Disqus - Latest Comments for martinos

Re: Introducing DigitalOcean Marketplace: Our Platform for Preconfigured 1-Click Apps and Tools

martinos — Fri, 08 Mar 2019 14:56:48 -0000

The big issue that I see for the moment is there are no ssl certificate installed on the droplets. I think that It would be great to have at least a procedure to add them.

Re: yield_stelf is more awesome than you could think

martinos — Fri, 28 Dec 2018 15:46:55 -0000

I think that ‘apply’ would have been clearer.

value.apply(&:computation)

Re: ReasonML: functions

martinos — Fri, 09 Mar 2018 06:10:37 -0000

I am wondering why you say the |> is for function composition . I am not able to compose two function together with it. I cannot do let add3 = add1 |> add2. In Elm the composition operator is >>, Is there such an operator in ReasonML?

Re: Deploying an Elixir + Phoenix + Elm Application with Nanobox

martinos — Wed, 17 Jan 2018 16:41:16 -0000

There is also two typos in:
li/nanoapp_web/template/layout/app.html.eex

it should be:
lib/nanoapp_web/templates/layout/app.html.eex

Re: Deploying an Elixir + Phoenix + Elm Application with Nanobox

martinos — Wed, 17 Jan 2018 16:17:28 -0000

The boxfile example that you show contains invalid chars. All the "-" are replaced by another unicode char. If you copy paste your example it won't work. Just to let you know.

Re: Simple Functional Strong Params In Ruby

martinos — Sun, 09 Jul 2017 18:53:09 -0000

Here's a small benchmark:

https://gist.github.com/mar...

10 micro second per call.

Re: Simple Functional Strong Params In Ruby

martinos — Sun, 09 Jul 2017 18:43:51 -0000

I wrote this post in order to explain partial application, I used strong parameters as an example to show how it can make code composable. I didn't want people to use this necessary, but I wanted to demonstrate that it can help solve problems in a different way. I agree that It's not very easy to understand at first, but once you got it it opens a new world.

I am wondering why you think that this solution is complicated? I would like to know maybe I could improve my post.

I agree that OO programmer are not used to think that a function can have each of its parameters applied in different context. In languages such Haskell, Elm and F# every function is curried, so it's very familiar to FP programmers. Not because we are not familiar with a concept that makes it complex.
This Rich Hickey presentation explains it well : https://www.infoq.com/prese....

If you're wondering about benchmarking, I think that this is really negligible. If it take 1 ms for processing a params hash, this will not affect the throughput in 99.9% of cases.

Re: Simple Functional Strong Params In Ruby

martinos — Fri, 07 Jul 2017 22:32:23 -0000

Thanks for the correction ! If you are interested in FP I think that Elm is a very nice language to start with and I think it teaches you better functional programming concept than Elixir. Elixir has its merit but it is not the best for that.

Re: Simple Functional Strong Params In Ruby

martinos — Fri, 07 Jul 2017 16:16:31 -0000

Thanks, I didn't know about itself.

Re: Simple Functional Strong Params In Ruby

martinos — Fri, 07 Jul 2017 09:06:28 -0000

I will create one, but I don't know exactly when, I am pretty busy at the moment. Thanks for your interest.

Re: How Elm Slays a UI Antipattern - Kris Jenkins' Blog

martinos — Wed, 18 Jan 2017 11:55:04 -0000

Last week I did an small Elm app in about 1hr of work without any tests or even running the app. The first time that I ran it, I got this:

http://static-pagesmc.s3.am...

I think that doing that doing a small program like that one is very hard to do in Javascript without running the actual code even with good test coverage.

For me this is a proof on how a great compiler can help you develop quickly and help you create rock solid code.

Note that all the error handling has been taken care with Kris library. In this case, Github is down the user will get a nice error message.

Don't get me wrong, I am a big fan of TDD but in Elm you need to do less tests. Less tests == less code, which means that it makes code easier to refactor.

Re: [Elm 011.5] Elm Weekly Drip #11 and Exercise: Support PresenceDiff Types

martinos — Mon, 19 Dec 2016 09:28:47 -0000

Sorry, I didn't realize that.

