workspace
stringclasses 4
values | channel
stringclasses 4
values | text
stringlengths 1
3.93k
| ts
stringlengths 26
26
| user
stringlengths 2
11
|
|---|---|---|---|---|
elmlang | general | <@Leoma> I believe I saw a post of yours where you were doing some live editing-rendering with your markdown language. Could you point me to it? | 2019-03-18T04:49:18.324100 | Jana |
elmlang | general | <@Jana> <https://elmlang.slack.com/archives/C0CJ671HU/p1552497726223800> | 2019-03-18T09:05:23.324700 | Leoma |
elmlang | general | So this is working pretty well, EXCEPT, there are two elements working together here, a transparent textarea and a div. I’m just having some CSS struggles with how these resize in sync with each other.
Wasn’t working on the weekend, but back at it today. | 2019-03-18T09:08:19.325000 | Leoma |
elmlang | general | Richard has some workshops that might help. One specifically for beginners | 2019-03-18T09:13:16.328200 | Rosa |
elmlang | general | Is it something that can be done online? | 2019-03-18T09:42:42.328800 | Monte |
elmlang | general | Richard's course on Frontend Masters is on-line. If you are asking about our service, then yes. We can use video calls with screen sharing. I currently work with <@Robyn> that way. You can ask him about his experience :smile: | 2019-03-18T09:44:45.329000 | Dorsey |
elmlang | general | I’m a very happy customer! | 2019-03-18T10:23:42.329700 | Robyn |
elmlang | general | :heart: | 2019-03-18T10:26:32.329900 | Dorsey |
elmlang | general | What's the worst name for a scheme written in elm? | 2019-03-18T11:35:28.330600 | Lindsey |
elmlang | general | so far I've got schelme. | 2019-03-18T11:38:18.330700 | Lindsey |
elmlang | general | Hey beautiful people.
I've got chat, and I load more messages above the current ones (like slack).
I want to keep the scroll position as the messages are added.
I am being able to successfully do that, but I have a flash of one frame where the scroll position is wrong before it being corrected.
This is the code I'm using
```
{-| Maintain scroll position after adding elements
above the current view.
-}
anchorScrollPosition : String -> Viewport -> Task Browser.Dom.Error ()
anchorScrollPosition id before =
let
distanceToBottom vp =
vp.scene.height - vp.viewport.y
adjustScroll after =
Browser.Dom.setViewportOf id 0 (after.scene.height - distanceToBottom before)
in
Browser.Dom.getViewportOf id
|> Task.andThen adjustScroll
``` | 2019-03-18T11:44:11.332800 | Tawnya |
elmlang | general | I suspect this is happening because Elm is using `requestAnimationFrame` on `getViewportOf` and `setViewportOf`, which means that it draws stuff, I get the viewport, and only in the next frame is the scroll adjusted. | 2019-03-18T11:45:03.333900 | Tawnya |
elmlang | general | What I would really need is to force a reflow by getting the new scroll position and setting it to the right value *after* the content is there, but *before* it is rendered on screen. | 2019-03-18T11:46:02.335600 | Tawnya |
elmlang | general | Is that possible? | 2019-03-18T11:46:09.336000 | Tawnya |
elmlang | general | That is correct. One thing you might want to look into is using `position: sticky` in CSS, though I don't know what browser compatibility is like | 2019-03-18T11:46:27.336500 | Huong |
elmlang | general | <@Huong> What do you mean? How could that help? | 2019-03-18T11:47:18.337200 | Tawnya |
elmlang | general | I’d try handling all of this with css.
I think the way you’re trying to do things won’t work without some convoluted solution.
You’ll definitely want:
- Html.Keyed.ol
You may need:
- Flex Column Reverse (This allows the user to load a page, at the very bottom.) | 2019-03-18T12:00:08.337300 | Delois |
elmlang | general | schelme-on-you ? | 2019-03-18T12:19:24.337700 | Velia |
elmlang | general | welp, its out there as schelme now. | 2019-03-18T12:40:57.337900 | Lindsey |
elmlang | general | I meant scheme-on-you, but schelme is most likely bad enough :laughing: I don't understand it though, my English is lacking. Is it related to the German word? | 2019-03-18T12:45:22.338100 | Velia |
elmlang | general | Or another word play? | 2019-03-18T12:45:40.338300 | Velia |
elmlang | general | its just scheme with an extra L! its scheme with elm inside, but technically it should be the other way around | 2019-03-18T12:58:38.338500 | Lindsey |
elmlang | general | :man-facepalming:of course | 2019-03-18T12:59:21.338700 | Velia |
elmlang | general | I am running into this a lot at work. Think it has to do with the the fact our projects have slightly different versions of elm compiler. Elm-bugfix2, bugfix5, bugfix6 etc. Have been thinking of setting ELM_HOME env variable per project but have not tried it yet. Tell me how it goes if you try it... | 2019-03-18T14:36:33.344100 | Tabatha |
elmlang | general | See: <https://github.com/elm/compiler/issues/1846> | 2019-03-18T14:37:45.344500 | Tabatha |
elmlang | general | I handled this same situation in JS with a mutation observer that tracked when elements were added and adjusted the scroll position | 2019-03-18T15:41:43.346600 | Lorilee |
elmlang | general | *[Only for those who program in Java too]*
*Custom Types* from ELM VS *Enums* from Java
Hi!
