<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Kapten on Medium]]></title>
        <description><![CDATA[Stories by Kapten on Medium]]></description>
        <link>https://medium.com/@kapten?source=rss-c3269b83a48d------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*0Z884JkeKe-apkyn4ruNew.jpeg</url>
            <title>Stories by Kapten on Medium</title>
            <link>https://medium.com/@kapten?source=rss-c3269b83a48d------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 08:15:03 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@kapten/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Kapten @ KotlinConf 2019]]></title>
            <link>https://medium.com/kapten-engineering/kapten-kotlinconf-2019-9dee877b8541?source=rss-c3269b83a48d------2</link>
            <guid isPermaLink="false">https://medium.com/p/9dee877b8541</guid>
            <category><![CDATA[kapten]]></category>
            <category><![CDATA[android]]></category>
            <category><![CDATA[kotlinconf]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[kotlin]]></category>
            <dc:creator><![CDATA[Kapten]]></dc:creator>
            <pubDate>Thu, 30 Jan 2020 15:23:39 GMT</pubDate>
            <atom:updated>2020-01-30T15:23:39.941Z</atom:updated>
            <content:encoded><![CDATA[<p>Last month, we thrilled to attend the biggest conference about Kotlin: KotlinConf! It took place in Copenhagen, Denmark, and we were lucky to be this many Kapten developers going: Android and back-end developers already working with Kotlin.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6GbZH1XU6I05gjLvPUnw0A.jpeg" /><figcaption>As mentioned, we were <strong>really happy</strong> to attend the conference!</figcaption></figure><p>In this article, we would like to highlight some of the talks we appreciated and why.</p><p>But before digging in details into the talks, we would like to give a big shout out to the speakers! About that, we were thrilled to see that 25% of the speakers were women (18 out of 70), which is great and should actually be pretty normal. Not only that, but we could also see lots of women developers, advocates, and/or simply working in the tech area.<br>Our hope of seeing more and more women speakers is in the same spirit as the one we’re having for our growing company at Kapten: continue bringing more diversity in teams to build a product with and for everyone.</p><h3>Opening Keynote</h3><p><a href="https://www.youtube.com/watch?v=0xKTM0A8gdI"><em>🔗</em></a><em> Andrey Breslav, Lead Language Designer of Kotlin @ JetBrains</em></p><p>Let&#39;s begin with the Keynote! Each excellent keynote starts with numbers, and that how <a href="https://twitter.com/abreslav">Andrey Breslav</a> opened his talk by congratulating the 4 million developers having used Kotlin this year. This incredible adoption was the opportunity to show JetBrains’ vision for his language. This vision could be shortened in one expression: <strong>one language to rule them all</strong>. Indeed, JetBrains’ dream is to have Kotlin as the default language to solve problems, creating a language available on any platform, at any scale, and accessible to everybody. And we have to say that they have done an impressive work until now to achieve this goal: having Kotlin available for server, mobile, desktop, browser, and even data science. All these topics were covered with very fascinating talks this year, you can easily find replays online if you are interested in one of them.</p><p>Aside from multiplatform, JetBrains worked hard to stabilize and improve the language in version 1.4. Code completion and Gradle imports are much faster. To really embrace multiplatform developments (and also to fix unresolvable bugs and performance issues), they started working on a new compiler. Not all of these compiler features will be available in 1.4, but Andrey Breslav shows some of the expected benefits, and it will definitely improve our life as a developer. One last announcement which will be a game-changer for Kotlin mobile is the ability to debug, test, and run iOS apps from Android Studio. It will unquestionably help reduce friction when adopting Kotlin on iOS ; thanks for this excellent initiative.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9EU1T8YTpT3XGiu0sRDwxw.jpeg" /><figcaption>Multiplatform announcement: run and debug iOS Kotlin apps in Android Studio.</figcaption></figure><p>A big announcement on JetBrains products side: <a href="https://www.youtube.com/watch?v=t1vMUV9jYRs">Space!</a> It’s an Integrated Team Environment, which seems promising! The interesting part is that the product (web, back-end, apps) is mostly written in Kotlin by sharing a lot of business logic.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*BwmNrDE4P-Gz7VSn.png" /><figcaption>Space, the new integrated team environment product announced by JetBrains during the conference.</figcaption></figure><h3>Asynchronous Data Streams with Kotlin Flow</h3><p><a href="https://www.youtube.com/watch?v=tYcqn48SMT8"><em>🔗</em></a><em> Roman Elizarov, Team Lead for Kotlin libraries @JetBrains</em></p><p>Roman Elizarov is well known for giving speeches at the KotlinConf about shiny libraries of Kotlin (does anyone remember his talks about coroutines?). And this year, he did it again with <a href="https://kotlinlang.org/docs/reference/coroutines/flow.html">Flow</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Pv9F8C6WUgZXL28du0p1iw.png" /><figcaption>An exemple of Flow sequences</figcaption></figure><p>In his presentation, he first reminded us about how coroutines and channels work. Coroutines allow us to do asynchronous work and channels to make them non-sequential. The problem is that channels are hot, so as soon as you register to them, you’ll receive some data. But if you are not ready to process the data, you’ll lose it. That’s why they created Flow, which is a new addition to the coroutine library. Flows are cold streams that are an answer to the problem brought by channels. It’s also declarative similar to Kotlin sequences. Flow handles reactive streams and presents itself as an alternative to reactive libraries such as Reactor or RxJava. What we found very interesting is that it’s interoperable with those libraries. Roman showed some numbers to prove that performance-wise Flow was better than them.</p><p>Also, he explained how easy it was to specify in which context you wanted your Flow to run on (UI, I/O, …) and how easy you could release a Flow resources. Flow will be stable in the coroutines 1.3.0 version, and some features that are missing will be released during 2020, such as sharing or caching a Flow.</p><p>We&#39;re really looking forward to test it!</p><h3>Kotlin runs taxes in Norway</h3><p><a href="https://www.youtube.com/watch?v=K8XxaAba65g"><em>🔗</em></a><em> Anders Mikkelsen and Jarle Hansen, consultant developers @ Systek</em></p><p>In this full talk, Anders Mikkelsen and Jarle Hansen, developers at Systek and consultants at the Norwegian Tax Administration (NTA), covered a lot of subjects such as: their experience with Kotlin, with Spring Boot and dug in some use cases. We especially loved their honest point of view on what worked and what they had to work around with the technos they used. One of their major pain points was with the usage of coroutines even if Spring 5.2 and Spring Boot 2.2 brought better support and solved most of the issues they encountered.</p><p>They mentioned Kotlin-Gradle, Domain-specific languages (DSLs) usage in the NTA, and the <a href="https://skatteetaten.github.io/">Aurora Platform as a Service they open-sourced</a>, and that runs part of the NTA’s digital tax and customs services.</p><p>They also shared their usage of <a href="https://docs.spring.io/spring/docs/current/spring-framework-reference/web-reactive.html#webflux">WebFlux</a>, the reactive-stack web framework from Spring. Data is represented as two primary types (Mono and Flux) of subscribable data. They showed some examples of codes with and without the usage of coroutines that makes things easier and more readable.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ejrtrdqRBY14Dd7c" /><figcaption>Reactive programming with WebFlux</figcaption></figure><p>Unfortunately, they encountered some issues with the coroutines, sometimes not supported by all used libraries.</p><p>A problem they had was with Spring Security specifically but a general issue with a lot of Spring Boot packages that are not WebFlux-specific libraries. As an example, they took the <em>@PreAuthorize</em> annotation that they could have used with the Spring Web MVC to authorize based on granted authorities and explain why it is quite impossible to make it work with suspended flows, even these are things you can workaround.</p><p>A great talk to see the up and downsides of refactoring with coroutines and reactive programming.</p><h3>LONA: Scaling server-driven UI</h3><p><a href="https://www.youtube.com/watch?v=Ir8lq4rSyyc"><em>🔗</em></a><em> Laura Kelly &amp; Nathanael Silverman, Design Language Team @ Airbnb</em></p><p>This talk presents an ambitious framework, called <a href="https://github.com/airbnb/Lona">LONA</a>, that the Airbnb team has been developing for <em>Component-based server-driven UI</em>. The main idea is to find a middle ground between the two usual “extreme” solutions:</p><ul><li>a native UI fetching raw data from a REST API</li><li>a WebView displaying HTML-based UI fetched from server</li></ul><p>With LONA, the server is providing a list of components, along with their configuration, and it is up to the native app to display these components natively. This approach allows us to leverage the benefits of the server-side flexibility and the advantages of the client-side native UI.</p><p>But the main challenge that comes with it is, of course, the versioning management because the components specification is ever-evolving and because users are using many different versions of the applications.</p><p>Airbnb team tackled this problem with Kotlin DSL (<em>Domain Specific Language</em>, not to be confused with <em>Design Language System</em>!). Once the component specification is described in Kotlin DSL (including the complete versioning), the LONA framework allows to generate a full set of tools for the developers :</p><ul><li><strong>JSON specs</strong> for the front-ends</li><li><strong>JSON schemas</strong> for produced JSON validation</li><li>Human readable <strong>Documentation</strong></li><li><strong>Page Builder</strong> to build pages with type-safety (work in progress)</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*lKSCdl7iUmDaJ35J" /></figure><p>A key point of this framework is that the LONA Spec DSL is a <em>single source of truth</em> for the full specification of the system. All the other artefacts are generated from the Spec DSL!