00:03 <hightower2> I have the code for it somewhere, will submit to the Benchmark module

01:28 <hightower2> I'm missing the Array#inject method

01:33 <FromGitter> <bigtunacan> @hightower2 you are looking for map

01:34 <hightower2> Yes and no, .inject is not the same - it keeps an implicit "accumulator" for the final value, and the result is that value rather than an array of elements

01:37 <FromGitter> <bigtunacan> Lol, yeah I just realized that

01:38 <FromGitter> <bigtunacan> I was thinking it was an alias for map, but it's an alias for reduce

01:38 <FromGitter> <bigtunacan> Which I don't see reduce defined on Array either

01:38 <hightower2> oh indeed, great

01:39 <FromGitter> <bigtunacan> Ah reduce is there

01:40 <FromGitter> <bigtunacan> Looks like it was inherited from Enumerable

01:40 <FromGitter> <bigtunacan> The General approach of Crystal is to not use a bunch of aliases with in the class, but two instead choose a single method name to reduce mental model.

01:48 <crystal-gh> [crystal] asterite pushed 1 new commit to master: https://git.io/vHDCv

01:48 <crystal-gh> crystal/master 4a9b867 Ary Borenszweig: Array: fix some bugs in #fill. Fixes #4539

02:19 <travis-ci> crystal-lang/crystal#4a9b867 (master - Array: fix some bugs in #fill. Fixes #4539): The build passed. https://travis-ci.org/crystal-lang/crystal/builds/241411952

02:19 <DeBot> https://github.com/crystal-lang/crystal/issues/4539 (Array#fill may result in an invalid memory access when given a negative argument.)

02:31 <FromGitter> <faultyserver> @bigtunacan I don't know how far you've gotten with the theory, but I've started writing up an introduction to compilers for a (currently) fictional language: https://github.com/myst-lang/myst/blob/master/COMPILERS_INTRO.md. It goes through some small examples of each step of a dynamic language compiler (lexing, parsing, visiting, and a VM), trying to help bridge the gap between the purely-theoretical side

02:31 <FromGitter> ... of compilers in most books today and real, practical compilers in the world today.

02:31 <FromGitter> <faultyserver> disclaimer: I don't claim to know everything about compilers, I just got annoyed that I couldn't find any good examples of anything other than a tree-walking compiler, which is pretty much useless in the real world

02:32 <FromGitter> <faultyserver> Charly lang (https://github.com/charly-lang/charly/blob/master/src/charly/syntax/parser.cr) is also a good example of a compiler that hasn't grown too much

02:47 <FromGitter> <faultyserver> I'm also just starting the implementation of that (currently) fictional language, so it could be interesting to watch how it grows from nothing to something basic?

08:08 <hightower2> Does crystal have "automatic" variable inside blocks, like 'it' in Groovy?

08:10 <FromGitter> <sdogruyol> nope

08:18 <hightower2> Have some weird issue. I do p myvars.map{|x|x.class} to confirm I am using all Float64s. Then I call another function with those vars (and perform the same p to make sure they're still Float64s). Then I subtract two of them there, and get: Couldn't find overloads for these types: - Float64#-(String)

08:19 <FromGitter> <sdogruyol> code please

08:20 <hightower2> Yeah, trying to come up with a usable isolated paset

08:20 <hightower2> paste*

08:28 <hightower2> This is what is happening - https://gist.github.com/docelic/318466c091367df7e0c1f44c1da4dd4b

08:28 <hightower2> It is only the final step of the problem, I need to work on isolating the code before

08:29 <hightower2> But this is the final 3 lines that give me trouble

08:32 <FromGitter> <mverzilli> that generally means at some point of your program `current_average` is assigned a String

08:33 <hightower2> verzilli right of course, but these 3 lines execute directly in the order shown. They report their class as Float64, and immediately in the next line I do a subtract and get this error.

