workspace
stringclasses 4
values | channel
stringclasses 4
values | text
stringlengths 1
3.93k
| ts
stringlengths 26
26
| user
stringlengths 2
11
|
---|---|---|---|---|
elmlang | general | With ports and base64-encoded strings, it can be done but it’s rather hacky | 2019-04-23T07:21:47.215700 | Huong |
elmlang | general | (got a bunch of dependencies that depend on `elm/http` v1) | 2019-04-23T07:21:52.215800 | Kimbery |
elmlang | general | I guess there's no way to install from github with 0.19? (I see there's a pending PR against elm-http-builder) | 2019-04-23T07:25:24.216300 | Kimbery |
elmlang | general | There is <https://github.com/Skinney/elm-git-install> but are you going to have two versions of `elm/http` in your app? | 2019-04-23T07:26:50.216700 | Lynne |
elmlang | general | I was going to fork http-builder and merge the PR in. | 2019-04-23T07:27:26.217300 | Kimbery |
elmlang | general | Ah, then that tool might help | 2019-04-23T07:27:42.217800 | Lynne |
elmlang | general | does elm-git-install work with 0.19? | 2019-04-23T07:27:46.217900 | Kimbery |
elmlang | general | Yes | 2019-04-23T07:27:50.218100 | Lynne |
elmlang | general | oh awesome - phew | 2019-04-23T07:27:57.218400 | Kimbery |
elmlang | general | that gets me out of a tricky spot | 2019-04-23T07:28:02.218600 | Kimbery |
elmlang | general | always used elm-github-install with 0.18, exactly for moments like this | 2019-04-23T07:28:44.219300 | Kimbery |
elmlang | general | hopefully the PR for elm-http-builder will be merged fairly soon | 2019-04-23T07:28:58.219800 | Kimbery |
elmlang | general | Actually, I don't know but someone mentioned here it was updated if I remember correctly | 2019-04-23T07:29:08.220100 | Lynne |
elmlang | general | looks like it from the readme yeah | 2019-04-23T07:29:19.220300 | Kimbery |
elmlang | general | `elm-git-install` was specifically made for 0.19. It just clones the dependency and adds it to your source-directories, so it’s not quite the same thing as installing a regular package (i.e. won’t work for packages with kernel code etc, which essentially means it’s a pretty safe thing to do!) | 2019-04-23T07:32:04.221700 | Huong |
elmlang | general | It used to exist in 0.18 as well | 2019-04-23T07:32:43.222100 | Lynne |
elmlang | general | that seems fine, shame it has to do a clone but I can live with it temporarily | 2019-04-23T07:32:59.222400 | Kimbery |
elmlang | general | different author, different tool :slightly_smiling_face: | 2019-04-23T07:33:24.222600 | Huong |
elmlang | general | Only the name stays :smile: | 2019-04-23T07:33:53.222800 | Lynne |
elmlang | general | close enough - elm-git-install <-> elm-github-install :smile: | 2019-04-23T07:34:24.223000 | Huong |
elmlang | general | Ah, I did not realize it was github-install | 2019-04-23T07:35:38.223300 | Lynne |
elmlang | general | so to recap, it’s a different tool with a different name by a different author, solving a related but different problem (elm-github-install specifically wanted to make it possible to install kernel code) :smile: | 2019-04-23T07:36:48.223500 | Huong |
elmlang | general | Yeah, duck test seems to pass | 2019-04-23T07:38:10.223700 | Lynne |
elmlang | general | is there a way to check if there are any pending Cmds still being processed? Ie. I do a Cmd.batch of HTTP requests, and I want to know if any of them are still loading? | 2019-04-23T07:58:02.225200 | Gertrude |
elmlang | general | There's no way of doing it via Cmds, you'd need to track their status yourself in the Model. | 2019-04-23T08:01:25.225900 | Kimbery |
elmlang | general | Depending on your needs you might consider <https://package.elm-lang.org/packages/krisajenkins/remotedata/latest/RemoteData> | 2019-04-23T08:02:06.226200 | Kimbery |
elmlang | general | <https://package.elm-lang.org/packages/elm/http/latest/Http#Progress> | 2019-04-23T09:11:19.226600 | Lea |
