workspace
stringclasses 4
values | channel
stringclasses 4
values | text
stringlengths 1
3.93k
| ts
stringlengths 26
26
| user
stringlengths 2
11
|
|---|---|---|---|---|
elmlang | general | i dont see any difference | 2019-04-09T06:25:37.801100 | Sunni |
elmlang | general | They can also overwrite anything in `/usr/bin` | 2019-04-09T06:25:59.801300 | Huong |
elmlang | general | even stuff `root` might execute | 2019-04-09T06:26:20.801500 | Huong |
elmlang | general | not necessarily we are giving permissions to folder individuals which has no write access to the group there will not be affected by this | 2019-04-09T06:29:48.801700 | Sunni |
elmlang | general | thats what i meant by not giving -R flag | 2019-04-09T06:30:31.801900 | Sunni |
elmlang | general | - you give your user permission to write to `/usr/bin`
- you do something like `npm install -g seemingly-innocent-package`
- that package can now, while installing, overwrite files in `/usr/bin` | 2019-04-09T06:31:36.802100 | Huong |
elmlang | general | Anyway, long story short: more restrictive permissions is better for security, so when there is an option to do things without widening permissions, that's generally to be preferred. | 2019-04-09T06:38:16.802300 | Huong |
elmlang | general | This link explains how to do safe global installs with npm without requiring root: <https://docs.npmjs.com/resolving-eacces-permissions-errors-when-installing-packages-globally#manually-change-npms-default-directory> | 2019-04-09T06:39:05.802500 | Huong |
elmlang | general | totally agreed :slightly_smiling_face: and again overriding default dir of npm is a way way better solution i just didnt know that | 2019-04-09T06:43:03.802700 | Sunni |
elmlang | general | RSS feeds of Elm packages releases with a website. | 2019-04-09T06:43:21.802900 | Velia |
elmlang | general | yeah, I've tried that, and at least on my Ubuntu machine it broke stuff and still had the same permissions issues | 2019-04-09T07:33:18.803300 | Gertrude |
elmlang | general | manually changing npm's location to a world-writable directory is the best option by far, especially because some packages download raw binaries instead of only source code | 2019-04-09T07:46:59.805400 | Inga |
elmlang | general | on other OSes and even some other Linux distros, npm ships by default to install to a directory in $HOME rather than global root | 2019-04-09T07:48:00.807200 | Gertrude |
elmlang | general | I don't understand why Ubuntu insists on the opposite. Even the NodeSource packages do it | 2019-04-09T07:48:15.808100 | Gertrude |
elmlang | general | and I advise to do it as soon as you can because after you do it, your old packages' linked executables will still be pointing at /usr/lib | 2019-04-09T07:48:37.808800 | Inga |
elmlang | general | meaning you'll have to reinstall everything to the new location or manually move the packages and re-link everything in $PATH, which is ambitious | 2019-04-09T07:49:23.810200 | Inga |
elmlang | general | for me it was weird that even /opt restricts access from the users group, so I just created a /programs world-writable directory where I install e.g. programs from Github and other stuff not coming from APT | 2019-04-09T07:54:17.814300 | Inga |
elmlang | general | and moved NPM's home to /programs/npm-modules | 2019-04-09T07:54:54.815000 | Inga |
elmlang | general | never had a single issue with any permission after it | 2019-04-09T07:55:13.815600 | Inga |
elmlang | general | Thanks for the kind words! I have subscriptions working in the upstream app, but I need to port an example back to the public repo. Probably by tomorrow afternoon (US eastern time zone.) | 2019-04-09T08:42:42.816200 | Dede |
elmlang | general | does `--optimize` have a big effect on runtime performance, or is it mostly bundle size? | 2019-04-09T08:52:13.817000 | Nana |
elmlang | general | Depends on the code I assume, but for my hobby project it’s a huge performance boost as well. | 2019-04-09T08:54:49.817600 | Timika |
elmlang | general | how huge? :smile: | 2019-04-09T08:55:27.817900 | Nana |
elmlang | general | like 2x or 10x? | 2019-04-09T08:56:04.818100 | Nana |
elmlang | general | No, not _that_ huge :smile: | 2019-04-09T08:56:18.818300 | Timika |
elmlang | general | I can only give you anecdotal fps values, but emulation ran like 10-20fps faster, just by adding ---optimize | 2019-04-09T08:56:39.818500 | Timika |
elmlang | general | Mostly because simple custom types just become integers and some opaque types are removed. | 2019-04-09T08:57:35.818700 | Timika |
elmlang | general | I wonder how much performance would be gained in a typical CRUD app. | 2019-04-09T08:58:02.819000 | Timika |
elmlang | general | The difference between the app I'm working on now in local, vs. the optimize build that gets deployed, is fairly noticeable | 2019-04-09T09:11:30.819300 | Gertrude |
elmlang | general | I don't have an instrumentation on it, but it's noticeably snappy, only slowed by network access | 2019-04-09T09:12:14.819500 | Gertrude |
elmlang | general | I’m looking forward to read ;D, thank you! | 2019-04-09T09:12:30.819800 | Shantell |
elmlang | general | The discovery about flip being removed in 0.19 left me hurting. It leads to the creation of things like this: <https://github.com/pilatch/flip/tree/1.0.0>.
