02:20 <FromGitter> <636f7374> @wontruefree No i mean i know ysbaddaden has been working hard to make http2 available as soon as possible.

02:22 <FromGitter> <636f7374> But, I took a look at the source code, which is a bit tricky, let alone `http3` (`QUIC`).

02:23 <FromGitter> <636f7374> I haven't started researching `http2`, the `http3` project has been launched.

02:25 <FromGitter> <636f7374> `MITMProxy` already supports `http2`, it doesn't seem to support `http3`?

02:29 <FromGitter> <636f7374> TCP Fast Open (`TFO`) feature may break `HTTP::Client`, `HTTP::Server`? I see `Golang` seems to have experienced these.

02:31 <FromGitter> <636f7374> Golang#4842 https://github.com/golang/go/issues/4842

02:35 <FromGitter> <636f7374> `http3` (`QUIC`) implementation on Haskell looks interesting. 😋 https://github.com/kazu-yamamoto/quic

02:41 <FromGitter> <636f7374> Actually what I said is double meaning, I also want to say: No one wants to implement `http3` like `ysbaddaden` (implement http2 (https://github.com/ysbaddaden/http2))?

09:59 <ua_> why is there no File#readline :(

10:00 <ua_> its actually useful when processing big files you know

10:01 <ua_> so you dont have to keep gigabytes of stuff in memory?

10:05 <ua_> read_line

10:05 <ua_> inherited from IO

10:05 <ua_> jesus

10:06 <ua_> im starting to wonder if just moving to python/cython is not much better choice

10:06 <ua_> than from ruby to crystal

10:12 <FromGitter> <yxhuvud> ua: there is also `each_line`

11:30 <repo> ua_: how come?

11:31 <repo> because crystal is too similar to ruby without being actually compatible?

11:33 <repo> ua_: ruby has grown organically for many many years and is full of aliases and method namings, that were adapted from perl and such. IMO it's totally understandable if crystal doesn't want to load itself with that baggage, and implement things in a way it deems best.

11:36 <ua_> ruby syntax similarity is biggest selling point of crystal, that and speed, but if one has to check every single thing because its actually different then one might just as well go to use more mature languages that are different all the wa

11:36 <ua_> way

11:37 <repo> yes, fair point. but crystal _has_ similar syntax to ruby. It just hasn't a necessary similar stdlib. but that's also not advertised

11:37 <repo> *necessarily

11:38 <repo> i really like the object orientedness and syntax of both ruby and crystal which is why for me the switch to crystal was obvious.

11:40 <repo> on the other hand even though i was a ruby dev for 4+ years and i still have to look up docs for many ruby stdlib methods

11:40 <repo> -and

11:43 <ua_> i was under the impression that method names fall under the syntax part

11:43 <ua_> but ok

11:45 <repo> of course not

11:46 <ua_> still kinda feels like false advertising

11:47 <ua_> syntax but not stdlib should be more emphasized lol

11:49 <repo> well.. it is fairly similar still

11:49 <repo> chances are stuff just works

11:49 <repo> you don't have that with a lot of languages

11:49 <repo> but i do understand your frustration

11:49 <repo> devdocs is your friend :)

14:00 <FromGitter> <alex-lairan> Hi guy's ! :) ⏎ ⏎ I have troubles with this Warning : ⏎ ⏎ ```code paste, see link``` ... [https://gitter.im/crystal-lang/crystal?at=5e29a6fa44c076313cd14bb0]

14:02 <FromGitter> <Blacksmoke16> The type that implements that method must declare the return type

14:02 <FromGitter> <Blacksmoke16> Also, the docs are pretty clear IMO

14:02 <FromGitter> <Blacksmoke16> > Have a syntax similar to Ruby (but compatibility with it is not a goal)

14:03 <FromGitter> <Blacksmoke16> Just because the syntax is similar doesn't mean it's a clone

14:03 <FromGitter> <alex-lairan> I don't understand, forall worked in previous versions

14:04 <FromGitter> <Blacksmoke16> Oh sorry, diff warning

14:08 <FromGitter> <Blacksmoke16> Could try just removing the type restrictions, since if you're not actually using U in the method it's essentially the same as not having one

14:27 <FromGitter> <alex-lairan> Ok, thanks @Blacksmoke16 https://github.com/alex-lairan/monads/commit/5b77aa00c5f831c383fadbaf918de52c3c8b6e95

14:28 <FromGitter> <alex-lairan> So specifying return type on Crystal 0.32+ is not recommanded ?