</p><h3>Note on Multi-Platform Programming evolution</h3><p><a href="https://youtu.be/oBv1QykLAXc"><em>🔗</em></a><em> MPP in 1.3.X and Beyond by Dmitry Savvinov &amp; Liliia Abdulina</em><br><a href="https://youtu.be/IKYsX6nBcsw"><em>🔗</em></a><em> Kotlin Multiplatform in Action by Alexandr Pogrebnyak</em></p><p>Last year we were very excited about MPP; it was the beginning of exciting projects (like SQLDelight migration to being a fully MPP library). And with Kotlin/Native released in beta, it was made easier to create an iOS app with Kotlin. Today, MPP is more and more mature while being used by more companies seeing it as a true option for production products. This was proven by a lot of talks about this subject this year at the KotlinConf, but also by JetBrains supporting iOS development in Kotlin by releasing a nice update of Android Studio in 2020. Indeed, we will be able to run and debug iOS apps developed in Kotlin in Android Studio!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ivKKJQ5ZwIawRXB2Z-HbEg.jpeg" /><figcaption>Shared Common between platforms.</figcaption></figure><p>This was a fabulous opportunity to attend the KotlinConf, and to watch so many inspiring talks. We want to implement as many best practices and tips as possible in our stack, and continue using more Kotlin, especially on the back-end side, but also on our Android apps as our usage of Kotlin is currently of 92% on the driver app and 72% on the user app (and growing!).</p><p>We hope that we&#39;ll be back again next year in another beautiful city!</p><p><em>P.S.: If you’re excited about working on some of these subjects or just interested in tackling European urban mobility challenges on an ever-growing platform, we have various openings! Check it out: </em><a href="https://www.kapten.com/uk/careers/engineering/"><em>kapten.com/careers/engineering/</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9dee877b8541" width="1" height="1" alt=""><hr><p><a href="https://medium.com/kapten-engineering/kapten-kotlinconf-2019-9dee877b8541">Kapten @ KotlinConf 2019</a> was originally published in <a href="https://medium.com/kapten-engineering">Kapten Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Kapten @ DotJS 2019]]></title>
            <link>https://medium.com/kapten-engineering/kapten-dotjs-2019-817d99cf64ea?source=rss-c3269b83a48d------2</link>
            <guid isPermaLink="false">https://medium.com/p/817d99cf64ea</guid>
            <category><![CDATA[web-development]]></category>
            <category><![CDATA[kapten]]></category>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[dotjs]]></category>
            <dc:creator><![CDATA[Kapten]]></dc:creator>
            <pubDate>Thu, 09 Jan 2020 11:24:55 GMT</pubDate>
            <atom:updated>2020-01-09T11:24:55.551Z</atom:updated>
            <content:encoded><![CDATA[<p>Last December - we once again had the great opportunity to attend the DotJS in Paris, two days dedicated to the current state of front-end and back-end development in Javascript, gathering people from around the world with top-notch speakers.</p><p>Here is an overview of what we consider to be the main topics that came out of those two days.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YvSAWHA-7rxEEFun1UUnqQ.jpeg" /><figcaption><a href="https://twitter.com/_tehem">Thomas</a> presenting how at Kapten, we build ready for production microservices in 10 minutes</figcaption></figure><h3>State of Components — Vue.JS</h3><p><strong><em>by </em></strong><a href="https://medium.com/u/4f198f5f1f12"><strong><em>Evan You</em></strong></a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*AvzLPgaMwWbpxWE-GSMUWg.jpeg" /><figcaption>Evan You — DotJS 2019</figcaption></figure><p><a href="https://twitter.com/youyuxi">Evan You</a> — creator and lead maintainer of <a href="https://vuejs.org/">Vue.JS</a> — raised the point in his talk that we see a new trend to define UI component — something different from what we were used to — which was a recurring subject during the conference. Not so long ago, most of the frontend frameworks were using classes to define a new component, but as Evan pointed out during his talk, «Classes are not bad… But maybe they are not the most natural fit for UI components either». He highlighted that inheritance does not work well for UI components, mixins have their fair share of problem, and that services classes are isolated from the component API.</p><p>Now we will rather extract stateful logic from the main component and encapsulate it into “composition functions.” These functions can be tested independently and reused across components.</p><p>Evan also gave his pros and cons between the different frameworks that follow this new way (<em>React with his Hooks — Svelte 3 and more recently Vue¹ with his new composition API)</em>.</p><h3>Evan You on Twitter</h3><p>Reactivity with different tradeoffs</p><p>Another topic that came out a lot during the talks was that TS is used widely for front-end technologies. The codebase of Vue 3 is written in Typescript, and Evan You mentioned that using it internally does not affect the user but helps him understand the library a lot better. Adding an excellent support for auto-completion and type information from most IDEs — it proves to be the go-to language for front-end developers.</p><p>Typescript comes from Microsoft; if you go on its Github repository, you can see many starter templates for libraries and frameworks such as Vue, React, Node, React-native, to name of few. It also happens to be the primary language for Angular application development.</p><h3>Denoland</h3><p><strong><em>by </em></strong><a href="https://medium.com/u/affe28bd6191"><strong><em>Bert Belder</em></strong></a></p><p><a href="https://twitter.com/piscisaureus">Bert Belder </a>— co-creator and NodeJS core developer — gave a fascinating talk on a new Javascript runtime called <a href="https://github.com/denoland/deno">Deno</a>, that he developed among others, with Ryan Dahl — co-creator of NodeJS.</p><p>Node now being a matured technology, Bert and Ryan looked back at how it was created and developed. They realized that its current implementation and architecture had some limitations and would need improvements. Having that in mind, they decided to work on this new runtime with the aim of creating a real alternative to NodeJS.</p><p>When it comes to performances, Bert explained that Node implemented promises to replace callbacks, but as they were removed not long after, it resulted in hard maintenance of the code. The GYP build system used to implement C/C++ add-ons was also a pain point, as it was hard to use and maintain. That is why Deno’s runtime is based on Rust and not C/C++, so it would not use GYP anymore.</p><p>It will implement <a href="https://tokio.rs/">Tokio</a> — an equivalent of <em>libuv</em> for Rust — to handle the event-loop workflow.</p><p>Another interesting key difference is how dependencies are handled. Every Node project has a <em>package.json</em> file, a package manager (<em>npm</em>|<em>yarn</em>), and <em>node_modules</em> with tons of files, resulting in a hell of module algorithm resolution. This makes Node highly bounded to <em>npm, </em>whereas there is no need for it to be. That is where Deno comes in with a major difference; it uses ES Modules to import dependencies. No need for a <em>package.json</em> file anymore. You can now import any module directly using a URL specifying where the module is stored.</p><p>This difference raises questions about resiliency and availability. Deno has an integrated cache system storing every module that you depend on. If we take a situation where the server handling dependencies is down, you would still be able to install your dependencies from this cache.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/2998fe356a091ab83ec9b640c223c0fb/href">https://medium.com/media/2998fe356a091ab83ec9b640c223c0fb/href</a></iframe><p>On a security side, one of the main issues with <em>Node</em> is that it has access to everything outside its “sandbox.” Every module or your code can access files, networks, or environments. You can think about Deno as a black box with absolutely no doors from the outside world, which makes it really secure by default. If you need to get permission, Deno comes in with CLI flags. That way, you are enabling access to any external resource you need for your project.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/e523d054dc5a9b01ce2fabd284509a17/href">https://medium.com/media/e523d054dc5a9b01ce2fabd284509a17/href</a></iframe><p>Deno, like <em>Node</em>, is built on top of V8, supports <em>Typescript</em> natively, and was initially written in <em>Go</em> but has been rewritten in <em>Rust</em>.</p><p>Two other talks also mentioned<em> Rust</em>, both on <em>WASM</em>, one by <a href="https://medium.com/u/9092f8979c9a">Sven Sauleau</a>, working at Cloudflare and core maintainer of Babel, and the other by <a href="https://medium.com/u/1c23c4e808b8">Vlad Filippov</a>, working at Mozilla. It promises excellent performances and a more straightforward implementation. It is interesting to see how years after years <em>Rust</em> is taking more and more space in the web development universe.</p><p>However, please don’t rush into Deno yet, it still has many bugs and a beta should come out beginning of this year.</p><h3>Matthew Cawley on Twitter</h3><p>Bert Belder (@piscisaureus) talking about Deno (@deno_land) at @dotJS 2019, #dotjs</p><p>At @Kapten, we love to give a try to new exciting technologies with strong potentials. We recently used <a href="https://nestjs.com/">Nest.js</a> — a new backend framework — on a proof of concept, and we think Deno could be the next one when it is more mature. As a reminder, Deno doesn’t want to be a substitute for <em>Node</em> but an alternative!</p><h3>JS saves the world, one line at a time</h3><p><strong><em>by </em></strong><a href="https://medium.com/u/3b15b3b516eb"><strong><em>Asim Hussain</em></strong></a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*6YLIKWmNmX2YdNGe" /><figcaption>Asim Hussain — DotJS 2019</figcaption></figure><p>Can we reduce our impact on the environment to fight climate change?</p><p><a href="https://twitter.com/jawache">Asim Hussain</a>’s talk raises the point on what does this question mean for the tech industry. How — as developers — can we make our applications greener and help fighting the climate emergency we currently find ourselves in.