is it really okay to remove a piece such fundamental FP helper simply because it's unknown to people not knowing FP? I think it will lead to people auto-including packages from there along with core wether they end up using it or not, or bringing up the next "lodash" to Elm. | 2019-04-09T09:54:46.822800 | Inga |
elmlang | general | Same applies to curry and uncurry, I do not want to have to think about installing those packages when I want to use them, so I'll just install by default in every project. | 2019-04-09T09:55:54.824700 | Inga |
elmlang | general | Then it becomes even worse than before because a beginner reading a code base will bump into those functions and not have been introduced to them. | 2019-04-09T09:56:54.826600 | Inga |
elmlang | general | As someone who came to Elm from a more imperative tradition, I haven’t felt pain in not having those. It may be that Haskellers will reinvent Haskell in Elm, but the Lispers will do the same, and etc. | 2019-04-09T09:57:34.827400 | Dede |
elmlang | general | well, have you taken the time to consider possible benefits? | 2019-04-09T09:58:27.828500 | Inga |
elmlang | general | The benevolent dictators have decided flip makes codebases harder to read, discussion over. | 2019-04-09T09:58:54.829300 | Niesha |
elmlang | general | Evan doesn't remove things from the language without thorough consideration, though... | 2019-04-09T09:59:38.831400 | Sharon |
elmlang | general | I'm not arguing to bring it back, I'm just saying it will happen that people will reach for those third-party packages when they need them | 2019-04-09T09:59:44.831800 | Inga |
elmlang | general | Me? No, there’s far too many people who would like to sell me possible benefits of the one true way for me to assess them all. I’ve found it comfortable to take Elm as given and work with it as much as possible. There are places I chafe (setters!!) but this isn’t one of them. | 2019-04-09T09:59:47.832000 | Dede |
elmlang | general | I’m curious — is there a formal place where people make pitches to Evan? I’m pretty sure Slack isn’t it…. maybe the Discourse? | 2019-04-09T10:00:40.832900 | Dede |
elmlang | general | The recommendation is to not use `flip` and instead in the rare occasions you might use it to instead write out a function | 2019-04-09T10:02:59.837800 | Earlean |
elmlang | general | To be clear I'm not pitching for anything, nor do I want to argue about Elm's philosophy, I just want to have a feel about your opinions adding this extra complication for the sake of making code "beginner-friendly" | 2019-04-09T10:03:09.838000 | Inga |
elmlang | general | I love Elm’s small size. | 2019-04-09T10:03:44.839500 | Dede |
elmlang | general | I guess it only becomes a problem if the libraries become popular? | 2019-04-09T10:03:52.840200 | Inga |
elmlang | general | I think that’s an advantage to beginners and advanced developers alike. | 2019-04-09T10:03:55.840400 | Dede |
elmlang | general | (you asked for my opinion ;-) | 2019-04-09T10:04:09.841200 | Dede |
elmlang | general | The way I see is: we're adding this extra step you have to do when using those "banished complicated functions", and their use will also be shunned upon or discouraged, so we're hinting for you not to use them... BUT... | 2019-04-09T10:06:10.845300 | Inga |
elmlang | general | It's not just for beginners, flip is often confusing for people well familiar with it | 2019-04-09T10:06:23.845700 | Earlean |
elmlang | general | When I see `flip (/)` I must pause and think hard about it, but `\x y -> y / x` requires little thought | 2019-04-09T10:07:07.846700 | Earlean |
elmlang | general | These kinds of discussions are a bit useless, the decision’s been made from above recently, and such things don’t change. | 2019-04-09T10:09:00.850800 | Kris |
elmlang | general | I guess it's aggravated for me that all of those, curry, uncurry and flip are such core pieces for expressing functional programming, and it saddens me to see them taken way
but let me reiterate that I don't want to start a discussion or anything | 2019-04-09T10:09:44.852000 | Inga |
elmlang | general | It's not like you can't use them if you want to | 2019-04-09T10:09:56.852700 | Inga |
elmlang | general | Wait, that makes no sense, you’ve definitely come here to start a discussion :wink: | 2019-04-09T10:09:59.852800 | Dede |
elmlang | general | I don't want to argue or prove a certain point so it doesn't qualify as a discussion, I just wanted to know your viewpoints | 2019-04-09T10:11:52.855600 | Inga |
elmlang | general | For 'core pieces' they were used very rarely in Elm code even before they were removed | 2019-04-09T10:12:17.856800 | Earlean |