14:30 <FromGitter> <Blacksmoke16> No, might be a bug related to forall as the return type

15:15 <sorcus> Crystal doesn't have rotate method for Slice?

16:09 <FromGitter> <lebogan> Trying to install amber on Ubuntu. ⏎ ⏎ ```code paste, see link``` ⏎ ⏎ got an error about resolving db (~> 0.7.0 ~>0.8.0). Did I miss something? [https://gitter.im/crystal-lang/crystal?at=5e29c557e177666195b677d7]

16:12 <FromGitter> <straight-shoota> @lebogan Probably an issue with version restrictions. Amber and its dependencies want different releases of `db`. You might manually edit the version in `shard.yml` to fix this.

16:13 <FromGitter> <lebogan> Okay, I'll give it a shot. Thanks

16:37 <FromGitter> <lebogan> @straight-shoota , after commenting and uncommenting db stuff, it seems like micrate is the culprit. I tried 0.3.0, 0.4.0, 0.5.0. Without it the make works. Hmmm

16:45 <FromGitter> <lebogan> @straight-shoota I changed the db version in my local lib/micrate and make worked. shards update still doesn't. Should I be concerned?

17:58 <FromGitter> <grkek> @Blacksmoke16 would you mind checking out the PR

18:19 <FromGitter> <Blacksmoke16> Can when I get home tomorrow night, am out of town for a bit

21:05 <FromGitter> <jwoertink> Question if there's a way to do this. You know when you have a method that returns `Thing?`, then you do `if thing`, and inside that if branch, the compiler knows that it exists? I have a method I'm passing the `thing` in to `if check(thing)`, and in my if branch I have to do `thing = thing.as(Thing)` or call` not_nil!`.

21:05 <FromGitter> <jwoertink> Is there another way to do this?

21:05 <FromGitter> <tenebrousedge> `try`

21:06 <FromGitter> <jwoertink> Right, but I'd have to do it several times.

21:07 <FromGitter> <jwoertink> I mean a way to tell the crystal that inside that branch, the object isn't nil without doing the variable reassignment

21:07 <FromGitter> <tenebrousedge> `try` or `not_nil!` or nested `if`s are kinda what you have to work with

21:07 <FromGitter> <tenebrousedge> at least at runtime

21:08 <FromGitter> <tenebrousedge> what are you trying to do?

21:09 <FromGitter> <jwoertink> ```code paste, see link``` [https://gitter.im/crystal-lang/crystal?at=5e2a0b85f85dba0aab02448b]

21:09 <FromGitter> <jwoertink> That's what I have currently which is cool and all, but I had to change it

21:10 <FromGitter> <jwoertink> ```code paste, see link``` [https://gitter.im/crystal-lang/crystal?at=5e2a0bb3cd53a80aac13515f]

21:10 <FromGitter> <jwoertink> but now I have this

21:11 <FromGitter> <jwoertink> I guess I could do `if thing && check(thing)` 🤔

21:11 <FromGitter> <jwoertink> yeah, I'll do that. Thanks!

21:11 <FromGitter> <tenebrousedge> :plus1:

21:35 <FromGitter> <tenebrousedge> Kai wrote a codewars kata https://www.codewars.com/kata/5e29d57ac6b9af00333abd19 ⏎ It's sorta derivative, but still good

22:36 <raz> why did the homepage motto change btw? 'fast as C, slick as ruby" was great.

22:38 <raz> "a language for humans and computers" sounds like placeholder text

22:38 <raz> like duh, who else would it be for ¯\_(ツ)_/¯

22:39 <FromGitter> <tenebrousedge> I would imagine that the speed claims were arguable, and the similarity to Ruby equally so

22:39 <raz> yes! that's part of what made it great

22:40 <raz> plenty publicity from kids getting their panties in a knot over it on reddit and hn

22:40 <FromGitter> <tenebrousedge> personally, I don't approve of trolling, for any purpose

22:41 <raz> nothing trolly about, it's the best one-line summary i've heard (and still give!) up to this day

22:41 <raz> i certainly don't say... "uh yea..um.. so, it's a language for humans and also for computers"