</p><p>Electricity is responsible for 30% of the greenhouse gas pollution, and on average, 80% of the world’s energy is produced from the burning of fossil fuels, so why not try making our apps use less electricity?</p><p>When it comes to carbon costs related to using a web app, servers are on top of the list. They run 24/7. Most of them need constant temperature to operate. And cooling can compromise a considerable amount of electricity usage in data centers.</p><p>Giving some figures, such as 470kg CO2e per year — the average waste of a single idle server per year — Asim made his point on the energy consumption impact of data centers.</p><p>As more and more web applications are running on the cloud, cloud providers have an essential role to play in this fight. Asim shared some interesting points on the 3 leading players that are Google Cloud, Microsoft Azure, and Amazon Web Services.</p><p>Interesting was the fact that GCloud and MS Azure are carbon neutral companies, as well as their data centers. As for AWS, it turns out that it is, globally, not carbon neutral, with the exceptions of the Dublin, Frankfurt, Oregon, and Canada regions. GCloud is 100% relying on renewable energy to run its data centers. MS Azure will reach 60% next year. AWS wasn’t mentioned, but after some research, it seems that they are on similar levels as Microsoft.</p><p>One of the main points of his talk was to consider serverless more, in particular, through the use of service functions, such as AWS Lambda, Google Functions, and Azure Functions.</p><p>The underlying point is that cloud providers are experts in optimizing server resources to be the most efficient possible, which in the end has a massive impact on energy consumption. So in relying more on their services, we will be able to run our apps in a greener way by avoiding wasting unnecessary computing resources, which could be used more efficiently.</p><p>To conclude his talk, Asim introduced us to <a href="https://climateaction.tech/">climateaction.tech</a>, a global community of tech professionals with a mission to empower people working in the tech industry to join forces by meeting, discussing, and taking climate change actions. Ready to be part of the movement?</p><h3>What TC39 is doing?</h3><p><em>by </em><a href="https://medium.com/u/ed0162ffe9b7"><em>Daniel Ehrenberg</em></a></p><p>The closing talk was by <a href="https://twitter.com/littledan">Daniel Ehrenberg</a>, former Google now working at Igalia — a free software cooperative — who has been working on V8 and web standards, he is a TC39 core member.</p><p>Daniel presented TC39 organization workflow and culture and how its committee manages it. It was interesting to hear more about the behind the scene on how its members — who are from different parts of the world — get together to define new versions of the language.</p><p>He also presented the 4 different stages a feature has to go through to be part of the language new version, a process that can take months up to years, with a strong focus on community decisions.</p><p>This talk was also the occasion to give a snippet of interesting features proposals that currently are in the pipe at different stages.</p><p>Here is our selection:</p><p><strong><em>Private methods and fields</em></strong></p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/cd32e828938e605f16e59ca05fc705bb/href">https://medium.com/media/cd32e828938e605f16e59ca05fc705bb/href</a></iframe><p>This proposal is about something JS lacks, which is addressing the fact that while classes have been available since ES2015, they do not handle privacy on fields and methods.</p><p>Daniel explained that there were some disagreements within the committee members, especially on the syntax to use. They finally adopted the tag operator <em>#</em> as a way to specify privacy within a class. Currently, in stage 3!</p><p><strong><em>Pipelines</em></strong></p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/6ac2affbf0cf9f29fecc43c7099d2fe5/href">https://medium.com/media/6ac2affbf0cf9f29fecc43c7099d2fe5/href</a></iframe><p>Functional programming feature inspired by languages such as <em>OCaml</em> and <em>Elm</em>, is-as described on its Github page — “a backward-compatible way of streamlining chained function calls in a more functional manner.”</p><p>It is in stage 1 — so under construction — as the syntax is unsettled with two competing proposals under consideration.</p><p><strong><em>Temporal</em></strong></p><p>This proposal was presented in another interesting talk by <a href="https://twitter.com/maggiepint">Maggie Johnson Pint</a> — also a TC39 core member.</p><p>It aims to provide standard objects and functions for working with dates and times.</p><p>Basically, it would be a new implementation of the Date class. Some key differences would be that it would use a domain model — something Date lacks — and objects would be immutable.</p><p>The focus is to make the use of dates and times less painful than it is at the moment (no pun intended 😉).</p><p>You can find all the implementation details <a href="https://github.com/tc39/proposal-temporal">here</a>. It is currently in stage 2 and is being reviewed to go to stage 3!</p><p>To conclude, Daniel discussed how the organization and people from the committee would like to grow TC39 community to make it more inclusive, open, and accessible to people wanting to participate in the Javascript evolution.</p><p>This dotJS was again a huge success — from the top-level organization to the fantastic keynotes — and confirms years after years to be the JS event not to be missed.</p><p>We cannot wait to be back next year!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1LzN07hICA78DilqxBl0gg.jpeg" /><figcaption>Credits: Nicolas Ravelli for dotConferences</figcaption></figure><blockquote>P.S.: If you’re excited about working on some of these subjects or just interested in tackling European urban mobility challenges on an evergrowing platform, we have various openings! Check it out: <a href="https://www.kapten.com/uk/careers/engineering/">https://www.kapten.com/uk/careers/engineering/</a>.</blockquote><p><em>Article written by Amélie Tamisier, Rémy Gouello, and Florent Violin.</em></p><p>¹<em> </em>Official release planned in Q1 2020</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=817d99cf64ea" width="1" height="1" alt=""><hr><p><a href="https://medium.com/kapten-engineering/kapten-dotjs-2019-817d99cf64ea">Kapten @ DotJS 2019</a> was originally published in <a href="https://medium.com/kapten-engineering">Kapten Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[We’re opening a Tech Hub in Barcelona!]]></title>
            <link>https://medium.com/kapten-engineering/were-opening-a-tech-hub-in-barcelona-d067ab0af920?source=rss-c3269b83a48d------2</link>
            <guid isPermaLink="false">https://medium.com/p/d067ab0af920</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[technewsblog]]></category>
            <category><![CDATA[coding]]></category>
            <category><![CDATA[technews]]></category>
            <dc:creator><![CDATA[Kapten]]></dc:creator>
            <pubDate>Wed, 17 Jul 2019 11:56:20 GMT</pubDate>
            <atom:updated>2020-01-09T10:59:35.392Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>We’re opening a Tech Hub in Barcelona and here’s why!</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wJDXWwm9h1wYDgpJT1vRuA.jpeg" /></figure><p><em>Our new space is an earthly paradise for developers — even if we say so ourselves.</em></p><p><strong>We’re super excited to tell you about our awesome new Tech Hub in Barcelona, one of the world’s most buzzing cities.</strong></p><p>As one of the most dynamic tech companies in Europe, we want to attract world-class tech talent to make our driver apps as awesome as possible for our driver partners. And that was a big part of our decision to base our new Tech Hub in Barcelona — one of Europe’s tech capitals.</p><p>When we started as Chauffeur Privé in 2012, we never guessed that in just a few years we’d have over 300 people across 17 nationalities working with us, or that we’d have 2 million happy customers using our app to make journeys with 25,000 driver partners in 8 European cities. It’s been a real trip!</p><p>We felt the time was right to create an environment that was all about the tech. A new space where devs can roam free and unleash their brilliant talents to the fullest, just like our other workplaces. Number one: because we’re hungry to expand internationally and double the size of our tech team since rebranding as Kapten earlier this year. Number two: because we have many exciting challenges around the corner. And to meet them, we need more devs!</p><p><strong><em>Our next-level workplace has been designed to give everyone the sense of freedom and fun they need to produce great ideas that really pop.</em></strong></p><p>Launching on 1st September with a crack team of engineers who will hopefully reach 50-strong by the end of the year, our new hub will be a dynamic working environment like no other. It’s about creating lots of cool features and having a really positive impact on people and society in general. And it’s also a great chance for developers at all levels to learn and develop their skills.</p><p>As well as diving into a pool of great talent, our new team will also work in a Spotify-like way that organises groups into Tribes, Squads and Guilds. Think 2-week Scrum sprints and the kind of enthusiastic energy that gets new features pushed live at speed.</p><p>We’re looking for engineers who can’t wait to get stuck into building the best possible app features from day one. Like a quick and easy driver onboarding process, reporting tools and bonus systems for our driver partners, as well as optimising and refining our real-time GPS platform to match passengers and drivers. And that’s just the start!</p><p><strong>If you’re a developer with big ambitions and an even bigger skillset, you’ll find a perfect home with us in Barcelona. </strong>We live by a ‘healthy body, healthy mind’ culture which fits right in with our location in the heart of Barcelona. There are plenty of perks and benefits for working with us (hello, free taxi rides) but most of all, it’s the sense that you’re making something special, in one of the most beautiful cities in the world, alongside some truly, awe-inspiringly talented people.</p><p>Does that sound like you? Drop us a line! <a href="https://www.kapten.com/uk/careers/engineering/">https://www.kapten.com/uk/careers/engineering/</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*H89-w1TwFKSwoYUtmL-nIQ.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d067ab0af920" width="1" height="1" alt=""><hr><p><a href="https://medium.com/kapten-engineering/were-opening-a-tech-hub-in-barcelona-d067ab0af920">We’re opening a Tech Hub in Barcelona!</a> was originally published in <a href="https://medium.com/kapten-engineering">Kapten Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Kapten @ ReactEurope]]></title>
