workspace
stringclasses 4
values | channel
stringclasses 4
values | text
stringlengths 1
3.93k
| ts
stringlengths 26
26
| user
stringlengths 2
11
|
|---|---|---|---|---|
elmlang | general | My advice is to detect the case, no `_ -> ` and write to an error holder on your model | 2019-04-19T21:24:06.077200 | Willodean |
elmlang | general | that way you can do something with the error, like send it to a port or show it to the user | 2019-04-19T21:24:37.077600 | Willodean |
elmlang | general | Or re-write the thing in something with HKT? Purescript? Does purescript have something like elm-ui? That's the lib that keeps me stuck on elm at the moment? | 2019-04-19T21:26:01.078900 | Buffy |
elmlang | general | HKT is not the issue | 2019-04-19T21:26:31.079300 | Willodean |
elmlang | general | the problem is you need to convince the compiler that unaligned cases are impossible | 2019-04-19T21:26:53.079800 | Willodean |
elmlang | general | that means that you need some for of dependent typing | 2019-04-19T21:27:04.080200 | Willodean |
elmlang | general | where the type of the function can be determined by the value of its input | 2019-04-19T21:27:15.080500 | Willodean |
elmlang | general | PureScript cannot do this either, unless they added existentials recently | 2019-04-19T21:27:49.081500 | Willodean |
elmlang | general | haskell can only barely do this | 2019-04-19T21:28:18.081900 | Willodean |
elmlang | general | and even then, its not ergonomic | 2019-04-19T21:28:27.082300 | Willodean |
elmlang | general | the truth is you are running down the type level road, and that road lies maddness | 2019-04-19T21:28:45.083000 | Willodean |
elmlang | general | you can invest hours upon hours of effort convincing the compiler this wont happen, and in practice have eliminated ZERO bugs | 2019-04-19T21:29:16.084000 | Willodean |
elmlang | general | Yeh, I get the picture. So for now, just `Debug.todo` then upgrade that to something to track if the app ever hits any of these cases... maybe by then something else will exist that handles this sort of thing nicer? | 2019-04-19T21:29:45.084400 | Buffy |
elmlang | general | In practice, I have this in my code
`_ -> singleton x` | 2019-04-19T21:30:08.084800 | Willodean |
elmlang | general | and never had a problem | 2019-04-19T21:30:13.085000 | Willodean |
elmlang | general | We don't need the compiler to statically check everything | 2019-04-19T21:30:25.085400 | Willodean |
elmlang | general | So you'd rather noop than fail? | 2019-04-19T21:30:33.085800 | Buffy |
elmlang | general | we need the compiler to staticly check the things we are worried about | 2019-04-19T21:30:34.085900 | Willodean |
elmlang | general | I have never needed to actually track stray messsages | 2019-04-19T21:30:50.086400 | Willodean |
elmlang | general | I did for a while | 2019-04-19T21:30:52.086600 | Willodean |
elmlang | general | but I deleted it because it was code to maintain that was not adding value | 2019-04-19T21:31:04.087000 | Willodean |
elmlang | general | How many real world errors are caught by `Maybe`, shitloads | 2019-04-19T21:31:43.087900 | Willodean |
elmlang | general | Gotta love that value stuff... what's that? kek. This is a project I'm workin' on in my free time so type level shenanigans are just questions I ask to learn. :slightly_smiling_face: | 2019-04-19T21:32:03.088600 | Buffy |
elmlang | general | how many real world errors are caught by dependently typed complexity to ensure the compiler wont allow a stray message. Like zero. | 2019-04-19T21:32:11.088800 | Willodean |
elmlang | general | Then you are in the wrong language | 2019-04-19T21:32:28.089300 | Willodean |
elmlang | general | go idris or agda | 2019-04-19T21:32:33.089600 | Willodean |
elmlang | general | idris has a javascript backend | 2019-04-19T21:32:41.089900 | Willodean |
elmlang | general | Idris has a javascript backend? | 2019-04-19T21:33:10.090500 | Buffy |
elmlang | general | yes | 2019-04-19T21:33:55.091400 | Willodean |
elmlang | general | go have some real fun :slightly_smiling_face: | 2019-04-19T21:34:07.091800 | Willodean |
elmlang | general | Elm has two use cases. Teaching beginners, and real world consumer facing applications where performance and space concerns are key. | 2019-04-19T21:34:48.093100 | Willodean |
elmlang | general | Learning about types is not one of them. As is writing good FP code. | 2019-04-19T21:35:05.093700 | Willodean |
elmlang | general | Just seems odd, this is the second time I've wanted to do dependently typed shenanigans... first time trying to map Http Headers to Persistent Queries for Pagination stuff... now randomly here trying to make update not have little weird edge case-y things...