elmlang | general | hmm, can i somehow install a global event handler in elm? | 2019-04-23T09:13:12.227200 | Emilee |
elmlang | general | You mean like keypresses and suchlike? I don’t know of a generic way, although Browser.Events covers a lot of the most common stuff. | 2019-04-23T09:24:17.228000 | Dede |
elmlang | general | scrolling, in particular | 2019-04-23T09:27:03.229700 | Emilee |
elmlang | general | ISTR having to write a global handler in JS and wire it up with ports, although I can’t find the relevant code lying around right now. | 2019-04-23T09:34:39.230600 | Dede |
elmlang | general | Something like `_window.addEventListener('scroll', scrolledOrResized);` which I just snagged from <http://guupa.com/elm-scroll-resize-events/> | 2019-04-23T09:35:03.231000 | Dede |
elmlang | general | i see :slightly_smiling_face: i thought there was some change that made it possible to register global (body-level) handlers from within elm, but that might have just been a fever dream :wink: | 2019-04-23T09:35:36.231700 | Emilee |
elmlang | general | i may just resort to grabbing animationframe, since scrolling doesn't actually catch every case I might recalculate my logic anyway. a lot of message spam, but that's okay i think^^ | 2019-04-23T09:36:26.232800 | Emilee |
elmlang | general | i hope the application doesn't become too much of a cpu hog that way ^^' | 2019-04-23T09:36:45.233200 | Emilee |
elmlang | general | no you can't check for "pending Cmds", and `Http.Progress` isn't really for this either | 2019-04-23T10:17:39.233900 | Nana |
elmlang | general | you can however have some value in the model which you update whenever you make a request, and when you get a response | 2019-04-23T10:20:56.234200 | Nana |
elmlang | general | Is there a pass-thru decoder for `elm/bytes`? I'm downloading a file with progress using http and I'm using `expectBytes` which takes a message and a decoder. I want the decoder to just return the same bytes instead of decoding it into an elm value. (The bytes will be passed onto `File.Download.bytes`) | 2019-04-23T11:17:35.237600 | Caroline |
elmlang | general | <@Caroline> Probably need to whip something up with `expectBytesResponse`:
<https://package.elm-lang.org/packages/elm/http/latest/Http#expectBytesResponse> | 2019-04-23T11:20:43.238300 | Earnest |
elmlang | general | <@Earnest> thanks - found a possible answer here I think - <https://stackoverflow.com/questions/52054917/downloading-file-from-post-request-with-elm> | 2019-04-23T11:22:19.239300 | Caroline |
elmlang | general | Any Miami Elm developers here? | 2019-04-23T11:54:26.239900 | Frieda |
elmlang | general | Orlando here, kinda close! | 2019-04-23T12:00:09.240400 | Alicia |
elmlang | general | Couldn't find anything on client-side hydration with elm. Can elm be configured to recycle/hydrate existing HTML before rendering the page? | 2019-04-23T12:15:12.241400 | Maxwell |
elmlang | general | Say I have an app that renders this HTML on the page `<div><div>10</div><button>Increment</button></div>`. | 2019-04-23T12:17:42.243300 | Maxwell |
elmlang | general | If I serve my index.html with some pre-existing HTML that matches the HTML I wrote above, will elm reuse the nodes ("hydrate") or ignore it (duplicating the HTML by default)? | 2019-04-23T12:18:59.245500 | Maxwell |
elmlang | general | In Elm, how much work triggered by an msg is synchronous? Is the call to `update` synchronous, the following render, calls make to Ports, etc… | 2019-04-23T12:21:45.247900 | Leoma |
elmlang | general | It depends. Certain DOM events will call `update` synchronously and redraw the DOM synchronously afterwards (to prevent the jumping cursor type bugs).
Other events, like an HTTP response coming in, will asynchronously trigger `update`, and `update` will then asynchronously schedule a render.