            <link>https://medium.com/kapten-engineering/kapten-reacteurope-c2004ab9ac4f?source=rss-c3269b83a48d------2</link>
            <guid isPermaLink="false">https://medium.com/p/c2004ab9ac4f</guid>
            <category><![CDATA[react]]></category>
            <category><![CDATA[javascript]]></category>
            <dc:creator><![CDATA[Kapten]]></dc:creator>
            <pubDate>Mon, 24 Jun 2019 09:05:31 GMT</pubDate>
            <atom:updated>2020-01-09T10:59:09.999Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dSt17jOwXL6Jb1Csk6OfOw.jpeg" /><figcaption>React Europe 2019</figcaption></figure><p>This year, ReactEurope took place in Paris the 23rd and 24th of May and addressed a lot of topics around front-end development with React.</p><p>So here is a selection of conferences that we thought are worth talking about.</p><h3>The Keynote</h3><p><em>by </em><a href="https://medium.com/u/fb7a3c353cc1"><em>Jared Palmer</em></a></p><p><a href="https://medium.com/u/fb7a3c353cc1">Jared Palmer</a> is a lead Engineer at The Palmer Group<strong> </strong>and the creator of <strong>Formik</strong> (React form builder), <strong>Razzle</strong> (create universal JS application with no configuration) and <strong>react-fns</strong> (Web APIs turned in React components).</p><p>He started to give us the React 2019 Roadmap; <a href="https://reactjs.org/blog/2018/11/27/react-16-roadmap.html#react-16x-mid-2019-the-one-with-suspense-for-data-fetching"><strong>Suspense for data fetching</strong></a><strong> </strong>and a new server renderer. Then, we went deep into this first topic.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3_dOgdyUQWJGRUVPFKqLXA.jpeg" /><figcaption>Benefits of Suspense</figcaption></figure><p><strong>Loading stuff is still hard</strong></p><p>A lot of our applications are made to load data. And thanks to React, it’s now easier! We want our applications fast, good and pretty. But these two last things conflict with each other. React provide some APIs (Lazy load, etc.) to help us make a smoother UX. It’s good for slow connections because you want to show to your users that your application works. However, it makes your UX bad for fast connections. Your users will see too much useless loading state like flashes.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/250/0*N2VPDDS81syjl2eD" /></figure><p>That’s what React Suspense aims to fix.</p><p><strong>Why React</strong> <strong>Suspense</strong>?</p><p><strong>Suspense for data fetching</strong> refers to React’s ability to “suspend” rendering while components are waiting for data. Jared wrote a small custom hook which returns a promise if the cache hasn’t corresponding data(s).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*yUjtofdwG42uRlIf" /><figcaption><em>React aim to deal with </em><strong><em>cache</em></strong><em> but Apollo and Relay already do it.</em></figcaption></figure><p>This hooks will throw a Promise if the data requested has not been cached.<br>When you throw a promise, the Suspense component behaves like a componentDidCatch and execute this asynchronous operation (in our case, a query).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/950/0*b0QOW3hZ6iiEHAwZ" /><figcaption><em>You can add a fallback to Suspense</em></figcaption></figure><p>After some changes, it looks like this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/250/0*8fNtVrCj_2vrLIMu" /><figcaption><em>Here, Jared add a small spinner on the right to simulate the loading of the user details</em></figcaption></figure><p>As you can see, the application is waiting for the required data to be fetched in order to render the page.</p><p><em>You can access the slides and the demo app on the </em><a href="https://github.com/jaredpalmer/react-europe-2019"><em>Jared Palmer GitHub account</em></a><em>.</em></p><h3>Move fast with confidence</h3><p><em>by </em><a href="https://medium.com/u/b97642f6668f"><em>Paul Armstrong</em></a></p><p><a href="https://medium.com/u/b97642f6668f">Paul Armstrong</a> is tech lead at Twitter and with his team, they rebuilt Twitter from ground up as a React.js PWA. Here are some of the lessons that they learned, some tools and strategies to move fast with confidence from development to production.</p><h4>Fail fast locally</h4><p>Have you ever pushed something to GitHub and then CircleCI took 15 minutes to tell you that your test failed? It could be just a linting error and we should have got alert in a few minutes.</p><p>To improve your productivity, it’s better to fail faster in a local development environment. Git hook can help to set up some necessary checks before commit:</p><ul><li>Only test related files</li><li>Only lint related files</li><li>Apply auto-formatting</li><li>Type check</li></ul><p>By using <a href="http://github.com/typicode/husky">Husky</a> and <a href="http://github.com/okonet/lint-staged">lint-staged</a>, we can set up the above checks with a few lines of configuration. Many projects like Create-react-app use this kind of tools. Paul shared some examples of config files in order to get a failure alert as soon as possible.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/902/0*ZuE9he5evq4V4B_Q" /><figcaption>Husky config file example</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*vVEUml3g7sIX-wVT" /><figcaption>lint-staged config file example</figcaption></figure><h4>Automate node module installation</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*kOOf2eg7Mm7nenMA" /></figure><p>Sometimes, we forgot that required npm dependencies change when we change code branch. Well, git hook can also take charge of node modules installation and it saves time!</p><p>Husky exposes $HUSKY_GIT_PARAMS variable, so we can check that and decide whether we install the node_modules again or not. Therefore, no more manual installation process is needed.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ivnAD33vUyEf-voJ" /><figcaption>Husky config file example</figcaption></figure><h4>Use <a href="https://buildtracker.dev/">Build Tracker</a> to track build size.</h4><p>As a developer, every time we added a new dependency, we would like to know how it affects the build size. Build Tracker, developed by Paul Armstrong, aims to automate bundle size tracking. With the dashboard showing the bundle sizes of every build over time, we are able to track the size growth and find out the ones which potentially affect application loading speed.</p><p>As Paul has said, it really helps Twitter to catch some really serious issues before going to production.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ZnLHsEDO-goUKyLe" /></figure><h3>Saving the web, 16ms at a time</h3><p><em>by </em><a href="https://medium.com/u/a11edf57e20a"><em>Joshua Comeau</em></a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gK4-q_g43imMQmvVdVs8pg.jpeg" /></figure><p><a href="https://medium.com/u/a11edf57e20a">Joshua Comeau</a> is a software engineer at <strong>Khan Academy</strong> and he gave a talk about the different ways to create <strong>smooth animations</strong> in the browser.</p><p>He explained the stages the browser runs in order to render one frame among the many necessary to perform an animation.</p><p>Those stages are the following:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*atrsGSJraXzueY44" /></figure><p>The most expensive stages are the <strong>layout</strong> and the <strong>paint</strong>. The layout stage is in charge of computing the size and the position of every element in the document while the paint takes care of applying colors, rendering images, shadows, etc…</p><p>There’s obviously a way to skip these stages by avoiding using some CSS properties in the keyframe or the transition.<br>For example, in order to skip the layout stage, we can use <strong>CSS transforms and opacity</strong> instead of using width, height, margins, padding, absolute positioning, etc…<br>With that in mind, we can optimize our animations to reach the holy grail a.k.a <strong>the 60 FPS frame rate</strong>.</p><p>For more complex animations (which involves a lot of particles for example), Joshua gave us a pretty unconventional tip which is inspired by the 2D video games industry: <strong>the animation sprite</strong>.<br>The purpose here is to create an image containing each frame of the animation, display it in a wrapper having the same dimensions than one frame of the animation, and move the image step by step using CSS transform to reproduce the animation.</p><p>Here is a code sample:</p><p>The sprite:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*GW774eL5__HS-pPs" /></figure><p>The component:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*LIbGzfD515FVC8MB" /></figure><p>And the styled components:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/952/0*gALvDslVXzpmIYAf" /></figure><p>And this will give us the following result:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/148/0*9ffnDDv53rbfQpGQ" /></figure><p>In conclusion, we can paraphrase Joshua by saying “Do not be afraid to break the rules”. Which means that it’s not necessary a bad idea to do unusual things as long as it offers a real improvement in terms of performance and user experience.</p><blockquote><em>P.S.: If you’re excited about working on some of these subjects or just interested in tackling European urban mobility challenges on an evergrowing platform, we have various openings! Check it out: </em><a href="https://www.kapten.com/uk/careers/engineering/">https://www.kapten.com/uk/careers/engineering/</a><em>.</em></blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c2004ab9ac4f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/kapten-engineering/kapten-reacteurope-c2004ab9ac4f">Kapten @ ReactEurope</a> was originally published in <a href="https://medium.com/kapten-engineering">Kapten Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Kapten @Kubecon]]></title>
            <link>https://medium.com/kapten-engineering/kapten-kubecon-499f90225dfb?source=rss-c3269b83a48d------2</link>
            <guid isPermaLink="false">https://medium.com/p/499f90225dfb</guid>
            <category><![CDATA[microservices]]></category>