I'm scared of what I'll find if I go down the Idris brick road? | 2019-04-19T21:35:17.093900 | Buffy |
elmlang | general | lol | 2019-04-19T21:35:17.094100 | Buffy |
elmlang | general | You will find madness | 2019-04-19T21:35:53.094300 | Willodean |
elmlang | general | You want to take a real hit off the Sandy type level programming? | 2019-04-19T21:36:12.094800 | Willodean |
elmlang | general | You are going to have diminishing returns in the extreme, as you invest more and more time in types that don't catch real world problems, and don't add features to your app. | 2019-04-19T21:36:43.095500 | Willodean |
elmlang | general | But its a beautiful madness, and an experience worth having | 2019-04-19T21:36:54.095800 | Willodean |
elmlang | general | been down that road and back again my friend | 2019-04-19T21:37:09.096100 | Willodean |
elmlang | general | you think you want dependent types, until you live with them in an enterprise production application | 2019-04-19T21:37:27.096500 | Willodean |
elmlang | general | I'll take a word from the wise and hold off then for now. :slightly_smiling_face:
Maybe one day there will exist a dependently typed language that solves all our problems with nice error messages :man-shrugging: | 2019-04-19T21:39:13.098000 | Buffy |
elmlang | general | Blodwen ftw | 2019-04-19T21:40:32.098200 | Willodean |
elmlang | general | :man-shrugging: | 2019-04-19T21:42:41.098400 | Buffy |
elmlang | general | Sincere thanks for you time either way :slightly_smiling_face: | 2019-04-19T21:46:22.098800 | Buffy |
elmlang | general | I am in the process of upgrading to elm-0.19 and I'm hitting this issue too ... works without `--debug`, but fails with that error when I use it in elm-live. My codebase is prohibitively large to extract a SSCCE :confused: | 2019-04-19T22:52:27.099000 | Tamika |
elmlang | general | Has any advice been published on how to get rid of the security messages (in git/github) that come with using a 0.18 dev environment. I feel bad providing my clients with repos that have security warnings, even if the warnings only apply to dev packages | 2019-04-20T03:49:28.100600 | Lynn |
elmlang | general | Hello, I hope somebody can help me with this weird behavior i've been getting: we have an application that runs 3 submodules which each have their own MVU cycle (own model, messages and update). We use this function to map these messages to the submodules <https://github.com/rtfeldman/elm-spa-example/blob/master/src/Main.elm#L280> . Now one of our submodules is doing a HttpRequest in a Command. We've put this command on the submodule's `init` function and can see the command being fired but the message never being received. However if we put the same command inside one of the submodules messages as a side effect it works fine. Why does this happen? | 2019-04-20T05:16:39.103800 | Olevia |
elmlang | general | Maybe you case on the `Msg SubMsg` and `Page` to be ‘active’ at the same time? | 2019-04-20T05:44:38.104100 | Lea |
elmlang | general | So stuff ends up <https://github.com/rtfeldman/elm-spa-example/blob/master/src/Main.elm#L276> | 2019-04-20T05:45:00.104300 | Lea |
elmlang | general | Hey, is there a concern with my API? I don’t allow Msg without a Model. The rule is, you must ALWAYS start with one of: a model, or an error. Now, operations on errors (adding a Cmd Msg) simply result in the error. As opposed to getting combined with a model. | 2019-04-20T13:34:08.108200 | Leoma |
elmlang | general | Rest in peace Joe Armstrong, you were a true functional programming brother, thanks for paving the way for the rest of us. | 2019-04-20T14:36:22.109300 | Ann |
elmlang | general | Hello,
I met Elm recently and ended up creating interest in learning it, because with Elm-UI I would save time not having to learn three different languages for front end.
It seems that there is no updated or complete content about the language. Maybe that's one of the reasons that contribute to the low popularity of some languages.