elmlang | general | so thanks everyone that gave me some feedback | 2019-04-09T10:12:19.856900 | Inga |
elmlang | general | I actually only have started to know Elm since two days ago so I'm trying to get a feel for how things are | 2019-04-09T10:14:23.860200 | Inga |
elmlang | general | <@Earlean> is there a formal way for people to make language proposals to the Elm core team? Or is that not so much of interest? | 2019-04-09T10:14:39.860900 | Dede |
elmlang | general | It's also worth considering that Elm is made for javascript developers, not haskell <@Inga> | 2019-04-09T10:16:31.862400 | Danika |
elmlang | general | purescript exists if you want some haskell clone on the front end | 2019-04-09T10:16:58.862700 | Danika |
elmlang | general | we want *everyone* and all the *folks* to feel welcome, don't we? | 2019-04-09T10:18:01.864500 | Sharon |
elmlang | general | <@Dede> there isn't as far as I know. A post to the Discourse is the best bet | 2019-04-09T10:18:11.864900 | Earlean |
elmlang | general | Awesome, thanks. | 2019-04-09T10:18:28.865500 | Dede |
elmlang | general | Well of course, I just mean from a design decision perspective | 2019-04-09T10:18:43.866400 | Danika |
elmlang | general | What <@Earlean> said, its also helpful when making such a proposal to include prior research in the field around the proposal and data that you've collected from the community and your own experiences. Evan highly values that sort of research. | 2019-04-09T10:19:16.868700 | Teddy |
elmlang | general | Yeah the post should show research in to the feature, how it's handled in other languages, and other solutions to the same problem etc | 2019-04-09T10:20:47.871500 | Earlean |
elmlang | general | This was mentioned in one of Evan's talks | 2019-04-09T10:21:30.873800 | Earlean |
elmlang | general | Elm as a whole seems to have a bunch of niceties that makes it enjoyable to work with, so that's where I'm weighing most of my likeness towards... I do care about having something that works nicely
For instance, the reason why I don't code in Haskell, or Purescript, or any of that is due to the toolchain's complications and slow feedback in general you get throughout the whole development process
Meanwhile with elm-make and the other tooling I get very fast feedback and it feels much more at home for me
I don't think it's helpful to say "just ditch Elm if you disagree with X" or have the narrative of "trying to make X in Elm is bad" because of all of the extra reasons you would choose to work with something. | 2019-04-09T10:28:18.884000 | Inga |
elmlang | general | Being such a "small" language is one great feature of Elm.
I use C# at work and I hate the syntax hell we face every day. | 2019-04-09T10:28:28.884400 | Malinda |
elmlang | general | > For instance, the reason why I don't code in Haskell, or Purescript, or any of that is due to the toolchain's complications and slow feedback in general you get throughout the whole development process
PureScript is quite fast so far. Dunno about slow. | 2019-04-09T10:30:05.888700 | Niesha |
elmlang | general | granted I of course trust what the language's creators are doing, otherwise I wouldn't be here in the first place | 2019-04-09T10:30:16.889400 | Inga |
elmlang | general | PureScript is awesome. | 2019-04-09T10:31:15.890600 | Huong |
elmlang | general | Said it a bit before, but a way to do away with unnecessary lambdas and their arguments *and* `flip` is to have functions with the arguments flipped and good names:
- `\x -> x / 10` => `divideBy 10`
- `\x -> x - 3` => `minus 3`
- etc. | 2019-04-09T10:31:34.891100 | Antonette |
elmlang | general | yes <@Antonette> and point taken, between this exchange and your comment there I was able to form an opinion | 2019-04-09T10:32:39.892700 | Inga |
elmlang | general | Works for backend too. | 2019-04-09T10:34:58.895500 | Niesha |
elmlang | general | To add a bit on this, one of my own "pains" in Elm compared to Haskell is the lack of typeclasses ; with them we could have optics that compose with `>>` ; but since composing is the main reason optics are interesting, we added lots of composition functions so we could have an idiomatic way to do so. | 2019-04-09T10:35:27.896100 | Antonette |
elmlang | general | So, instead of having it the Haskell way:
```
optionalFromAtoD = lensFromAtoB >> optionalFromBtoC >> prismFromCtoD
```
We can still do:
```
optionalFromAtoD =
lensFromAtoB
|> Compose.lensWithOptional optionalFromBtoC
|> Compose.optionalWithPrism prismFromCtoD
``` | 2019-04-09T10:37:26.898300 | Antonette |