Reading the appendix* from the guide I did not really find there is a difference in terms of cardinality between a custom type from ELM and an Enum from Java. Both will give no room for invalid inputs. I agree it is easier to create a custom type though.
What are your thoughts on this guys?
* Source:
<https://guide.elm-lang.org/appendix/types_as_sets.html>
Thank you! | 2019-03-18T21:24:43.350500 | Paulita |
elmlang | general | i can't remember if java enums are simple Ints, if so then there's one difference, a custom type can hold state or other custom types.
however the biggest difference is the handling of custom types by the compiler. Elm will fail to compile if in a case statement, you do not handle all cases of a custom type. I believe java does not do this? ( sorry, no longer program in java ) | 2019-03-18T21:34:19.352700 | Ruthann |
elmlang | general | IIRC, under the hood, each value in a Java Enum is just a singleton object. They can’t carry data because they’re all constructed at startup. And the type checking definitely isn’t detailed enough to ensure comprehensive switch coverage. | 2019-03-18T21:46:59.354200 | Dede |
elmlang | general | Like so much of Java, it’s really just syntactic sugar on objects. | 2019-03-18T21:48:00.354800 | Dede |
elmlang | general | I wish, under the hood they are just Ints, if they were proper objects maybe they could carry some state | 2019-03-18T21:49:57.355700 | Shondra |
elmlang | general | Hilariously, they are objects, and they can carry methods, and they can even carry the data with which they are constructed. Just.. .construction only happens once, so that comparison is just object comparison. | 2019-03-18T21:54:21.356700 | Dede |
elmlang | general | <https://docs.oracle.com/javase/tutorial/java/javaOO/enum.html> | 2019-03-18T21:54:28.356900 | Dede |
elmlang | general | <@Paulita> maybe this proxy is closer to ADTs in Java than Enums: <https://github.com/spotify/dataenum> | 2019-03-18T21:56:41.358200 | Ruthann |
elmlang | general | They can, those are c enums | 2019-03-18T22:06:06.358900 | Kris |
elmlang | general | I think I've found a bug in the compiler, what's the course of action? | 2019-03-19T01:51:13.359900 | Shondra |
elmlang | general | <@Shondra> write a <http://www.sscce.org/> that triggers the bug and report it at <https://github.com/elm/compiler/issues> | 2019-03-19T01:59:52.360700 | Earlean |
elmlang | general | once you have a sscce you might want to post it here first to see if anyone knows anything about it | 2019-03-19T02:00:43.361600 | Earlean |
elmlang | general | this breaks with elm: `Map.!: given key is not an element in the map` | 2019-03-19T02:46:21.362600 | Shondra |
elmlang | general | sure, what I've found tinkering around is that it only happens if I put types from a library into a Msg type of my own | 2019-03-19T02:46:27.362900 | Shondra |
elmlang | general | where Form comes from elm-form 4.0.0 | 2019-03-19T02:49:42.363300 | Shondra |
elmlang | general | but if I use it in a record it works | 2019-03-19T02:49:53.363600 | Shondra |
elmlang | general | <@Shondra> ah, that's a well known one. It's being worked on at the moment. | 2019-03-19T03:18:05.364600 | Earlean |
elmlang | general | if you're using `--debug` it's likely you can workaround this issue by not using `--debug` | 2019-03-19T03:19:07.365800 | Earlean |
elmlang | general | yep, that "fixes" it | 2019-03-19T03:25:11.366200 | Shondra |
elmlang | general | <@Earlean> is there a issue I can follow so I can know when it's fixed? | 2019-03-19T03:27:05.367200 | Shondra |
elmlang | general | <https://github.com/elm/compiler/issues?utf8=%E2%9C%93&q=is%3Aissue+is%3Aopen+map.%21> there are a bunch of issues reported | 2019-03-19T03:29:13.367500 | Earlean |
elmlang | general | oh :disappointed:, in my madness I downloaded and compiled elm, maybe I'll try to take a stab at it later :simple_smile: | 2019-03-19T03:33:44.368600 | Shondra |
elmlang | general | I wouldn’t say that’s the part that’s madness, madness is that this bug has gone so long without being fixed :wink: | 2019-03-19T06:25:56.369000 | Garnett |
elmlang | general | Indeed <@Ruthann>, maybe this proxy is closer to a Custom Type from ELM. There is even a comparison with Data types from Haskell. | 2019-03-19T07:56:04.371300 | Paulita |
elmlang | general | Elms non-reactionary development is intentional, if not frustrating | 2019-03-19T08:23:40.372100 | Danika |