08:35 <FromGitter> <mverzilli> without seeing a little more code is hard to tell. it could be a bug in Crystal's "class" method

08:35 <hightower2> I am working on trying to come up with an isolated example

08:38 <FromGitter> <mverzilli> oh, wait

08:38 <FromGitter> <mverzilli> "class" returns the runtime class

08:38 <FromGitter> <mverzilli> run this and see what happens

08:40 <FromGitter> <mverzilli> ```code paste, see link``` [https://gitter.im/crystal-lang/crystal?at=593bb06c02c480e6722daacb]

08:42 <hightower2> It reports: (Float64 | Int32 | String) for both variables

08:42 <FromGitter> <mverzilli> `current_average.class` should be returning `String`

08:42 <FromGitter> <mverzilli> while `typeof(current_average` should be `Float64 | String`

08:43 <FromGitter> <mverzilli> are you running the exact code snippet I pasted here?

08:43 <hightower2> These values are coming from a text file, and I produce them using: File.read( "test.txt").split(' ').map{ |value| value.to_f}

08:44 <FromGitter> <mverzilli> right, I just produced a snippet to explain what's happening

08:44 <hightower2> mverzilli oh no, I added typeof() to my code and reported that. The one you pasted, if ran exactly as you gave it, works like you report

08:45 <FromGitter> <mverzilli> cool, that is to show the difference between typeof and class

08:45 <FromGitter> <mverzilli> for the compiler, current_average might be a String or a Float64

08:46 <FromGitter> <mverzilli> so it can't let you run `-`, because there's no way to subtract a String from a Float

08:46 <crystal-gh> [crystal] ujifgc opened pull request #4540: add HTTP::STATUS_CODES, use official IANA Registry (master...http-status) https://git.io/vHD2O

08:46 <hightower2> How does it figure out it can be a string when it is produced as a result of string.to_f ?

08:46 <FromGitter> <mverzilli> `class` returns `Float64` because that's what it is at runtime

08:47 <hightower2> and/or how to avoid that then?

08:48 <FromGitter> <mverzilli> to understand why it says it could be a string we would need to see more code

08:49 <FromGitter> <mverzilli> in general, you don't want to avoid that, what you want to do is find out where you might missing something... this is Crystal catching a potential bug for you

08:49 <hightower2> Right, sure, though the code is simple, don't see a problem. I'm going over it once again.

08:50 <hightower2> Btw, how do I "expand" an Array into list of arguments, e.g. if I have arr.size == 4, how do I call a method with 4 arguments? *arr doesn't seem to work.

08:53 <FromGitter> <mverzilli> you can do it with tuples, not with arrays AFAIK

08:56 <hightower2> oh wow, not only that worked, but it got rid of the Float64/String problem. Now I am returning {1,2,3,4} from my function, whereas previously I was returning [1,2,3,4]. And I call nextmethod(*tuple) now instead of nextmethod(arr[0],arr[1],arr[2],arr[3]) and that seems to work

08:57 <FromGitter> <mverzilli> nice, and as a bonus now you're removing some pressure from the GC :)

09:36 <hightower2> Hm, how can I run Benchmark.ips() without having it print anything to the screen? I don't see a way to prevent report() from being called.

09:41 <hightower2> Oh I see, by using Benchmark::Ips directly instead of via Benchmark.ips

11:07 <FromGitter> <ultra2mh> hi i have a problem can someone please help me

11:07 <FromGitter> <sdogruyol> hey

11:07 <FromGitter> <ultra2mh> im building app with kemal framework

11:08 <FromGitter> <ultra2mh> how should i use bootstrap.css in layout file ?

11:10 <FromGitter> <ultra2mh> i use file location directly but it dont load

11:10 <FromGitter> <sdogruyol> check this repo and code https://github.com/f/kemal-react-chat/blob/master/src/views/index.ecr#L8 @ultra2mh

11:13 <FromGitter> <ultra2mh> thanks simple ...but im sure i use correct path for file but not load

11:13 <FromGitter> <sdogruyol> where's your file located

11:13 <FromGitter> <sdogruyol> under /public or /src/public?