Can anyone help me? | 2019-04-20T17:53:14.111600 | Sandi |
elmlang | general | what are you looking for in particular? | 2019-04-20T17:59:58.111900 | Virgie |
elmlang | general | Learn how to use elm-ui. I need recommendations on where to look for content. | 2019-04-20T18:19:04.112100 | Sandi |
elmlang | general | there is a <#C4F9NBLR1|elm-ui> channel for questions, and an examples folder in the repo <https://github.com/mdgriffith/elm-ui/tree/master/examples> | 2019-04-20T18:26:57.112900 | Virgie |
elmlang | general | Thanks for your help. I'm going to read it now | 2019-04-20T18:31:05.113100 | Sandi |
elmlang | general | Also of interest is this talk from the author of elm-ui: <https://www.youtube.com/watch?v=Ie-gqwSHQr0> | 2019-04-20T18:36:25.113500 | Carman |
elmlang | general | I've been meaning to PR the tuple bits to the main library, but I think there's some better was to handle this stuff now so I haven't bothered upgrading the library. | 2019-04-21T02:27:32.114600 | Myrtle |
elmlang | general | The package docs are great too | 2019-04-21T03:59:53.115500 | Bebe |
elmlang | general | Any suggestions on how I can builder a decoder of `Int` that does not succeed in this case
```
> Decode.decodeString <http://Decode.int|Decode.int> "2.0"
Ok 2 : Result Decode.Error Int
``` | 2019-04-21T04:36:11.116200 | Lynn |
elmlang | general | JSON only has `number` and doesn't disambiguate between floats and integers, so as far as JSON (and JS) are concerned, that's _exactly_ the same as json with `2`. In other words, I can't really think of a way of handling this that doesn't involve manually parsing the actual json string | 2019-04-21T04:39:36.116500 | Huong |
elmlang | general | Or to explain it slightly differently - Elm internally uses `JSON.parse(jsonString)` because it's much more efficient to do that than to implement a custom parser. The issue in this case is that, on the resulting value, there is - as far as I can think of - no way to say whether the original JSON string specified `2.0` or just `2` | 2019-04-21T04:43:14.116700 | Huong |
elmlang | general | (i.e. in JS, `2.0 === 2`) | 2019-04-21T04:44:40.116900 | Huong |
elmlang | general | so to disambiguate, it would have to happen during parsing rather than during decoding, which we currently do not control | 2019-04-21T04:45:28.117100 | Huong |
elmlang | general | Decode it as a string, andThen call toInt maybe? | 2019-04-21T04:56:15.117800 | Danika |
elmlang | general | `Decode.string` only works on things that are actually a string _in the JSON_. That was my first thought, too, but it wouldn't work :smile: | 2019-04-21T05:00:34.118000 | Huong |
elmlang | general | it's somewhat easier to see if we adapt the original example to `Decode.decodeString (Decode.field "foo" <http://Decode.int|Decode.int>) "{\"foo\": 2.0}"` | 2019-04-21T05:02:07.118300 | Huong |
elmlang | general | Hmm | 2019-04-21T05:02:56.118800 | Danika |
elmlang | general | As soon as `JSON.parse` is used, you end up with this value `2.0` which is _exactly_ the same thing as `2` as far as javascript (and JSON!) goes | 2019-04-21T05:04:23.119000 | Huong |
elmlang | general | ok. hoped I wasn't missing anything. I may need to add a tag to the serialised values then | 2019-04-21T05:04:55.119200 | Lynn |
elmlang | general | Yeah, that's probably a good idea. Alternatively (but also icky) you could use `{"foo": "2.0"}` and interpret that string | 2019-04-21T05:06:14.119400 | Huong |
elmlang | general | indeed, I can see that causing issues as I will use this decoder very generally | 2019-04-21T05:07:24.119600 | Lynn |
elmlang | general | Does Elm use deep compare after an update to prevent rerenders? I've noticed it compares deeply when using (==), but I'm not sure if the VDom is written in Elm w/ kernel code or if it's entirely JS and uses something like shallow compare | 2019-04-21T13:36:31.122500 | Dayna |
elmlang | general | I’m pretty sure it doesn’t do a deep compare. | 2019-04-21T13:39:16.123100 | Cornell |
elmlang | general | I know `lazy` compares by reference, except for primitives like `String` and `Int`, which it will compare directly. | 2019-04-21T13:40:08.123800 | Cornell |
elmlang | general | Though maybe it does :thinking_face: Im realizing now I don’t know | 2019-04-21T13:46:05.125300 | Cornell |
elmlang | general | In the redux community, performance is a huge topic and usage of libraries like reselect is very common. It seems strange that even with escape hatches and many performance optimizations (no prop drilling like in Elm) people complain about Redux's performance, but I've never seen any performance complaints/solutions for Elm, is that because the community is still small? | 2019-04-21T13:47:35.126400 | Dayna |
elmlang | general | Looks like the diff does use `===`: <https://github.com/elm/virtual-dom/blob/5a5bcf48720bc7d53461b3cd42a9f19f119c5503/src/Elm/Kernel/VirtualDom.js#L706> | 2019-04-21T13:48:43.127000 | Cornell |