Ports are (afaik) always triggered synchronously. | 2019-04-23T12:24:40.248100 | Huong |
elmlang | general | Note that a lot of this changed between 0.18 and 0.19, but most of those changes shouldn’t have obvious impact on users, except a few things have now transparently become possible (such as trigger things through a port JS in response to a click event, within the same tick) | 2019-04-23T12:25:55.248300 | Huong |
elmlang | general | Thanks. I wanted to call `.focus` on mobile safari. I’d have to do it over a port. Problem is that mobile Safari devs thought it would be nicer to only have `.focus` behave if called synchronousely from a user event. | 2019-04-23T12:25:58.248500 | Leoma |
elmlang | general | So, I think this is impossible with Elm. | 2019-04-23T12:26:19.248700 | Leoma |
elmlang | general | You should be able to do that in 0.19 | 2019-04-23T12:26:23.248900 | Huong |
elmlang | general | if it is in response to a user event, that is | 2019-04-23T12:26:39.249100 | Huong |
elmlang | general | Oh. So the update via user event will be sync. And the render, AND the port calls? | 2019-04-23T12:27:05.249300 | Leoma |
elmlang | general | pretty much! | 2019-04-23T12:27:23.249500 | Huong |
elmlang | general | Oh great news! I’ll give that a go. THANK YOU for the info. | 2019-04-23T12:27:45.249700 | Leoma |
elmlang | general | <@Maxwell> it’s not really supported right now, you can do it if you are using `Browser.element` because of a quirk of the implementation (it doesn’t clear out the node), but if you are using `Browser.application` there is no way | 2019-04-23T12:35:03.251400 | Alicia |
elmlang | general | So, I can't find it now, perhaps I was hallucinating. ISTR there was an issue where batched Cmds could run asynchronously where the individual ones would run synchronously. | 2019-04-23T12:35:27.251600 | Dede |
elmlang | general | relevant bits are <https://github.com/elm/browser/blob/master/src/Elm/Kernel/Browser.js#L45-L53> and <https://github.com/elm/browser/blob/master/src/Elm/Kernel/Browser.js#L83> | 2019-04-23T12:35:40.251900 | Alicia |
elmlang | general | Perhaps it's closed, or I'm misremembering the details. | 2019-04-23T12:35:41.252000 | Dede |
elmlang | general | Thanks <@Cleotilde>. I made this ellie to show what I mean: <https://ellie-app.com/5mg8w4VdnBLa1> | 2019-04-23T12:36:07.252500 | Maxwell |
elmlang | general | I'll check out Browser.element. | 2019-04-23T12:36:14.252700 | Maxwell |
elmlang | general | I've always wondered, is Elm's kernel JavaScript code, e.g., (src/Elm/Kernel/Browser.js) written by hand or by a compiler? | 2019-04-23T12:42:24.254000 | Maxwell |
elmlang | general | hand | 2019-04-23T12:42:32.254200 | Huong |
elmlang | general | Oh, I see. | 2019-04-23T12:42:43.254600 | Maxwell |
elmlang | general | Hi.
What are the guaranties regarding call order of cmd ports subscriptions on the javascript side?
I have a case where 2 calls of my top-level update function writes to a port. The 1st call also returns a Random.generate. The unexpected behavior is that the data written to my port by the second update is passed to my javascript subscription before the data written by the first update.
I know the commands execution order is not at all guaranteed, but I always thought it applied only to the commands emitted by a single update.
Should I consider the behavior I observe as being normal? | 2019-04-23T12:47:45.259900 | Barney |
elmlang | general | <@Erlene> You could hash (and possibly base64 encode) the string. That would make for a valid and unique class name. | 2019-04-23T12:57:54.260000 | Kenya |
elmlang | general | I think it will work with `Browser.sandbox` as well, if you click `Compile` in ellie it will start with that `10` as the initial value in the HTML, and it won’t change until it re-renders | 2019-04-23T14:45:01.260700 | Alicia |
elmlang | general | if you set the initial value to 10 on the elm side it should be seamless | 2019-04-23T14:45:14.260900 | Alicia |
elmlang | general | Nope. | 2019-04-23T14:45:38.261100 | Maxwell |
elmlang | general | As you can see in the ellie I posted, the DOM nodes are created from scratch every time. | 2019-04-23T14:45:52.261300 | Maxwell |
elmlang | general | (That's what the animation/keyframes is for there in the CSS; to show a fade-in animation). | 2019-04-23T14:46:23.261500 | Maxwell |
elmlang | general | A better question might be: where does Elm stand wrt SSR? Is it even desirable? | 2019-04-23T14:49:11.261900 | Maxwell |
elmlang | general | oh I see what you mean, I guess it is not possible then :disappointed: | 2019-04-23T14:56:11.262200 | Alicia |
elmlang | general | I thought it would diff starting from the node you give it and only replace if needed | 2019-04-23T14:56:26.262400 | Alicia |
elmlang | general | Do you think this would be a desirable thing to have? | 2019-04-23T14:56:52.262600 | Maxwell |