            <category><![CDATA[tech]]></category>
            <category><![CDATA[kubecon]]></category>
            <category><![CDATA[technology-trends]]></category>
            <category><![CDATA[technology-news]]></category>
            <dc:creator><![CDATA[Kapten]]></dc:creator>
            <pubDate>Tue, 11 Jun 2019 10:10:38 GMT</pubDate>
            <atom:updated>2020-01-09T10:59:55.023Z</atom:updated>
            <content:encoded><![CDATA[<p>From May 20th to May 24th, 1 member of our Production Team and 2 members of our Platform team went to KubeCon + CloudNativeCon Europe 2019 in Barcelona. It is the leading conference on Kubernetes and cloud-native applications.</p><p>Around 7,700 fellows came from all over the world, from different industries and background that were reunited during 4 days to discuss, learn and share their knowledge.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/900/0*N9Z53M37NxfqOAlp" /></figure><p>Kubernetes turns 5 this year, and this conference was also the occasion to take a look back and observe its technical maturity five years later.</p><p>As such, various talks were given during these 4 days and we learned a lot of things, so we were excited about talking about some of the more prominent themes permeating the conference:</p><ul><li>Kubernetes as the new standard</li><li>The Cloud native ecosystem</li><li>Service meshes and SMI</li></ul><h3>Kubernetes as new standard</h3><p>Kubernetes is becoming right now the de-facto standard that people are starting to adopt really fast.</p><p>Here one extract of <a href="https://trends.google.com/trends/explore?date=2016-01-01%202019-06-02&amp;q=docker%20swarm,apache%20mesos,kubernetes">Google Trends</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*-7gWfIQEF0OZld0y" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*JekIoH3RklTpk6p-" /></figure><p>During those 4 days, we had the chance to listen to successful and failure stories on migrating to k8s, how companies had to adapt their organizations in order to facilitate the transition to cloud-native and many in-depth technical talks from various companies, from banks to retail, all with their own specificities.</p><p>Thanks to the wide community behind Kubernetes and to the large adoption by the industries, Kubernetes is nowadays a production ready technology that can give companies more flexibility and control over their infrastructure.</p><p>This is why we at Kapten, decided to migrate all of our infrastructures to Kubernetes, a project we achieved mid-April 2019. At Kapten from our code crafting pipeline up to production infrastructure, all the steps are massively based on Kubernetes.</p><h3>The CNCF ecosystem</h3><p>Kubernetes and its related technologies were at the center of this conference, but the CNCF (Cloud Native Computing Foundation) has a broader goal in mind: helping developers navigating a cloud-native landscape, by building an ecosystem that answers various needs, such as:</p><p>- Distributed environments communication with NATS, a go messaging queue that focuses on simplicity &amp; performance, which is releasing its 2.0 in June.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*WSogWBinXXFkwYAk" /><figcaption>NATS Deep Dive talk</figcaption></figure><p>- Unified logging layer with Fluentd — an open source data collector — which has graduated recently from incubation.</p><p>These projects are a small example of this ecosystem’s reach, with more than 690 projects!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*IsdYbXid7hIf_FlO" /><figcaption><a href="https://landscape.cncf.io/">CNCF Cloud Native Interactive Landscape</a></figcaption></figure><p>At Kapten, to keep up with the growth of our microservice architecture, we’ve been working on one of the conference trending topics: distributed tracing, which is one of the 3 pillars of observability, with logging and metrics: these are ways to make monitoring, debugging our distributed systems accessible.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*kxTOqrD8TNxD_n_p" /><figcaption>The pillars of observability keynote talk</figcaption></figure><p>Whereas the two former pillars focus on giving visibility from a single service at a time, distributed tracing is a method used to profile and monitor applications, especially those built using a microservices architecture. It helps pinpoint where failures occur and what causes poor performance.</p><p>But how do we implement it in our systems? Fortunately, there are CNCF incubating projects for that:</p><p>- Application code is instrumented with the OpenTracing project, which delivers vendor-neutral APIs and instrumentation for distributed tracing.</p><p>- Jaeger: a distributed tracing system collects tracing data and offers a UI for transaction monitoring, root cause analysis, service dependency analysis, and performance/latency optimization.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*j05doqLR-ASapQCr" /><figcaption>Example of Jaeger trace UI</figcaption></figure><p>Finally, as an effort to concentrate community effort &amp; ecosystem consistency, an announcement was made at the conference keynote to introduce OpenTelemetry, which is a backward compatible merger of OpenTracing and OpenCencus — another instrumentation project originating from Google.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*OeKT_5TqLge5AKzF" /><figcaption>OpenTelemetry keynote announcement</figcaption></figure><p>Another big theme of this conference was the community, as an ecosystem thrives on their efforts: growing an inclusive environment, while accompanying users to become contributors, then maintainers, remains a critical focus.</p><h3>Kapten Engineering on Twitter</h3><p>An inspiring talk on contributing to kubernetes from @TheNikhita &amp;amp; @kubernetesonarm, with diversity at its core! #kubecon #cloudnativeconf #opensource</p><h3><strong>Service mesh and SMI</strong></h3><p>During the KubeCon, many talks spoke about service mesh. A service mesh is a way to control how different parts of an application share data with one another. Unlike other systems for managing this communication, a service mesh is a dedicated infrastructure layer. The service mesh is usually implemented by providing a proxy instance, called a <em>sidecar</em>, for each service instance.</p><p>The mesh provides critical capabilities including service discovery, load balancing, encryption, observability, traceability, authentication and authorization, and support for the circuit breaker pattern.</p><p>In the Kubernetes ecosystem, there are many service mesh providers (Istio, Linkerd, aspen mesh, HasiCorp Consul, …).</p><p>The problem is that developers who turn to mesh technologies must choose a provider and write directly to those APIs. They become locked into a service mesh implementation. That is the reason why during the keynote Microsoft announced Service Mesh Interface (SMI) which defines a set of common, portable APIs that provide developers with interoperability across different service mesh technologies including Istio, Linkerd, and Consul Connect. SMI is an open project started in partnership with Microsoft, Linkerd, HashiCorp, Solo.io, Kinvolk, and Weaveworks.</p><p>The idea behind Service Mesh Interface is not new. It is the line of existing Kubernetes resources like Ingress and Network Policy.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/930/0*84MEL5wSdINiORvI" /></figure><p>At Kapten, we are very excited about the idea to add a service mesh in our Kubernetes clusters. And now with SMI we should not become locked into a service mesh vendor.</p><p>You can find more information about SMI on the official website <a href="https://smi-spec.io/">here</a>.</p><p>This conference was a great opportunity to get a view of the current landscape, how it got here, but also where it’s headed: we talked about a few subjects for which we’re really excited about at Kapten, but we’re always looking into ways to improve our platform, so we’ll definitely follow the cloud-native ecosystem closely!</p><p>Fortunately, KubeCon + CloudNativeCon will be returning next year in Amsterdam, from March 30th to April 2nd, and we’ll definitely try to be there!</p><p>What talks/subjects really stood out to you? Where do you think this ecosystem is going?</p><blockquote>P.S.: If you’re excited about working on some of these subjects or just interested in tackling European urban mobility challenges on an evergrowing platform, we have various openings at <a href="https://www.kapten.com/uk/careers/engineering/">https://www.kapten.com/uk/careers/engineering/</a>.</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=499f90225dfb" width="1" height="1" alt=""><hr><p><a href="https://medium.com/kapten-engineering/kapten-kubecon-499f90225dfb">Kapten @Kubecon</a> was originally published in <a href="https://medium.com/kapten-engineering">Kapten Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Kapten @ International Javascript Conference 2019]]></title>
            <link>https://medium.com/kapten-engineering/kapten-international-javascript-conference-2019-41b83d37144b?source=rss-c3269b83a48d------2</link>
            <guid isPermaLink="false">https://medium.com/p/41b83d37144b</guid>
            <category><![CDATA[coding]]></category>
            <category><![CDATA[tech]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[javascript]]></category>
            <dc:creator><![CDATA[Kapten]]></dc:creator>
            <pubDate>Wed, 05 Jun 2019 13:58:31 GMT</pubDate>
            <atom:updated>2020-01-09T11:00:13.872Z</atom:updated>