elmlang | general | I saw that piece of code too, but that's for diffing the DOM, I'm talking about the process before that - deciding if dom diffing is even needed | 2019-04-21T13:49:22.128600 | Dayna |
elmlang | general | It could be that the DOM is always diff'd in Elm though, I'm not familiar with the implementation | 2019-04-21T13:50:12.129800 | Dayna |
elmlang | general | The assumption of immutability in Elm allows the compiler and runtime to do most of the optimizations that you need to do by hand with Redux and React. | 2019-04-21T13:50:47.130100 | Lashawnda |
elmlang | general | Also, deeply nested codepaths and data structures in Elm are intentionally difficult, and flatter code has lead to less expensive update function calls | 2019-04-21T13:52:31.132300 | Lashawnda |
elmlang | general | I guess so, but derived data will still create new references | 2019-04-21T13:56:19.135300 | Dayna |
elmlang | general | It tends not to matter. Even in the largest and most complicated programs we have at NoRedInk garbage collection is not a bottleneck. | 2019-04-21T13:57:39.137000 | Lashawnda |
elmlang | general | I'm not talking about GC, I'm talking about derived data like List.filter/map etc always producing a new reference even though the contents might be the same. | 2019-04-21T13:58:42.137600 | Dayna |
elmlang | general | As far as I understand lazy would skip diffing a subtree | 2019-04-21T13:59:31.138500 | Dorotha |
elmlang | general | For example, you can run a selector that uses List.map 5 times using the same model and it will produce 5 list references that have the same contents, but if you use shallow compare it will seem like 5 changes have happened, it's a huge issue when writing redux apps and that's why libraries like reselect exist. | 2019-04-21T13:59:45.138700 | Dayna |
elmlang | general | Oh I see. I’m not sure why that isn’t a big issue here. `Html.Lazy` exists for this, but there’s no data to tell us how often it is used | 2019-04-21T14:04:18.141300 | Lashawnda |
elmlang | general | So with <https://www.youtube.com/watch?v=XpDsk374LDE> (Elm Europe 2017 - Evan Czaplicki - The life of a file) in mind.
Got some modules that are becoming quite a pain to navigate, but I don't want to split out certain groups of functions into another module because I don't want to expose the `Msg` and `Model` that these functions would work on.
Started plopping these in as a way to sort of have sub modules within my modules.
```
{-|
_____ _
| ___| ___ ___ | |__ __ _ _ __
| |_ / _ \ / _ \ | '_ \ / _` | | '__|
| _| | (_) | | (_) | | |_) | | (_| | | |
|_| \___/ \___/ |_.__/ \__,_| |_|
-}
gotoFoobar : ()
gotoFoobar =
()
```
So I get a visual delimiter when scrolling (readable in minimaps as well) and a symbol with a common prefix to jump to.
Quite interested what everyone else is doing for large modules in their codebases. | 2019-04-22T00:12:02.145900 | Lizabeth |
elmlang | general | Lots of people have the classic ```
{---------------
Model
---------------}
``` banners. But if your module is getting hard to navigate then better tooling might help | 2019-04-22T00:53:07.146600 | Earlean |
elmlang | general | The Elm add-ons for most editors includes features like 'go to definition' or 'peek definition' that provide a way to find functions you're using | 2019-04-22T00:55:46.146800 | Earlean |
elmlang | general | Trying to find an approach that isn't dependant on a specific editor.
The useless `gotoFoobar` function is simply there to leverage existing 'go to definition' of most editors.
I suppose grouping functions with a common prefix would work too. | 2019-04-22T01:03:28.147000 | Lizabeth |
elmlang | general | I’ve been unable to find any information about such a feature but what I want is to be able tell my editor to represent a big file as several smaller files. | 2019-04-22T01:16:33.147300 | Jae |
elmlang | general | So I’d be able to define a new file that starts at line 500 and ends at line 1500. Then if I made changes in this virtual file, it would be inserted into the original file | 2019-04-22T01:18:36.147500 | Jae |
elmlang | general | I just put a
```
-- MODEL
-- UPDATE
-- VIEW
```
and search for the uppercase names | 2019-04-22T04:43:10.148000 | Earnest |
elmlang | general | I think vscode has bookmarks. I'm not sure, I'll check it out and will report back. | 2019-04-22T04:54:55.148400 | Lory |
elmlang | general | Hello, I am using em-verify-examples to test a function that returns a float. I am getting the following message ` Do not use Expect.equal with floats. Use Float.within instead.` What can I do about it? | 2019-04-22T05:06:29.148700 | Ayesha |
elmlang | general | You cannot really test for equality with float, due to rounding errors | 2019-04-22T05:13:18.148800 | Hoa |
elmlang | general | I’m wondering if the error is referring to this function: <https://package.elm-lang.org/packages/ianmackenzie/elm-float-extra/1.0.1/Float-Extra#equalWithin> | 2019-04-22T05:14:03.149000 | Hoa |
elmlang | general | So is it really `equalWithin`? | 2019-04-22T05:15:14.149200 | Hoa |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.