm-lang.org/search.json>` on my machine | 2019-01-25T10:34:19.990700 | Lynne |
elmlang | general | Getting 200, so it must be an issue with your corporate proxy | 2019-01-25T10:35:08.991100 | Lynne |
elmlang | general | Oh. My. Lord! That is awesome! love how he’s like, it’s “asymptotically slower than elm/core” but works well on small-ish Dict’s.
I was just going to add how irritable I was working with any-dict but I hadn’t heard of assoc-list! Woo! | 2019-01-25T10:59:21.991300 | Buffy |
elmlang | general | Yes it's very useful, here is the announcement:
<https://discourse.elm-lang.org/t/introducing-assoc-list-a-dict-substitute-with-custom-keys/2901> | 2019-01-25T11:00:51.991500 | Velia |
elmlang | general | Noice! | 2019-01-25T11:04:17.991800 | Buffy |
elmlang | general | This might help if performance is acceptable for your use case:
<https://package.elm-lang.org/packages/pzp1997/assoc-list/1.0.0/> | 2019-01-25T11:05:53.992000 | Velia |
elmlang | general | Yeah it all makes sense, it's just that in this case my life would have been a lot easier if I could have just `deriving Eq, Ord` on my `alias` :expressionless: | 2019-01-25T12:16:10.992500 | Bernardo |
elmlang | general | Perhaps I'm just spoiled by Elm | 2019-01-25T12:16:17.992700 | Bernardo |
elmlang | general | (expecting things to be so easy) | 2019-01-25T12:16:27.992900 | Bernardo |
elmlang | general | Love it <@Nana>! | 2019-01-25T12:17:58.993200 | Bernardo |
elmlang | general | Amusingly I actually use that
```
type RemoteData e a
= NotAsked
| Loading
| Failure e
| Success a
```
pattern elsewhere, when submitting a form. For some reason it didn't click that I could follow it for loading also :tada: | 2019-01-25T12:18:58.993400 | Bernardo |
elmlang | general | I remember seeing a few (IIRC) <@Leonie> talks which feature a slide showing a loop of (IIRC) Build :arrow_right: Learn :arrow_right: Refactor :rewind:.
Does this sound familiar to anyone? | 2019-01-25T12:59:32.995400 | Bernardo |
elmlang | general | vaguely yes, though i think instead of Learn it was Discover | 2019-01-25T14:02:35.995900 | Lenore |
elmlang | general | <@Bernardo> Is it this one? <https://youtu.be/x1FU3e0sT1I?t=2348> | 2019-01-25T15:31:39.996900 | Mammie |
elmlang | general | That's it! :tada: Thanks <@Mammie>, what a hero! | 2019-01-25T15:32:27.997600 | Bernardo |
elmlang | general | Hi folks! I have a somewhat interesting question. I'm writing a parser (using `elm/parser`) for a fairly simple syntax I've written for commands, which is sort of like a command line if you will with the possibility of multiple statements. Some of those statements will ultimately mutate some of the global state (add items for example), and then future statements might refer to those items added. In an Elm-esque way, i'm trying to figure out how to have a sort of "new stuff created" stack as part of the parser, so that I can check the *real* state and also check the *will be created but not created yet* state to let the user know about potential errors prior to running their commands | 2019-01-25T19:38:23.000200 | Loyce |
elmlang | general | I'm not really sure how to go about that, somehow perhaps passing a simplistic model throughout all the parsers? | 2019-01-25T19:38:45.000600 | Loyce |
elmlang | general | I guess there are many ways to do the same thing but I would keep the parsers about just parsing and creating the AST and then I’d do the type checking or interpretation part, which would raise those errors you are talking about if everything parsed correctly | 2019-01-25T19:45:52.003400 | Bebe |
elmlang | general | So do a first pass | 2019-01-25T19:46:18.003800 | Loyce |
elmlang | general | Then run it through another piece that will take that and do a sort of "fake" execution | 2019-01-25T19:46:29.004000 | Loyce |
elmlang | general | and then assuming that checks out, do the real execution | 2019-01-25T19:46:50.004200 | Loyce |
elmlang | general | That makes sense I think; once I have my Statement objects etc and I run them through a checker, I would be able to keep whatever state necessary | 2019-01-25T19:47:12.004400 | Loyce |
elmlang | general | Thanks! | 2019-01-25T19:47:58.004600 | Loyce |
elmlang | general | Anyone have a better way of describing a graph (for web audio) where the nodes can multiple inputs or outputs than the following:
```
graph
[ node "apple" data
, node "banana" data
, node "watermelon" data
, node "pineapple" data
]
[ connect {output = "apple", input= "banana"}
, connect {output = "apple", input "watermelon"}
, connect {output = "banana", input "pineapple"}
, connect {output = "watermelon", input "apple"}
]
``` | 2019-01-25T20:31:13.010200 | Cammy |
elmlang | general | or a less verbose version of `connect`: `connect ("banana", "apple")`
I’m looking more for another paradigm or structure. | 2019-01-25T20:32:45.011400 | Cammy |
elmlang | general | Is this a directed graph? | 2019-01-25T21:47:48.011800 | Carman |
elmlang | general | I wonder if you could do something more concise like:
```
graph
[ node "apple" data [ "banana", "watermelon" ]
, node "banana" data [ "pineapple" ]
, node "watermelon" data [ "apple" ]
, node "pineapple" data []
]
``` | 2019-01-25T21:51:52.013200 | Carman |
elmlang | general | where the third argument is a list of node ids you connect to | 2019-01-25T21:52:16.013600 | Carman |
elmlang | general | Is there a way to encode a `Decode.value`? What I want to do is: `API request -> decode as Value, because I don't care about the structure, nor can I predict it -> Send this value through a port -> work with it in JS -> send the updated value back to elm` | 2019-01-26T04:11:27.017800 | Moshe |
elmlang | general | I think `Encode.Value` and `Decode.Value` are interchangeable | 2019-01-26T04:12:08.018400 | Lynne |
elmlang | general | Ok. Then I got at least that part right :smile: So how to use it in an encoder? I don’t know what to put at `???`: `Encode.object [ ("some_value", ??? obj.someValue ) ]` | 2019-01-26T04:13:49.019800 | Moshe |
elmlang | general | Does not it work with just putting this value into the tuple? In other words, if you have `someValue : Decode Value` can't you just use `Encode.object [("some_value", someValue)]`? | 2019-01-26T04:15:24.021200 | Lynne |
elmlang | general | `type alias Value = Json.Encode.Value` | 2019-01-26T04:16:41.021700 | Lynne |
elmlang | general | Well… beat me to it. Wait a moment ^^ | 2019-01-26T04:16:44.021900 | Moshe |
elmlang | general | From `Decode.elm` | 2019-01-26T04:16:45.022000 | Lynne |
elmlang | general | This is the same type | 2019-01-26T04:17:05.022300 | Lynne |
elmlang | general | Why is it idiomatic to make subscriptions `mySubscription : (value -> msg) -> Sub msg` instead of just returning a `Sub value` and expecting the caller to call `Sub.map` like we do for `Html` and `Cmd`? | 2019-01-26T08:33:02.024300 | Rico |
elmlang | general | <@Rico> I'm not sure what you mean, it's far more common for functions to take a function that will convert a value to a `msg` than to use `Html.map` or `Cmd.map` | 2019-01-26T08:37:27.026300 | Earlean |
elmlang | general | Look at all the functions in the Html package or all the functions that return `Cmd msg` in the official packages | 2019-01-26T08:38:26.028100 | Earlean |
Subsets and Splits