elmlang | general | bug fixes ought to be treated differently than features though? | 2019-03-19T08:26:38.372400 | Nana |
elmlang | general | the fix here is simple (do not use `--debug`) and therefore it did not warrant a patch release (so far) | 2019-03-19T08:37:52.373000 | Virgie |
elmlang | general | the debug mode needs a lot of work anyway so patching it up now (while it is not really that useful) doesn't make sense | 2019-03-19T08:38:34.373500 | Virgie |
elmlang | general | <@Virgie> I use the debugger extensively. Moving to `0.19` I had a significant drop in the quality of my debugging. Up until last week I used a fork that solved some of the issues with `Map.!:` but that fork stopped working too. | 2019-03-19T08:56:46.377500 | Maida |
elmlang | general | how is it useful for you? my issues are 1) cannot read full message name/contents, 2) I get swamped by messages I don't need, hard to find the one I want | 2019-03-19T09:04:27.378600 | Virgie |
elmlang | general | hey everyone! I work on an enterprise project using Elm and it has a pretty large codebase | 2019-03-19T09:20:39.379400 | Todd |
elmlang | general | build times have become pretty unhealthy over time and I was wondering if some of you could share your build times (without previous build cache) and approx. lines of code? I'm interested to see if ours is a unique issue | 2019-03-19T09:21:17.380300 | Todd |
elmlang | general | <@Todd> are you on Elm 0.19? | 2019-03-19T09:21:58.380900 | Earlean |
elmlang | general | not yet, but I can't imagine the difference between the two versions to be that critically huge, especially with larger codebases | 2019-03-19T09:22:27.381400 | Todd |
elmlang | general | It's quite a huge difference | 2019-03-19T09:22:59.382000 | Earlean |
elmlang | general | I heard it can be around 40x? :smile: | 2019-03-19T09:23:17.382300 | Nana |
elmlang | general | 0.18 has some pathologic cases with n^2 complex | 2019-03-19T09:24:10.383300 | Earlean |
elmlang | general | If compile time is your biggest concern then moving to 0.19 should be a high priority. | 2019-03-19T09:26:45.384400 | Earlean |
elmlang | general | it is a high priority, but we also suspect that there are problems coming from modules that have too many dependencies/dependents | 2019-03-19T09:27:37.385400 | Todd |
elmlang | general | I also wanted to see what kind of experiences others have with big codebases | 2019-03-19T09:27:59.386200 | Todd |
elmlang | general | In terms of relationships between modules you want to structure your app so that modules that change often have few modules dependent on them because all modules that depend on a changed module need recompiling | 2019-03-19T09:31:59.390200 | Earlean |
elmlang | general | A common mistake is to make too many modules depedent on a module that changes often. Eg. A shared Msg or Model module imported by every other module | 2019-03-19T09:33:31.392400 | Earlean |
elmlang | general | Yes, we recently realized how much of a problem this has become. When our codebase was small, it was not a big deal, but it quickly snowballed and we didn't even notice | 2019-03-19T09:34:10.393900 | Todd |
elmlang | general | Or putting modules containing views/css (that designers are changing often) at the very bottom of a dependency chain of modules | 2019-03-19T09:35:20.395000 | Earlean |
elmlang | general | as a datapoint, our codebase went from 8 minutes -> 12 seconds | 2019-03-19T09:37:10.396500 | Huong |
elmlang | general | (for a clean compile, that is) | 2019-03-19T09:37:53.396700 | Huong |
elmlang | general | how large is this codebase?? :anguished: | 2019-03-19T09:39:22.396900 | Todd |
elmlang | general | There are a few numbers out there:
- <https://discourse.elm-lang.org/t/our-experience-upgrading-to-0-19-at-humio/3258>: 81kLOC, from 77s to 2.73s
- <https://discourse.elm-lang.org/t/contasystemer-as-current-stats-and-thank-you-the-community/3348/2>, 2.2s for 20.7kLOC
- <https://elm-lang.org/blog/small-assets-without-the-headache>, 49kLOC in 1.9s
- there are more that I cannot find right now
I would seriously port to 0.19 before refactoring anything. I have not yet heard of anyone having compilation time issues with 0.19. | 2019-03-19T09:41:49.397100 | Velia |
elmlang | general | I found this read really valuable recently | 2019-03-19T09:58:35.397700 | Teri |
