DotNext 2018 Piter program overview

Conference : DotNext 2018 Piter
Date : April 22-23, 2018
Location : St. Petersburg, Park Inn by Radisson Pulkovskaya Hotel

Next DotNext will be less than a month. Enormous work was carried out on the program and reports, and in order to acquaint you with it, we asked Roman Nevolin to tell about it. The novel is included in the PC DotNext, and as a developer and speaker specializes in working with data in high-tech projects, is engaged in research in the field of Machine Learning and developing its own machine learning tools. You could be present at his reports like “F # to the glory of Data Science” or “Machine Learning on the .NET platform” . I give the word to Roman.

Good day, Habr! My name is Roman, and I am one of the members of the DotNext Conference Program Committee. We this year, to admit, are proud of our program as never before. We were able to invite a bunch of people we have long wanted to see at the conference, and most importantly - to achieve a balance between different themes in order to show .NET in all its diversity. From security to Xamarin, from harsh enterprise development to quantum computer programming, we tried to take as many topics as possible and show them from various sides.

And since in all this diversity it is easy to get lost, we have made for you an overview of the reports of the conference, divided into two articles. In this part - the most hardcore themes that DotNext has always been famous for. I would like to warn in advance that the article is based on my impression of the materials that we have right now - records of training sessions, slides, a report outline, and so on. And since many speakers have the good habit of redoing half of the report’s content a week before the conference, something may change.

Performance

Performance has always been one of the most interesting topics for the DotNext audience - performances about performance always collect full halls, so every year we try to find new, interesting reports (and speakers!) On this topic.

For me, the most interesting in this category is a report from Andrei Alexandrescu, famous for his excellent books and articles on various topics, as well as the programming language D, one of whose authors he is. Unlike many other performance reports, this one doesn’t touch the inside of the framework. Instead, Andrew will talk about the principles of developing high-performance algorithms, demonstrating them with clear examples. This approach makes the report useful even for those who have never thought about the performance of their applications. But I warn you - algorithmic hardcore here just pours over the edge and you should prepare for it in advance. As one of the responses from Andrei’s website says: “All I have to say is: mind == blown”.

Federico Lois looks at this problem from the other side: he talks about those aspects of performance that are directly related to iron. For example, you can learn how the CPU architecture affects the various performance parameters of your application and what changes CoreCLR 2.1 has made to this. All this will be demonstrated on the example of RavenDB - a high-performance NoSQL database on .NET, developed by Federico. By the way, last year his report was included in the top 5 best reports of the conference. Surely this time will be no less interesting.

CLR and Runtime

In this category there are reports on low-level implementation features of the framework, which we encounter not so often and know very little about them. And the more valuable is the opportunity to learn about the subtleties and features of .NET from people who work with it every day.

One of the topics that is not discussed very often in the .NET community is the Ahead-Of-Time compilation. And if you are interested to know what it is and why it is needed, you should listen to the report from Elizaveta Golenok “We use AOT-compilation correctly”. She has a lot of experience working with different types of compilations and therefore she can clearly explain whether you need an AOT compilation, how to live with her and that she will help you win, but at the same time she will tell you that actually AOT is everywhere.

Another interesting report from this category will be made by Evgeny Peshkov from SKB Kontur, who will tell us about special exceptions in .NET. It would seem that there could be such a new and unknown? However, there are a lot of exceptions that we encounter quite rarely, and it would be better not to encounter them at all, because they occur with operating system level and runtime errors, and each such exception promises you a long debugging, tinkering in memory dumps and issues on StackOverflow. It is possible that this report will someday save you time and nerves by letting you know in advance about the causes and consequences of such exceptions.

And the last in the program is a report on the framework's internals - “Multi-threaded structures in .NET. How it works?". He will reveal to you the features of multithreading implementation in .NET, showing what lies behind the standard competitive collections. If you have never had the heart to figure out how it all works, Dmitry Ivanov’s report is what you need. He has, on the one hand, a vast experience with multithreading (the development of ReSharper and Rider - this is not to rivet web forms), on the other - the constant experience of speaking at meetings and conferences, which will allow you to transfer this experience to an understandable form and pass it on to you.

Debugging and Testing

Perhaps more time than inventing names for variables, the programmers take only one thing - debugging the code. How much time is spent on unsuccessful attempts to find an insidious bug that is hidden where it is not clear where it is manifested only on odd Fridays - do not count! Reports in this category will help you take advantage of someone else's experience and save a lot of time, making the code better, and life easier.

I will begin with the report of Sergey Shchegrikovich with the intriguing title “The Main Secret of Debugging”. Straight and I want to believe that you just have to listen to him - and that’s it, the bugs will be fixed by themselves, and the question “why does this damn thing fall” will never torture you again. So, as a person who has already heard this report, I have to disappoint you - the bugs didn’t fix themselves. But I got the Universal Error Search Algorithm, I realized when it is worth shouting loudly: “This is not my bug!” (No spoiler in any), and in general I systematically systematized my knowledge. In my opinion, paying attention to the report is absolutely for everyone - one will like the debugging algorithm, others will have plenty of life examples, and certainly no one will be disappointed with the narrative style: the report is heard as a set of detective stories, with conclusions and morality.

While Sergey talks about universal things, the report “Advanced .NET debugging techniques from a real world investigation” from Christophe Nasarre and Kevin Gosse talks about things that are much more mundane - correct metrics that help to notice anomalies, convenient tools for debugging, and various situations where it can help you. All this is shown on a large number of real-life examples, and almost half of the report is a practical demonstration of the proper use of certain tools. It is recommended to listen to everyone who has not fully mastered the variety of tools for debugging or just wants to see how these tools help solve everyday problems.

And it all works fine, but not for cloud applications. In the cloud, their problems, bugs and debugging methods. And therefore, Maarten Balliauw will talk about debugging cloud applications in a report “What is going on? Application diagnostics on Azure. In an hour, literally from scratch, you will figure out how to receive and interpret various information about your application hanging in Azure, what problems it may have and how to find errors even before they occur. By the way, many of the things that Marten talks about are not only suitable for Azure - tools like structured logging and AppInsights are good for any application. Therefore, it makes sense to look, even if you do not plan to send your code to the cloud.

The last report in the section approaches the problem from the other side. How to catch bugs with logs, metrics and dumps, where it’s better to reveal them by auto tests, right? And the report of Yulia Kovaleva “The Rageless Pizzeria” is about this - about tests and code testing. Julia will talk about how they made the Dodo IS system a mess, introducing a property-based approach and correct code testing practices. Why is it important to be friends with testers and how to write tests, where the code will be less, and the coverage - more? In general, this is a good reason to look at autotests from the perspective of a professional tester and improve your code testing skills.

Security

This year, the author of two excellent books on this topic and the creator of the Inferno crypto library, Stan Drapkin, is responsible for .NET Security on DotNext. Stan will come to us with the report “Pitfalls of high-level cryptography”, which he himself has positioned as a logical continuation of reports on this topic from past conferences (in particular, the report from Vladimir Kochetkov with DotNext Moscow 2015) - they talked about low-level cryptography , and this report is intended to show that at a high level, too, is not so simple. And I have to say, Stan did it. In his report, he showed how the correct high-level library working with cryptography should look like, how this high-level API should be used in order not to fall into any of the underwater holes and in what situations it is worth using (or not using) certain cryptography methods. In addition, the report is simple and pleasant to listen to, even for a person who is far from the subject of cryptography. In general, I recommend listening to the report to anyone who is at least interested in the topic of security - it's worth it.

We remind you that tickets for DotNext can be purchased on the official website .