            <content:encoded><![CDATA[<p>Three weeks ago we were present at the <a href="https://javascript-conference.com/">International Javascript Conference</a>, taking place in beautiful Westminster in London.</p><p>These two days of conferences were quite intensive but really fun and exciting! First because as French we always like to meet with our British cousins, but also because the talks were engaging and full of insights. We also had the opportunity to talk with many participants at our booth to discuss experiences, best practices, and so much more…</p><p>This short trip in London was also the opportunity for us to test the Kapten service before the official launch that took place on May 16th! 🚀</p><p><strong>In this article, you’ll find the main takeaways we observed during these two days.</strong></p><h3>Performance Budgets: the What, Why and How</h3><p><em>by </em><a href="https://medium.com/u/c7829fed8795"><em>Marcos Iglesias</em></a></p><p><a href="https://medium.com/u/c7829fed8795">Marcos Iglesias</a> from <a href="https://medium.com/u/341ba3c1cbab">Eventbrite Engineers</a> gave us a captivating talk about Performance Budgets. By beginning his speech with “Performance matters”, he clearly wants to highlight the importance of performance with the rise of web development in today’s world.</p><p><strong>Performance budgets consist of defining the performance threshold that we do not want to exceed</strong>. Marcos showed us why performance budgets are critical from its Eventbrite experience: create a Web Performance culture in the company, improve the user experience and enhance team accountability.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*nmI4DgRKZuskjZb7" /><figcaption>“Performance Budgets: the What, Why and How” talk by <a href="https://medium.com/u/c7829fed8795">Marcos Iglesias</a> from <a href="https://medium.com/u/341ba3c1cbab">Eventbrite Engineers</a></figcaption></figure><p>But when it comes to applying the concept, Marco explained to us what to do and how to comply with performance budgets:</p><ol><li><strong>Optimize an existing feature or asset</strong></li><li><strong>Delete an existing feature or asset</strong></li><li><strong>Do not add this new feature</strong></li></ol><p>To set up a performance culture inside your company, a shared performance vision should be integrated into teams. Marco introduced four strategies that lead to a good performance budget plan:</p><ul><li><strong>Faster than your competitors</strong></li><li><strong>Optimize business value</strong></li><li><strong>Follow Guidelines</strong></li><li><strong>Better than current</strong></li></ul><p>As a result, the question that we had was: how can we monitor our performance budgets? He replied with three clear metrics:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IfEKd7U6LTpJc0idpXNufw.jpeg" /></figure><p><strong>Rules-based metrics can be performed by some famous tools for performance:</strong></p><ul><li>LightHouse</li><li>WebPageTest</li></ul><p>He advised attendees to set up custom metrics in their company. These metrics are the best, mainly because you define them yourself. The metrics should be determined <strong>when the user can interact with your core business value</strong>.</p><p>Among the most famous, we have the <em>“</em><a href="https://blog.twitter.com/engineering/en_us/a/2012/improving-performance-on-twittercom.html"><em>time to first tweet</em></a><em>”</em> from Twitter. At Kapten,<em> “</em><strong><em>time to first ride”</em></strong> would be an efficient metric!</p><p>To conclude, Performance budgets adoption in your projects will allow you to put performance back to the place it deserves, to see a feature in its own right. The implemented process will allow you to make sure that new functionality will not degrade your system performance and mostly the user experience.</p><p><a href="https://docs.google.com/presentation/d/e/2PACX-1vSVI0ABJbAhn32522bTV7KOzigrJRE5p2-bRr0HuPSW1rytofeIHXUyS4Htmd9T2JEIkwAkWbGl4OLJ/pub"><em>Here are the slides from the talk</em></a></p><h4><strong>How can you make a JavaScript Website successful in Google?</strong></h4><p><em>by </em><a href="https://medium.com/u/c77d8c36315"><em>Tomek Rudzki</em></a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xgo9yUuSbIee1s8_MEpZXA.jpeg" /></figure><p>We also had another great talk on performance in Javascript application by <a href="https://medium.com/u/c77d8c36315">Tomek Rudzki</a>, R&amp;D Specialist &amp; Technical SEO at <a href="http://twitter.com/onely">@onel</a>y. SEO is becoming more and more important with the rise of modern Javascript frameworks like React, Angular or Vue. In this talk, he explained why it’s crucial for companies to care about SEO.</p><p>Above all, <em>“Time is money”</em> concept highlights the fact that faster web apps allow companies to make more money. The speed of web applications can directly impact the revenue of the company. For example, <a href="https://www.fastcompany.com/1825005/how-one-second-could-cost-amazon-16-billion-sales">Amazon’s study</a> reveals that just one second of website load slowdown can result in 1.6 billion in sales drop each year, simply amazing!</p><p><strong>Tomek showed us how to diagnose and solve Javascript SEO issues:</strong></p><ul><li>Use the URL inspection tool to verify if Google can render your content</li><li>Ensure you don’t block any JavaScript files</li><li>Use the URL Inspection tool for fixing JavaScript errors</li><li>Check if your content has been indexed in Google</li></ul><p><em>For more details on Tomek’s talk, check out his excellent and very detailed : </em><a href="https://moz.com/blog/diagnose-and-solve-javascript-seo-issues"><em>here</em></a></p><h4>How to re-architect a JavaScript Class System</h4><p><em>by </em><a href="https://twitter.com/tyoushe?lang=fr"><em>@tyoush</em></a></p><p>Olga Petrova did a fascinating <a href="https://javascript-conference.com/web-development-architecture/how-to-re-architect-a-javascript-class-system/">talk</a> on how Sencha managed the evolution of ECMAScript versions with their internal needs to offer the most up-to-date technology for their clients. A presentation of the original strategy followed in Sencha to manage objects inheritance before its later availability in es6. Based on <em>Object.create</em> function, the object management evolved to manage many other aspects in the Class System: inheritance, config properties, dependencies, mixins, overrides, and build tool.</p><p>While adopting the new es6 standard, Sencha had to manage evolutions progressively without injecting any breaking change for these clients and introduced decorators’ patterns for features available in their previous implementation not ready yet or planned to be supported at the moment in ECMAScript like for the overrides of object methods.</p><p><em>Slides available </em><a href="https://speakerdeck.com/olgapetrova/how-to-re-architect-a-javascript-class-system"><em>here</em></a></p><h4>Visual Testing with Cypress</h4><p><em>by </em><a href="https://medium.com/u/1104a5297728"><em>Jesse Sanders</em></a></p><p>Here <a href="https://medium.com/u/1104a5297728">Jesse Sanders</a> introduced us the concept of visual testing.</p><p>This testing layer can be useful to test the pages of your web application in a pixel perfect way. This is different from unit testing that checks that a function behaves as expected and from integration testing which checks that when clicking on a button the corresponding action happens but does not verify that the button is correctly rendered.</p><p>Jesse presented two tools to perform this kind of testing:</p><ol><li><a href="https://github.com/mjhea0/cypress-visual-regression">Cypress Visual Regression</a></li><li><a href="http://percy.io/">Percy.io</a></li></ol><h4>1. Cypress Visual Regression</h4><p>This module allows you to perform actions on your app pages and to take screenshots of the result page, and then to compare the result with the reference screenshots. In the case of a difference, a new image highlighting the difference between the two is created, to spot the difference and fix it quickly.</p><p>The only downside to this solution is that you have to set up manually the whole testing environment, including configuration and automation, before being able to test your pages.</p><h4>2. Percy.io</h4><p>Percy is the ultimate SaaS solution for visual testing. It provides the same features as Cypress Visual Regression (the tool is also based on Cypress) but with more features on top of that.</p><p>The Percy interface allows to see test run reports, and in case of regression it shows the before and the after screenshot, and it provides filters to really highlight the difference between the two. If some common component present on multiple pages has changed, it is also smart enough to group all the impacted screens into a single regression, so that we can easily see what actually has changed and therefore fix it faster. These highlighted differences then either have to be fixed or we can update the reference screenshots if we intended to make visual changes.</p><p>The configuration is also way more straightforward, and especially after creating your project, you can connect it to your Github repo, and it will be integrated into the CI validators. This way, tests and reports are automatically triggered when pushing your code, and you can be sure to block merging pull requests if there is a regression somewhere.</p><h4>Serverless explained by someone not selling you anything</h4><p><em>by </em><a href="https://twitter.com/taylorkrusen?lang=fr"><em>@TaylorKrusen</em></a></p><p>Taylor Krusen did a <a href="https://javascript-conference.com/web-development-architecture/serverless-explained-by-someone-not-selling-you-anything/">high-level presentation</a> on what is the serverless concept and how to implement it for real. From the initial buzzword, we had a concrete look thanks to an amusing example: anyone in the room was able to send jokes on Twitter and these jokes were immediately ingested into a simple spreadsheet. The code has been presented then deployed by uploading a zip file on AWS Lambda.</p><p>High-level concepts have been covered and event-based architecture pattern implications have been discussed as well.</p><h4>Functional JavaScript</h4><p><em>by </em><a href="https://twitter.com/olivers"><em>@olivers</em></a></p><p>Functional programming has always been a subject of great discussion in the programming world. While some languages are specifically made to develop in a functional way (think <a href="https://www.haskell.org/">Haskell</a> for example), this programming paradigm can also be used in other languages, such as Javascript.</p><p><a href="https://javascript-conference.com/web-development-architecture/functional-javascript/">In this talk</a>, Oliver introduced us different ways to develop in a functional way in Javascript, starting with logic done by hand, and continuing with various libraries helping us to develop faster.</p><p>The first example showed us how to perform functional programming manually by introducing the currying concept and composing methods.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*MO7jhT7DqumrUwQU" /><figcaption><a href="https://github.com/oliversturm/functional-js/blob/master/samples/fpbasics/manual.js">https://github.com/oliversturm/functional-js/blob/master/samples/fpbasics/manual.js</a></figcaption></figure><p>Then Oliver introduced <a href="https://lodash.com/">lodash</a>, which is a great utilitary library. This library can really do many things, but the one thing we focused on is the ‘curry’ method, which has the same logic as shown before, but is even more modular as it can handle a variable number of arguments, and even keep placeholders for arguments that will be passed later in the logic.