elmlang | general | On the minus side: more functions and boilerplate-y stuff to care about. | 2019-04-09T10:37:53.898800 | Antonette |
elmlang | general | On the plus side: it's kind of a little bit more readable if you squint hard. | 2019-04-09T10:38:13.899300 | Antonette |
elmlang | general | I have a good number of gripes about Elm, but I keep for myself unless asked. :slightly_smiling_face: I will never agree 100% with Evan on everything. As long as the general good feeling about the language survives I’m hanging around here. | 2019-04-09T10:42:39.907100 | Hoa |
elmlang | general | Im beginning to really annoy my Haskell friends with my questions about why custom typeclasses are needed at all. Im about 6,000 LOC into a Haskell project, and so far for me, typeclasses in practice are just having to tediously add `deriving (Eq)` or writing an `instance Show Thing where` for all my types. | 2019-04-09T10:44:20.909400 | Ashton |
elmlang | general | 99% of the time I disregard "pain points" because I understand it's a different beast, and even before that, the question was raised to bring a core-lib decision to the table (as in, affects the whole) which is not the same as a feature change (e.g. the removal of custom operators, which I don't mind either, but that's once again beside the point)
just to REALLY say once more that I don't want to have anything beyond what was asked, nor I think it's right or wrong | 2019-04-09T10:44:47.910200 | Inga |
elmlang | general | Are you sometimes using `do` syntax? | 2019-04-09T10:46:05.911600 | Antonette |
elmlang | general | Or `fmap`? | 2019-04-09T10:46:18.911800 | Antonette |
elmlang | general | Or optics? | 2019-04-09T10:46:23.912000 | Antonette |
elmlang | general | Ive avoided `do` and `fmap`, but I use them when I have to. | 2019-04-09T10:46:34.912200 | Ashton |
elmlang | general | I dont know what optics are. Is that like lenses? | 2019-04-09T10:46:42.912400 | Ashton |
elmlang | general | Yep :wink: lenses are optics. | 2019-04-09T10:47:19.912700 | Antonette |
elmlang | general | The kinds of types I `map` are `Maybe`, `Either`, and `List`. For those I do `List.map` and `Data.Either.Extra.mapRight` like I might in Elm. | 2019-04-09T10:47:57.913100 | Ashton |
elmlang | general | Are you suggesting that typeclasses are much more valuable to codebases that use `do` and `fmap` a lot? | 2019-04-09T10:48:41.914000 | Ashton |
elmlang | general | They're necessary to `do` syntax and `fmap`. | 2019-04-09T10:49:01.914400 | Antonette |
elmlang | general | `do` syntax relies on the "invisible" type being a `Monad` instance (thus having a `>>=`/`andThen` implementation) | 2019-04-09T10:49:43.914700 | Antonette |
elmlang | general | `fmap` is the generic `map` that maps over any kind of functor. | 2019-04-09T10:50:10.914900 | Antonette |
elmlang | general | Now, the cool thing with those is that they allow you to not rewrite any functions that rely on the core typeclass functions to provide that functions functionality. | 2019-04-09T10:50:57.915100 | Antonette |
elmlang | general | <@Ashton> What kind of project are you working on? Seems strange that in 6k LOC you didn't have to work with a foreign data format (derive FromJSON with aeson), or had to serialize data for a database (derive FromRow with postgresql), both of which you have to write boilerplate decoders/encoders in Elm for. | 2019-04-09T10:52:05.915800 | Dayna |
elmlang | general | It's a bit less of a pain in Elm since there's less need for monadic acrobatics (like using monad transformer and such), and I agree implementing them might lead developers making quite unreadable codebases. | 2019-04-09T10:52:06.915900 | Antonette |
elmlang | general | The main reason I'd like them sometimes is optics (since I've worked with codebases where models were deeply nested) :wink: | 2019-04-09T10:53:39.916100 | Antonette |
elmlang | general | But then again I've worked around the main pain point by adding the `Compose` module to `elm-monocle` and it works pretty well for me now. | 2019-04-09T10:54:38.916300 | Antonette |
elmlang | general | I spent a bunch of time researching the best optimize / compression settings for elm and uglify. This library is a homegrown bundler that does the basic things that we need to make our app tiny.
The uglify settings are here. This lib is opinionated (uses SASS for ex) and has config for auto-recompile and production bundling. Feel free to use it or use it as a starting point for your bundling needs.
<https://gitlab.com/doanythingfordethklok/ms-pacman/blob/master/lib/script.js#L110> | 2019-04-09T10:58:21.917400 | Charity |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.