46th Monthly Technical Session (MTS) Report

46th Monthly Technical Session (MTS) was held on May 18th, 2018. MTS is a knowledge sharing event, in which HDE members present some topics and have QA sessions, both in English.

Our first speaker for this session was Tanabe. He explained how gVisor works. gVisor is a user-space kernel, written in Go, developed by Google. More specifically, it implements Linux in a user-space in Go. gVisor exists because container cannot provide true isolation and machine level virtualization is too expensive. gVisor provides virtualization by intercepting system calls made by application in user space, then decides based on rules to forward it to real OS below it or deny it providing better security in sandboxed environment.

Next was Okubo. He brought up the presentation to invite us to enjoy Azure together, titled Let’s enjoy Azure. He began his presentation by introducing some problems he faced and his decision to choose Azure as deployment platform. He compared various Azure services with AWS with emphasis on Azure offers more managed services than AWS. For example, Azure provides service called App Service for Containers. To achieve the equivalent said service benefit that Azure offer, AWS mandates user to manage 3 services, those are Elastic Load Balancer, Autoscaling Group, and Elastic Container Service.

Our third speaker was Bagus. He introduced Amazon GuardDuty. Amazon GuardDuty is a fully managed service from AWS to monitor and to detect threats. He explained how he implemented this service on HDE accounts.

Next speaker was Koji. His presentation title was Go with Clojure. He began his presentation with the birth of Clojure. He then explained language feature of Clojure and its similarity to other programming languages such as asynchronous programming, channels, pipe, etc.

The last three presenters were our Global Interns. The first intern to present was Yuhao from China, currently studying in Osaka University for his Ph.D. His presentation title is Learning a New (natural) Language. It was not about programming language but it was about Japanese language. He explained why he needed to learn Japanese and how he learned it and provided us some tips to remember the words better using app. However, the app he was using required him to enter the words to remember manually and it was a hassle to put the words he wanted to remember every time he saw new words. Thus he came up with an elegant solution to let script do the job for him. To achieve this he extracted the words he marked when reading e-books on his kindle database, then he looked up the words in online dictionary, and finally put those result into the app. When he did the task manually, it took him 2~3 minutes on average to put 1 word into the app, but when he automated it, it only took 1~2 seconds on average to put 1 word into the app.

Our second Global Intern was Ray from Hong Kong. His presentation was about React.js. He titled his presentation Mini review of React in 2018. He began his presentation with flashback to 2016 when he thought:

I believe Single Page Application will be the future.
I had to choose React or Angular.

After that, he explained his reason to pick React as his tool to build single page application. The reasons he chose React were:

• Simple, not too much to memorize
• Straightforward for JavaScript developers
• Forces us to have a good understanding on JavaScript
• Render only, easily replace parts of the application
• Works well with new technology, e.g. GraphQL, RxJS, PWA (service worker)
• Large community
• Easy to find answers
• Many 3rd party libraries

He also highlighted several reasons he still uses React to this date as follows:

• Works well with type checker, e.g. TypeScript, Flow
• Refactor confidently
• React Native, Electron
• Develop mobile / desktop apps with my familiar work flow
• Good things will still be good things.

Our last but not least Global Intern was Silvia from Canada. She presented Self-driving Cars 101. As the title implies, she explained various things related to self-driving cars from the definition of autonomous vehicles, 5 automation levels, instruments and technologies developed for autonomous vehicles, engineering reasons behind the inception of self-driving cars, legal status, and many more. She emphasized her presentation on reasons to go for driverless cars.

Benefits of self-driving cars are:

• Save more lives by reducing crashes than human driven cars
• The “Drivers” never get distracted
• Eating, texting, TV… You may be to do more in the driver seat!

Self-driving cars have some drawbacks as follows:

• No sophisticated regulations
• The public is hesitant to accept new technology
• Expensive. E.g. A Velodyne Lidar (VLP-16) costs ~$4,000 = ~45万円

As nothing is perfect in this world, so is self-driving cars. On March 2018, a self-driving Uber struck and killed a pedestrian crossing the street in Arizona. Silvia then explained the consequences of the incident:

• Uber was suspended indefinitely from autonomous vehicle testing in Arizona
• Other autonomous cars project may also be affected
• Loss of public trust in the technology

She closed her presentation with the state of self-driving cars in the 5 automation levels she mentioned in the beginning of her presentation. Self-driving cars still have long way to go to reach the full automation level 5, no driver needed at all situation, as currently, self-driving cars are still on level 2 of automation, that is Partial Automation, which still requires human intervention on various cases.

As usual, we had a party afterwards :)