A little bit about programming trends in 2016: a series of conferences #ITsubbotnik

The very beginning of the coming year is a great time to take stock of the past year and look at some of the programming trends that will be exactly relevant in 2017.

October 29 in St. Petersburg and Samara and on November 19 in Saratov a series of conferences for programmers #ITsubbotnik took place, bringing together information technology specialists from various fields. EPAM experts discussed the main trends in IT 2016 and tried to touch on most of the sought-after languages ​​and technology stacks. Let's see how it was.

What is #ITsubbotnik?

3 cities, 40 reports from experienced professionals and more than 800 participants.
')
How was it in Petersburg

This year it was possible to visit the streams devoted to:

• Java
• Big data
• QA
• Javascript
• .Net
• Front end
• Iot

There were really a lot of reports, and in each city their themes and headliners came to the fore. Trying to retell the material of the speakers for themselves is a thankless task, but we will try to identify some of the key presentations for you. And for those who finish reading to the end, we have prepared a tasty bonus.

Big data

Today, specialists working with Big Data and related technologies are very popular. An interesting fact: in the recent elections in the USA, the headquarters of one of the candidates predicted victory even before the elections, creating a complex mathematical model that predicts the potential behavior of voters based on the analysis of a large amount of data on states over the past several decades. The result we all already know.


This is probably what Spark developers look like.

At #ITsubbotnik in St. Petersburg, Alexey Zinoviev, Senior Training and Development Specialist, spoke about Big Data with examples of real-life cases and plunged into the Spark jungle with participants:

“It all started in 2014 with a talk on how to do Data Science in Java. Then we went deep into NoSQL and Hadoop. The next in line was Spark (this is logical), and not just Spark, but real-time mode. In the summer, a hot and interesting Spark 2 came out. I have some combat experience with its “endless DataSet,” a special layout of objects in memory and in solving the problems of this revolutionary, but unfinished release. ”

During the report, Alexey told what was going on at the junction of the data processing engine itself and its friends (Kafka / Samza / MLlib), who supply this same data in groups or individually. The report was planned in two parts. Continuation will be released in the spring of 2017.


Lecture by Alexei Zinoviev, viewing time: 55 minutes

At #ITsubbotnik in Samara, Evgeny Kozhevnikov, EPAM Solution Architect, told how to work with Big Data correctly and dismantled important aspects in nuances.

Over the past 10 years, the principles of working with large or constantly growing volumes of data have changed a lot. The best of them are now decided to combine the concept of Big Data, and the technologies that form the platform are usually afraid. This is understandable - simple in theory, the principles of building distributed systems in practice are overgrown with millions of details and nuances. And it was the practical side of the issue that this workshop was devoted to.

As it was in Samara

Armed with his own production experience in the field of big data analysis, during the report from a clean slate he built prototypes of real Big Data architectures, which is called “on the knee”, in the comfortable conditions of a home computer. Right on the hardware that was at hand (on a not very powerful laptop), he emulated a real cluster, unfolded the necessary components, made friends with each other and at the same time made this distributed zoo realize the task.



In the final report, Yevgeny admitted:

“Of course, for the production-solution, many unnecessary gestures were made - on a real project, as a rule, you will have to wait for a ready cluster of already friendly Big Data technologies. But do not be naive, only the hardcore experience covered in this workshop will allow you to understand and effectively use such an environment. And most importantly, next time, trying to comprehend Zen of a new Big Data technology, you will know which way to approach it, so that everything will turn out right ”

Front end

“JavaScript is gaining in popularity more and more. Perhaps this is due to a large number of languages, since adding the ability to write a Desktop application in JS is easier than adding support for other languages ​​to browsers. ”

- considers Dmitry Turlychkin, EPAM Software Engineer.

In his report on #ITsubbotnik in Saratov, Dmitry addressed the following topics:

• The task of creating applications in one language for everything - both for mobile applications and for desktop solutions - can be solved with the help of JS. Ionic, React Native and other technologies
• Native JS support in devices on different OS
• Intel's Edison Platform and IoT Market
• Simplify the creation of applications on various frameworks, the developer community and the struggle for users

Deep learning

Another important area of ​​technology that has been very popular lately is, of course, neural networks and machine learning. Telegram chat bots familiar to everyone, smart keyboards in smartphones and Tesla autopilot are all from there.

Vsevolod Rodionov, Lead Software Engineer, in his report told how to make a simple neural network, without switching from JavaScript to Java, Python or R. By the way, he already spoke about this in more detail recently.

What is neural networks, what distinguishes linear and nonlinear dependencies, what the final code looks like and where to find the sources for building your own network - all this can be learned from the material of Vsevolod.

We write a neural network in an hour in JS

VR

Today, a lot of information is devoted to the development of AR and VR technologies, however, articles about all the features of testing such software cannot be found. In his report, Anton Tulyaganov, Senior Software Testing Engineer, shared his experience in developing projects in EPAM. It is useful for beginner testers as well as for those who are already working on AR / VR testing.

The report contains information on how AR / VR technologies work, on the types of devices for them. Anton discussed with the audience what to pay attention to when testing and planning it, described the types and features of software testing for AR / VR.

Lecture by Anton Tulyaganov

QA

For 4 years, EPAM developers worked on creating a Report Portal, a tool for automated testing. The product turned out so promising that the authors decided to post it in open source. The product manager, Dmitry Gumenyuk, who has been working at EPAM for 9 years, not only read a great report on #ITsubbotnik, but also gave an interview to dev.by about this popular tool.

“Report Portal is a tool that makes it possible to see the reasons for the loss of automation efficiency, to categorize test drops and, based on accumulated results, to recognize new drops automatically.

Report Portal provides analytics in the form of visual graphs, can collect statistics, store history and screenshots of what is happening, and also simplifies working with reports on the launch of automated tests. Accordingly, the problems are visible much earlier and can be solved more quickly.

The authors hope to distribute the product among the communities of automators, testers and developers. ”

You can familiarize yourself with the development by the links one and two.

For those who read

As promised, we will share something interesting with those who have reached this place. But first we want to thank all those who are interested in our events, listened and talked to our speakers, who invest time in developing themselves in the profession.

We offer to see the full record of all lectures from St. Petersburg #ITsubbotnik on our channel on Youtube .

And here you can see the reports from Samara .

Stay in touch and see you soon!