elmlang | general | I think so, there’s been talk about it before, server side rendering was going to be part of 0.19 but was dropped due to time constraints | 2019-04-23T14:57:29.262800 | Alicia |
elmlang | general | How would SSR work in elm (in theory) since the compiler generates javacript that runs in the browser? | 2019-04-23T14:58:55.263000 | Maxwell |
elmlang | general | oh, got a lead, it looks like it works when you aren’t in debug mode | 2019-04-23T15:02:05.263300 | Alicia |
elmlang | general | None | 2019-04-23T15:02:55.263500 | Alicia |
elmlang | general | In your gif, when you start clicking the button, the app has already started. | 2019-04-23T15:05:46.263900 | Maxwell |
elmlang | general | There's a 1.5 second delay before calling `Elm.Main.init`. | 2019-04-23T15:06:23.264200 | Maxwell |
elmlang | general | That's on purpose to see if the nodes are created twice. | 2019-04-23T15:06:35.264400 | Maxwell |
elmlang | general | try it out, with `debug` on it replaces it twice like this | 2019-04-23T15:28:10.264700 | Alicia |
elmlang | general | but without debug on it doesn’t replace it, the events only get hooked up after the 1.5s delay, but the nodes are there | 2019-04-23T15:29:54.265100 | Alicia |
elmlang | general | I am trying to find out if I can start a good PWA project using Elm. Please help | 2019-04-23T16:03:23.266100 | Jesenia |
elmlang | general | Elm(TEA)+PWA=? | 2019-04-23T16:05:06.266400 | Jesenia |
elmlang | general | what is PWA? | 2019-04-23T16:28:13.266800 | Willodean |
elmlang | general | Progressive Web App? | 2019-04-23T16:29:18.267000 | Briana |
elmlang | general | Any Bulgarian Elm developers here? Please PM me. | 2019-04-23T17:03:53.267600 | Dayna |
elmlang | general | <@Jesenia> : Try Create-elm-app from NPM. Two simple CLI commands and your Elm PWA is running :) Its amazing both for development and production. | 2019-04-23T17:11:36.271400 | Monnie |
elmlang | general | <#C4HTCLHPV|server-side-renderer> i think you want to ask in here | 2019-04-23T17:50:23.271600 | Keneth |
elmlang | general | <@Dayna> Just a stab at it: if the derivation happens within view code then the reference being compared can be to the unchanging value from the model. | 2019-04-23T17:58:40.272000 | Milda |
elmlang | general | I'm trying to have a layout manager that I can reuse across different applications, that will manage each application component in a container that its responsible for. The way I'm trying to set it up right now is to have messages on the layout manager side that are specific to it (e.g. `type LayoutMsg = Resize | Move | Add | Remove | ... `), & the content that its containing will have application-level messages (e.g. `type AppMsg = NoOp | QueryBackendForSomething | ChangeViews | LayoutMessage Layout.LayoutMsg | ... `).
Right now I'm running into issues with 2 different message types (generic `msg` which is the app-level message and `LayoutMsg`) not being able to exist in one div's attributes (styling & event handlers for the container, using `LayoutMsg`s) & html (what the container is containing, which has `AppMsg`s that will get handled by the application's model update). I get the feeling that this isn't the right approach for this problem, so I'm wondering what the correct direction might be. | 2019-04-23T20:06:51.279600 | Fidela |
elmlang | general | <@Fidela> you need another type to unify your LayoutMsg with your AppMsg | 2019-04-23T20:11:46.280900 | Earlean |
elmlang | general | something in layout like, `type LayoutMsg msg = AppMessage msg | Resize | Move | Add | Remove | ... `? | 2019-04-23T20:12:44.282600 | Fidela |
elmlang | general | Yep | 2019-04-23T20:14:14.282900 | Earlean |
elmlang | general | in retrospect that makes sense; I kept trying to think how to join them on the application level; thank you :slightly_smiling_face: | 2019-04-23T20:14:48.283500 | Fidela |
elmlang | general | Thanks you two, that was exactly the kind of insight I was hoping for | 2019-04-23T23:51:12.285100 | Jillian |
elmlang | general | Hey everyone, is there any way on 0.19 to detect user leaving the current view? Meaning if user clicks the anchor we're handling loading new way as defined by stuff in `Browser.application` and that results in next requests. Would it be technically possible to found out currently loaded view and call functions producing commands for specific requests that are being tracked? Next thing is, I guess, to just combine cancel commands and new commands to init next view according to next URL | 2019-04-24T00:40:19.287200 | Floy |
elmlang | general | I’m not sure I completely follow what behavior you’re after. It seems like you’re familiar with `Browser.application`, and therefore I assume `onUrlRequest` and `onUrlChange`…? | 2019-04-24T00:44:24.288200 | Dede |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.