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*uSgXj66comA0O60d" /><figcaption><a href="https://lodash.com/docs/4.17.11#curry">https://lodash.com/docs/4.17.11#curry</a></figcaption></figure><p>The next example was still using lodash, but especially a small sub-module that is not really well known: (<a href="https://github.com/lodash/lodash/wiki/FP-Guide">lodash/fp</a>).</p><p>This module allows to compose functions as in the first manual example, and also has the same methods as the regular lodash (map, reduce,…) but with a different API.</p><p>Here, instead of applying a method directly on one variable in one place, you build your method that can then be used multiple times at multiple places on different variables.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Fh7qvvDp6FDoOdYQ" /><figcaption><a href="https://github.com/lodash/lodash/wiki/FP-Guide">https://github.com/lodash/lodash/wiki/FP-Guide</a></figcaption></figure><p>The final example was using the <a href="https://github.com/rtfeldman/seamless-immutable">seamless-immutable</a> library to manipulate immutable objects to avoid any potential side effect in a method, in a truly functional way.</p><p><em>You can find </em><a href="https://github.com/oliversturm/functional-js"><em>here</em></a><em> the presentation slides and the code samples used to show everything.</em></p><p><em>Founded in 2012, </em><strong><em>Kapten</em></strong><em> has become a major player within the French ride-hailing market. We are made up of 270 employees, 15 nationalities, more than 2 million clients, and a presence in the major European cities. We’re looking for new talents to reach our ambition: becoming the European leader in urban mobility. If you think, as we do, that there’s no reason an American should take this spot… come and meet us : </em><a href="https://www.kapten.com/uk/careers/engineering/"><em>https://www.kapten.com/uk/careers/engineering/</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=41b83d37144b" width="1" height="1" alt=""><hr><p><a href="https://medium.com/kapten-engineering/kapten-international-javascript-conference-2019-41b83d37144b">Kapten @ International Javascript Conference 2019</a> was originally published in <a href="https://medium.com/kapten-engineering">Kapten Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Kapten @dotGo]]></title>
            <link>https://medium.com/kapten-engineering/kapten-dotgo-7529fb394bf?source=rss-c3269b83a48d------2</link>
            <guid isPermaLink="false">https://medium.com/p/7529fb394bf</guid>
            <category><![CDATA[golang]]></category>
            <category><![CDATA[karten]]></category>
            <category><![CDATA[dotconferences]]></category>
            <category><![CDATA[dotgo]]></category>
            <dc:creator><![CDATA[Kapten]]></dc:creator>
            <pubDate>Mon, 08 Apr 2019 18:36:18 GMT</pubDate>
            <atom:updated>2020-01-09T11:00:38.521Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*S8wFpK6b927GH7m2" /><figcaption>“Théâtre de Paris”, full of Gophers ready to attend the talks.</figcaption></figure><p>Last week we attended the dotGo — European Go conference that took place in the beautiful and historical Théâtre de Paris, localized in the heart of Paris.</p><p>It was a day full of inspiring people and talks. We were thrilled to be part of this conference as a Silver Sponsor. Thanks to this sponsorship, we had the chance to have our own booth to present our tech stack to the dotGo attendees.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*O0_eaJ1CeX8PQTl7DPONWQ.jpeg" /><figcaption><em>The Kapten team in front of the booth, thrilled to talk with passionate Gophers</em></figcaption></figure><p>If you’ve never been to a Dot Conference, you should know that the format is quite unique: there’s no specific schedule for each talk but instead only multiple 1h30~ session with various discussions, which allows attendees to discover topics that they may have missed otherwise.</p><p>In this article, we’ll focus on speeches that impacted us the most and will probably change the way we work with Go on a daily basis.</p><h3>Go as a scripting language</h3><p>By Ignat Korchagin <a href="https://twitter.com/secumod">@secumod</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*9Kq0sce1R2zp5RO_" /><figcaption>Ignat Korchagin — System engineer at Cloudflare — gave a compelling speech about using Go as a scripting language (so without compilation) to replace Python by Go (maybe this is a troll :) )</figcaption></figure><p>Why should we use Go as a scripting language? The answer is simple: Go is easy to learn, not too verbose and there is a massive ecosystem of libraries.</p><p>On Unix system, script files begin with a <a href="https://en.wikipedia.org/wiki/Shebang_(Unix)">shebang</a> line. It starts with ‘<strong>#!</strong>’ and specifies the script interpreter to be used to execute the script (for a bash script, it starts with ‘<strong>#!/bin/bash</strong>’). In Go language, there is the <strong>run</strong> command which interprets Go files, the shebang to Go script could be something like <strong>#!/usr/bin/env go run</strong>.</p><p>But they are few issues with this approach:</p><ul><li>The shebang line is not valid in <strong>.go</strong> files because ‘<strong>#</strong>‘ is not valid at the beginning of a line</li><li><strong>go run </strong>doesn’t return the script error code back to the operating system</li></ul><p>During his presentation, Ignat showed us some tricks and workarounds to try to get round of these issues. However, all methods presented have some drawbacks.</p><p>The last solution presented by Ignat was actually the right one: using binary support module in the Linux Kernel (<strong>binfmt_misc</strong>) to specify that <strong>.go</strong> files should be executed with <strong>gorun</strong>.</p><p>You could find more informations <a href="https://blog.cloudflare.com/using-go-as-a-scripting-language-in-linux/">here</a>.</p><h3>Multi-Module Repositories</h3><p>By Jean de Klerk <a href="https://twitter.com/jadekler">@jadekler</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*1k0n4KRqCwo1zAk5" /></figure><p>Jean de Klerk is an engineer at Google specialized on Go libraries. He’s been working in Go since 2013. Now that Go modules are considered as the future of dependency management in Go, we can start learning about advance usages of Go modules.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*5-ZhAlTe-fuNYKTD" /></figure><p>Jean started by refreshing our memory on Go modules and introduced us to the multi-modules repository which is self-explanatory: a repository that contains multiple modules, each with its own go.mod file.</p><p>He then goes over the various steps of creating a multi-module project (git tag prefix, go.mod generation, etc.) and also the various issues and solutions with adding a submodule to an already existing package (import conflict).</p><p>To learn more about Go Modules, you can find the very well written blog post by the Go team <a href="https://blog.golang.org/using-go-modules">here</a>.</p><h3>Updating the standard error library</h3><p>By Marcel van Lohuizen <a href="https://twitter.com/mpvl_">@mpvl_</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*HMAguL9R-TP4-6pv" /></figure><p>Marcel van Lohuizen works in the Go team at Google. His past researches were on rethinking errors for Go 2. At the dotGo conference, he talked about enhancements done on the standard library’s packages to better manage errors (more information on this update: <a href="https://go.googlesource.com/proposal/+/master/design/29934-error-values.md">Proposal: Go 2 Error Inspection</a>).</p><p>To help the transition and have these new features working without breaking change with older versions of Go, he published a new package that will work with Go 1.13 and previous versions: <a href="https://godoc.org/golang.org/x/xerrors">golang.org/x/xerrors</a>.</p><p>Some changes that will appear in standard libraries:</p><ul><li>Error formatting in <em>fmt. </em>Now fmt.Errorf method will use FormatError if called with a format string ending with “: %s”, “: %w” or “: %v”, such as: fmt.Errorf(“error: %v”, err) . It enables a more explicit error to be returned.</li><li>Error inspection in <em>errors. </em>func <strong><em>Is(err, target) bool</em></strong> allows us to inspect error type and determine if any error in err string matches targeterror, true is returned in this case.</li></ul><p>This day was full of enriching talks. We want to thank the dotConferences team that has organized such a fantastic event with high-quality speakers. It was an immense pleasure to meet so many developers from all around the world.</p><p>— — — — —</p><p><em>Founded in 2012, Kapten has become a major player within the French ride-hailing market. We are made up of 270 employees, 15 nationalities, more than 2 million clients, and a presence in the major European cities. </em><strong><em>We’re looking for new talents to reach our ambition: becoming the European leader in urban mobility. </em></strong><em>If you think, as we do, that there’s no reason an American should take this spot… come and meet us: </em><a href="https://www.kapten.com/uk/careers/engineering/"><em>https://www.kapten.com/uk/careers/engineering/</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7529fb394bf" width="1" height="1" alt=""><hr><p><a href="https://medium.com/kapten-engineering/kapten-dotgo-7529fb394bf">Kapten @dotGo</a> was originally published in <a href="https://medium.com/kapten-engineering">Kapten Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What we liked about the KotlinConf 2018]]></title>
            <link>https://medium.com/kapten-engineering/what-we-liked-about-the-kotlinconf-2018-a822298f5e06?source=rss-c3269b83a48d------2</link>
            <guid isPermaLink="false">https://medium.com/p/a822298f5e06</guid>
            <category><![CDATA[kotlin]]></category>
            <dc:creator><![CDATA[Kapten]]></dc:creator>
            <pubDate>Mon, 22 Oct 2018 08:40:48 GMT</pubDate>