Re: [Elm 011.5] Elm Weekly Drip #11 and Exercise: Support PresenceDiff Types

martinos — Sun, 18 Dec 2016 10:15:50 -0000

Just to let you know, there are some episodes that are marked as "Free" but when I click on the Episode there is no video.

Thanks

Re: Episode 277: Metaprogramming Model Validation Tests

martinos — Mon, 14 Nov 2016 15:32:18 -0000

I didn't even know that this was possible, maybe it works because the list comprehension is a macro itself. I find this very interesting but it is pretty hard to reason about, personally I would prefer something like:

test "validates required fields are present", %{fields: fields} do
for {field, message} <- fields do
changeset = build(:user) |> User.changeset(%{field => nil})
refute changeset.valid?
assert changeset.errors == [{field, {message, []}}]
end
end

Re: 249 - Webpack, Phoenix, and Elm | Elixir Sips

martinos — Mon, 13 Jun 2016 15:29:49 -0000

Thanks, this is a pretty good reason. By the way, I enjoy your elm-phoenix series a lot. I really think that the elm-phoenix combination has a lot of potential. Thanks again.

Re: 249 - Webpack, Phoenix, and Elm | Elixir Sips

martinos — Mon, 13 Jun 2016 15:17:44 -0000

I am wondering why do you prefer Webpack over brunch for Elm development ?

Re: Getting started with Guardian › Stuff 'n Things

martinos — Sat, 23 Apr 2016 13:11:09 -0000

What is the suggested way to create the secret_key?

Re: Rosemont met fin au stationnement à 5 mètres d’une intersection

martinos — Thu, 24 Sep 2015 12:17:36 -0000

Un compromis intéressant serait de ne pas permettre le stationnement au coin des sorties de rues à sens uniques mais sur les rue transversales. En tant qu'automobiliste et cycliste. il me semble que c'est à la sortie des rues et non à l'entrée qu'il y a un problème. Mais bon, je n'ai pas toutes les données pour porter un jugement éclairé, si au moins ce genre de données étaient disponibles aux citoyens.

J'ai l'impression que Projet Montréal n'est pas le genre à accepter les compromis.

Re: Rosemont met fin au stationnement à 5 mètres d’une intersection

martinos — Wed, 23 Sep 2015 15:37:00 -0000

Merci Yannick

Re: Rosemont met fin au stationnement à 5 mètres d’une intersection

martinos — Wed, 23 Sep 2015 15:30:35 -0000

Je réponds à ma propre question: article 386.

Re: Rosemont met fin au stationnement à 5 mètres d’une intersection

martinos — Wed, 23 Sep 2015 14:55:35 -0000

Je n'ai jamais entendu parlé de cette loi. Quel article?
http://www2.publicationsduq...

Re: 110 Catch and Throw | RubyTapas

martinos — Wed, 26 Jun 2013 09:23:46 -0000

I think that catch and throw are a nice feature of Ruby, but since they are not very common it debugging I bit more tricky. Last week, I tried to debug a controller action that was not doing what it was expected to. I step into the code and I realized that the method was returning before the end.

I said to myself, it's probably an exception raised, so I added a rescue Exception and nothing got catch. I was really clueless for many minutes.

I knew that throw and catch existed, I wanted to know who threw the object, so I tried to find a catch all for the throw method and I have found nothing on the Web.

I think that without a catch all method that allows us to see the callstack, which helps us to find who has thrown the object, makes debugging very difficult.

Do you know how to do a catch all?

Re: Simple way to test IO in Ruby

martinos — Mon, 28 Jan 2013 18:00:48 -0000

Another way to do the same thing:
https://gist.github.com/466...

local_io("line1\nline2\nline3"){CLI.run}

Re: DOCTYPE - Episode 2 - Charts, Graphs, and Diagrams!

martinos — Wed, 01 Dec 2010 11:50:09 -0000

You should look at the Google Visualization API. http://code.google.com/apis...

Those graphs are greats because they are interactive since they are made of Javascript.

Re: git ready » interactive adding

martinos — Thu, 25 Nov 2010 10:02:19 -0000

Aleks,
git diff --cached is the diff between the index and the HEAD, not the working dir and the index.

Also I don't want to leave the interactive console.