11:14 <FromGitter> <ultra2mh> under src/public/css/

11:14 <FromGitter> <ultra2mh> i put bootstrap and style.css in latest folder "css"

11:15 <FromGitter> <sdogruyol> you should not put that in src

11:15 <FromGitter> <ultra2mh> what is correct path?

11:16 <FromGitter> <ultra2mh> where shoud i locate my css files

11:16 <FromGitter> <sdogruyol> (https://files.gitter.im/crystal-lang/crystal/X1KP/Screen-Shot-2017-06-10-at-14.16.05.png)

11:16 <FromGitter> <sdogruyol> public is a root folder just like src

11:19 <FromGitter> <ultra2mh> thankyou so much

11:19 <FromGitter> <ultra2mh> solved

11:19 <FromGitter> <sdogruyol> cool :+1:

14:17 <FromGitter> <domgetter> Is it possible to create a Linux kernel module with Crystal or does a crystal program depend on userspace stuff?

14:29 <FromGitter> <mhsjlw> i guess you technically *could*...

14:29 <FromGitter> <mhsjlw> if there is a way to disable the standard library

14:30 <FromGitter> <mhsjlw> but to be honest i'd really just expose a C api from Crystal and write the kernel module itself in C

14:30 <FromGitter> <mhsjlw> @domgetter

16:43 <hightower4> How would I write a YAML.mapping() for Benchmark::IPS::Entry class? I have problems with 'action' which expects a block, and with all other properties that are required or must have non-null value.

16:48 <FromGitter> <bararchy> @domgetter some people have already created a whole Kernel with Crystal

16:57 <Papierkorb> RX14: Linux Kernel module in Crystal? Challenge accepted.

16:57 <FromGitter> <bew> ahahah

16:58 <Papierkorb> I however can predict the outcome: Yes it's possible, but it's not a practical idea

16:59 <Papierkorb> <mhsjlw> if there is a way to disable the standard library <- In fact you can by passing `--prelude empty` to the `build` command.

17:01 <Papierkorb> But it's a pain. The Int#/ implementation in the stdlib is not decoration, it is used and necessary if you need such fancy features such as integer division.

17:38 <FromGitter> <bararchy> Well :) ⏎ ⏎ https://github.com/TheKernelCorp/NuummiteOS

18:40 <hightower4> I have: @archive = YAML.parse( File.read "yaml").as_h and I am getting: Can't infer the type of instance variable '@archive'

18:41 <hightower4> The YAML content corresponds to {} of String => Array(SomeClass), but I don't see where to specify that, since these types are incompatible with YAML::Any/YAML::Type or whatever YAML module does

19:12 <oprypin> @jwoertink, it looks like you're writing a game framework. any reason that you're passing up CrSFML? seems like you're reinventing a lot of what it has. oh and its tutorials would also give you basic examples for the game loop.

19:55 <Papierkorb> hightower2: Prefer `YAML.mapping` over `YAML.parse` if possible

19:57 <Papierkorb> hightower2: Crystal can't know that YAML::Any#as_h's hash will be compatible with your hash. In fact, `YAML.parse` can't know that it's an Array of `SomeClass`. You can do all this though `YAML.mapping` easily

21:17 <crystal-gh> [crystal] oprypin opened pull request #4542: Turn Random::System into a module (+docs) (master...random-module) https://git.io/vHyk6

22:34 <FromGitter> <jwoertink> @oprypin I started work on my framework before I saw CrSFML, and contributing to the SDL shard with @ysbaddaden has been a great learning experience for me.

22:53 <FromGitter> <bigtunacan> @jwoertink @oprypin For people that are already familiar with SDL or SFML some don't want to switch so having both supported seems good

23:28 <hightower2> Papierkorb, I have trouble with non

23:29 <hightower2> ... with the object properties using YAML.mapping. For each property, it complains that it is not clear if it's a mistake or a property is nulable.

23:30 <hightower2> and also the class I want to map has a block as one of the required arguments in the constructor, and I didn't figure out what to do with that part.

23:31 <hightower2> And also what if my YAML isn't just that object, but e.g. an array of objects, or e.g. a Hash of Scalar => Array(SomeClass)