            <atom:updated>2018-10-22T12:32:24.580Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*380JXETCesGZMKAMPtwwmg.png" /></figure><p><em>Two weeks ago, Amsterdam hosted during 2 days the big Kotlin event of the year: </em><strong><em>KotlinConf</em></strong><em>. Chauffeur Privé’s Android team took the plane on a cold wednesday morning to participate to this Konference: 60+ speakers, more than 2 000 developers, 48 hours+ of coding, fun and technical talks with the best Kotlin developers around the world among which Andrey Breslav, Jake Wharton, Roman Elizarov, Romain Guy…</em></p><p><em>Here’s what we learned during these two amazing days! You’ll find </em><strong><em>6 selected topics</em></strong><em> we choose to mention because we believe they’ll have a real impact on the Kotlin community and on our daily job. And also because we badly want to implement them in the future!</em></p><h3><strong>Keynote</strong></h3><p><em>Author: Thomas Depinoy</em></p><p><strong>Andrey Breslav</strong> gave the start to this 2 days event by sharing with us some interesting information about the Kotlin language in general. First of all, the growth of Kotlin. Since 2016 the number of Kotlin <strong>developers</strong> has been <strong>multiplied by nearly 10</strong>, started from 156K to reach <strong>1.5M active developers</strong>. While Kotlin has gained a lot of popularity since the GoogleIO’17 announcement to be the officially supported language for Android applications, Andrey Breslav showed us that mobile developers are only a part of it and that Kotlin is designed for everyone.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Rx4qNzObeXRCqmaI" /><figcaption>Credits: KotlinConf 2018</figcaption></figure><p>He also spoke about how Kotlin is built day after day, by the Kotlin team (~50) but also by active developers among the community (~250). He insisted on how the feedback loop is important and how features become experimental (prone to feedback) before any release. <a href="https://kotlinlang.org/docs/reference/whatsnew13.html">Kotlin 1.3</a>, which was announced during the Keynote, is a perfect example of the team working with the community since the coroutines finally reached release state after being experimental on prior versions.</p><p>Video available here: <a href="https://www.youtube.com/watch?v=PsaFVLr8t4E">https://www.youtube.com/watch?v=PsaFVLr8t4E</a></p><h3>Getting a Grip on GraphQL</h3><p><em>Author: Jérôme Grondin</em></p><p><a href="https://twitter.com/brwngrldev">Annyce Davis</a> gave an interesting talk about UFOs with an example of how we can query a service to retrieve UFOs sightings between 2013–2014 with <strong>GraphQL</strong>! She presented what GraphQL is: a specification and a query language, and she really made a point when comparing a classic REST style API with GraphQL. In particular, she showed how to implement it server side (thanks to Kotlin data class) with schemas and resolvers, two main concepts in GraphQL. One disturbing thing was the client example which uses <a href="https://github.com/apollographql/apollo-android">ApolloAndroid</a>, a Java-based GraphQL client.</p><p>She also gave some tips and tricks if you want to start GraphQL on Android: a nice IntelliJ plugin for GraphQL, a Retrofit Converter, some cool GraphQL APIs and of course the full specification of GraphQL. She ended up the session with a <em>simple-non-controversial-question:</em> <strong>is REST dead?</strong></p><p>You can check <a href="http://adavis.info/2018/06/talk-getting-a-grip-on-graphql.html">her full presentation here</a> to make your own decision.</p><p>Video available here: <a href="https://www.youtube.com/watch?v=1NmXeZgtkvg">https://www.youtube.com/watch?v=1NmXeZgtkvg</a></p><h3>Gradle Kotlin DSL</h3><p><em>Author: Yann Vaillant</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*GS70OuqE_lEkbSkv" /></figure><p><a href="https://twitter.com/eskat0s"><strong>Paul Merlin</strong></a> a Lead Software Engineer at Gradle introduced us to the <strong>Gradle Kotlin DSL</strong>, a new way to write build scripts with Kotlin to improve on the aging Groovy.</p><p>Writing your build scripts with Kotlin allows for a much deeper IDE integration with easier refactoring and debugging, avoiding surprises at runtime with a faster feedback loop.</p><p>The improved tooling also grants us better API and model discovery with navigation to sources. He showed us a few examples of migration from Groovy to Kotlin and how easy it is. You can find it <a href="https://github.com/eskatos/android-sunflower/blob/7_app_script/app/build.gradle.kts">on this Github repo</a>, navigate through the branches (0_xxxx, 1_xxxx, etc…) to see a complete migration from Groovy to Kotlin Script with dependencies refactoring. If you want to know more, follow<a href="https://eskatos.github.io/kotlinconf2018-type-safe-build-logic/#/"> this link</a> to see his (beautiful) slides. We at Chauffeur Prive are already convinced and migrated our build scripts to the Kotlin DSL in less than a day with no issue and lot of benefit!</p><p>Video available here: <a href="https://www.youtube.com/watch?v=mAtrEPeAJSc">https://www.youtube.com/watch?v=mAtrEPeAJSc</a></p><h3>Coroutine</h3><p><em>Author: Lucien Guimaraes</em></p><p>The star of this year was definitely <strong>Coroutine</strong>, allowing to work with asynchronous programming, concurrency, actors and more. Further, than being simple to implement and easy to read, Coroutines are part of Kotlin itself, opening elegant possibilities with multi-platform.</p><p>The number of talks proved the popularity of this feature. Our favorites among them were from <a href="http://chris.banes.me/suspenders">Chris Banes</a>, <a href="https://youtu.be/jT2gHPQ4Z1Q">Venkat Subramaniam</a>, and Roman Elizarov. The first one focused on how to use them inside an Android app, and how helpful it can be compared to existing APIs (Loaders, Future, AsyncTask …). There is an available CodeLabs for Coroutine in Android <a href="https://codelabs.developers.google.com/codelabs/kotlin-coroutines/index.html">here</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/878/0*yveSKkBXTwylf6oL" /></figure><p>The second talk was interesting to understand the basics, how it works and how easy is the implementation in an existing project, without boilerplate.</p><p>Last but not least, Roman Elizarov (Coroutine’s guru), took a deep dive in the API to point out memory leaks and how to avoid them. Then we had an overview of <a href="https://kotlinlang.org/docs/reference/coroutines/channels.html">Channels</a>, which provide a way to transfer streams of values. To end his talk, he explained the differences between <a href="https://kotlin.github.io/kotlinx.coroutines/kotlinx-coroutines-core/kotlinx.coroutines.experimental/-coroutine-scope/index.html">CoroutineScope</a> and <a href="https://kotlinlang.org/api/latest/jvm/stdlib/kotlin.coroutines.experimental/-coroutine-context/">CoroutineContext</a> and how to wisely use them.</p><p>Video available here: <a href="https://www.youtube.com/watch?v=a3agLJQ6vt8">https://www.youtube.com/watch?v=a3agLJQ6vt8</a></p><h3>Multi-platform</h3><p><em>Author: Etienne Pinault</em></p><p>We had <strong>huge</strong> expectations about multi-platform development with Kotlin: we have not been disappointed. First of all, Andrey Breslav announced Kotlin Native has now reached Beta. It produces native code using the LLVM API targeting iOS, macOS, Linux, Windows and embedded systems (x86, ARM, …). By adding this support, Kotlin is now available for JVM, JS and Native developers, which means that you are now able to build your mobile and desktop applications in Kotlin as well as your backend, and even your frontend.</p><p>Marcin Moskala presented one possible multi-platform architecture fully written in Kotlin. In our opinion, this was one of the most amazing parts of the whole event.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*GLCfHlE-RsRNPJ6I" /></figure><p>He advised to let iOS developers taking care of the view part using the common code processed by the Kotlin Native compiler into an Apple framework.</p><p>If you are interested in the subject and want to know more about it, you can check the GitHub of the KotlinConf application which is a <a href="https://github.com/JetBrains/kotlinconf-app">very good starting point</a>. We will probably try it out ourselves in the next weeks.</p><p>Video available here: <a href="https://www.youtube.com/watch?v=UyTBXEZ983g">https://www.youtube.com/watch?v=UyTBXEZ983g</a></p><h3><strong>Annotation processing</strong></h3><p><em>Author: Quentin Huet</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*bQzamlhqe4Se0uC6" /></figure><p><a href="https://speakerdeck.com/hzsweers/annotation-processing-in-a-kotlin-world">Zac Sweers from Uber introduced</a> us to meta-programming and talked about annotation processing in Kotlin. He reminded us of annotation processing in Java and talked about some weird behaviors you can experience in a Kotlin project. He used Dagger2 as an example and explained why we had to do fixes like adding `@field` or `@SupressJvmWildcard` into our code. If you already used Dagger2 with Kotlin, you should probably know what he means…</p><p>Then, he went deeper into annotation processing in Kotlin, explaining how it works and exploring concepts like the use of the Kotlin.MetaData annotation, protobuf representation of our class and how to nicely generate code with KotlinPoet.</p><p>He finished with an example of how to build a processor for a new annotation, sharing advice and good practices.</p><p>If you want to learn more about this topic, why it can be useful for your app or library and how you can implement it in a proper way: take a seat and <a href="https://youtu.be/a2RoLFzrFG0">watch Zac’s video</a>, it’s inspiring!</p><p>Psst: You can deep dive into the future of meta-programming by looking to the <a href="https://youtu.be/w-GMlaziIyo">plugin compiler talk</a> made by Kevin Most from JetBrains. It’s still not public but should become an experimental feature soon, stay tuned ;)</p><p>Video available here: <a href="https://www.youtube.com/watch?v=a2RoLFzrFG0">https://www.youtube.com/watch?v=a2RoLFzrFG0</a></p><p>If you want to discuss with us, feel free to ask in comment or drop a mail to <strong>medium@chauffeur-prive.com</strong></p><p><em>If you’re searching for a new job, you can find </em><a href="https://chauffeur-prive.workable.com/"><em>our openings here</em></a><em>.</em></p><p><em>A huge thanks to our awesome Office Manager Lénaïc who organize this trip for us with ❤️</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a822298f5e06" width="1" height="1" alt=""><hr><p><a href="https://medium.com/kapten-engineering/what-we-liked-about-the-kotlinconf-2018-a822298f5e06">What we liked about the KotlinConf 2018</a> was originally published in <a href="https://medium.com/kapten-engineering">Kapten Engineering</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>