elmlang | general | <https://programmingisterrible.com/post/139222674273/write-code-that-is-easy-to-delete-not-easy-to> | 2019-03-19T09:58:35.397900 | Teri |
elmlang | general | 62630loc (excluding blanks and comments) | 2019-03-19T10:07:10.398300 | Huong |
elmlang | general | which includes a couple of very large `case..of` statements, which makes a big difference | 2019-03-19T10:07:31.398500 | Huong |
elmlang | general | wow, that is fascinating! thank you for telling me this, it's going to make the upgrade a lot easier to sell hopefully | 2019-03-19T10:16:59.398900 | Todd |
elmlang | general | Can’t say I totally understand what you’re seeing there, but it appears that your open command is getting to the JS code. Don’t know if you’ve instrumented things enough to see if its getting to your socket server, and coming back to your handler through `PortFunnels.cmdPort`. | 2019-03-19T10:21:54.399100 | Granville |
elmlang | general | Thanks Bill. I can see some data coming through now. As I’m sure you experienced, it takes a lot more wiring up than for 0.18 | 2019-03-19T10:37:57.399500 | Lynn |
elmlang | general | Sometimes I know I have compilation errors but the Elm code is compiled without any unless I go to the file with error and save it to trigger the compilation process. | 2019-03-19T10:55:50.400700 | Frieda |
elmlang | general | Why that happen? | 2019-03-19T10:56:05.401100 | Frieda |
elmlang | general | It happens with webpack elm loader in watch mode | 2019-03-19T10:56:35.401600 | Frieda |
elmlang | general | I think it's a known bug in the compiler. I find myself deleting elm-stuff in its entirety a few times a day. | 2019-03-19T11:04:58.402000 | Dede |
elmlang | general | it used to happen to me a lot, but then it stopped and doesn't happen at all now, don't remember what I changed :thinking_face: (using webpack, elm-webpack-loader and elm-hot-webpack-loader) | 2019-03-19T11:06:31.402900 | Nana |
elmlang | general | <https://github.com/elm/compiler/issues/1809> | 2019-03-19T11:06:41.403100 | Dede |
elmlang | general | Thank you | 2019-03-19T11:27:04.403500 | Frieda |
elmlang | general | I have rather short message names and minimal nesting. | 2019-03-19T12:07:05.403600 | Maida |
elmlang | general | also, the main thing I’m interested in is the state of the model. | 2019-03-19T12:07:25.403800 | Maida |
elmlang | general | I would love if I would also have the full message above the model representation BUT, between what I have in the sidebar and what I have in the model, I have access to a lot of information. | 2019-03-19T12:08:34.404000 | Maida |
elmlang | general | Anyone been creating an SPA with Browser.application and realized suddenly that any given http request fired is still active after navigation? Do people then put trackers on all requests and cancel them? | 2019-03-19T13:32:28.406700 | Rosaria |
elmlang | general | Any elm folks in Budapest? I’m going to be there next week. | 2019-03-19T13:46:16.407000 | Hoyt |
elmlang | general | I just have per-page message wrappers, and discard messages intended for a page that isn't current. | 2019-03-19T13:55:31.407100 | Dede |
elmlang | general | That's what we did too. But some messages are gloablly handled and it was creating race conditions. | 2019-03-19T14:00:03.407400 | Rosaria |
elmlang | general | Hm. Can you provide more detail? | 2019-03-19T14:01:46.407600 | Dede |
elmlang | general | On init API requests fly out on a given page. If one comes back 401 then we redirect the user to a 404 page with the ability to login there. But a second in flight request triggers the same thing and we get redirected a second time. | 2019-03-19T14:11:46.407800 | Rosaria |
elmlang | general | I think having login/access denied as a "route" and using redirects is always awkward | 2019-03-19T14:21:02.408000 | Nana |
elmlang | general | Open to suggestions. It's not so much a route as it's a global message that is handled in Main, it's handled with a redirect in Main however. | 2019-03-19T14:21:59.408200 | Rosaria |
elmlang | general | nicer to just have it as a conditonal view - ie. if current route requires auth and auth is missing, display login screen. if auth is pending, display a loading indicator, if everything is fine, display the actual page | 2019-03-19T14:22:40.408400 | Nana |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.