Conference Program PYCON RUSSIA 2015

Hi, Habr!



Three weeks left before the third international conference of python developers Pycon Russia 2015 , which will be held in Yekaterinburg on September 18-19.
')
We finish the final grid and discuss the details of the latest reports, but 90% of the program is ready.

So what awaits you this year.

Reports PyCon Russia 2015

The developer of CPython , a member of numerous Python projects, including PyPy and py.tes, Benjamin Peterson ( San Francisco, USA ) will give a talk on the topic “Python Packaging Progress”.

Python packaging is infamous and it is hard to use. Thankfully, a lot of progress has been made in recent years. Piping, PyPI, virtualenv, and wheels.
I'd say to talk to the advanced audience. Ideally, it will be a clear idea of ​​how to optimize. :)

Researcher at the National Research Institute INRIA in France , library developer scikit-learn, joblib, Mayavi and nilearn, member of PSF Gael Varoquaux will tell PyCon how to create an advanced data-processing environment for the budget.

As a penniless academic I wanted to do “big data” for science. Open source, python, and simple patterns were the way forward. Staying on top of todays datasets is an arm race. Data analytics machinery — clusters, NOSQL, visualization, Hadoop, machine learning, ... —can be a resource thin. It makes it easy to find out how to use it. For the out-of-the-core and parallel computing.

Identity of data processing with Python. What are the emerging patterns in data processing? How can modern data-mining ideas be used? What constraints and design trade-offs govern projects like scikit-learn, Mayavi, or joblib? Simple framework-less code?

For those who are learning, studying or interested in learning python, Daisuke Saito from Japan will give a talk on “Python Learning with Minecraft”.

This report propose Python-learning method with Minecraft. Minecraft is a popular sandbox game in the world. We are using Mincraft and Mod (Extending Minecraft Functionality) for Python learning. Mod is using RaspberryJam to provide Python to Minecraft. You can use minecraft in Python by using it.

One of the founders of PyConRu, a member of the program committee of the two past conferences, and now a resident of Bonn and a programmer at the United Nations Anton Patrushev will give a talk for hardcore pythonists.

Often, when developing, we are faced with situations where, to increase the productivity of an application, it is necessary to do large refactorings, to change the architecture. But what to do if there is no time for this and you need performance right here and now? Of course, first of all, you need to resort to traditional medicine - to profile and correct the bottlenecks. But much has already been said about this. I would like to talk about several unconventional techniques that once helped me overcome performance problems.

Alexander Sibiryakov , Python-developer at Scrapinghub (Czech Republic, Prague) , will give a talk about the new open source framework developed at Scrapinghub.

Frontera allows you to build a distributed robot for downloading pages from the Internet in large volumes in real time. It can also be used to build focused robots to siphon a subset of previously known websites. In addition to the description of the framework and system requirements, I will talk about our experience of downloading Spanish Internet using Fronter and present some statistics.

Python Core Developer, engineer at DataRobot , organizer of PyCon Ukraine, one of the program directors of PyCon Russia, Andrey Svetlov (Kiev), will talk about Aiohttp, the most popular asyncio-compatible library today. For those who want to try the asyncio approach, but do not know where to start.

In the report: Web-client. Primitive request. Use authorization and sessions. Web server Primitive example. Complex routes. Web sockets. Client and server. Processing parallel tasks (asyncio.wait) Testing. Explicit loop Databases. Redis, Postgres, Mongo. Sessions Authorization and access rights. Debugtoolbar. * PEP 492

Andrei Vlasovskiy (St. Petersburg) , developer of PyCharm and Vim emulator for programming environments in JetBrains. The author of the funcparserlib library for writing Python parsers in a functional style will tell you what type annotations can be useful for you.

PEP 484 introduces type annotations for Python 3. Type annotations can improve code readability for both people and code analysis tools. They make the code better and safer. In this report we will discuss why this is so.
We briefly review what type annotations are and move on to practical examples of where they can be useful and what they give. Next, we discuss some useful tips on writing and using annotations. We will see that in most cases simple type annotations based on classes and built-in collections are sufficient for the public APIs of our libraries. We will also discuss what type annotations can provide for third-party libraries.

Python Backend Developer in Toptal and QA Automation Lead in Wargaming.NET Andrey Soldatenko from Kiev will tell you about full-text search using Python.

Nowadays the world is drowning in textual information. All these years, existing technologies have focused on storing and structuring data. But what to do if we need to make a decision in real time using this data? I’ll be happy to share my experience in developing web applications using full-text search, which are written in Python using the Django framework, and also talk about the results of research from various search engines and integration of web applications with haystack and elasticsearch. Also talk about the future of full-text search in the context of Django 1.9 and Python.

Grigory Petrov (Moscow) , a technical evangelist at VoxImplant , the speaker of many IT conferences, will make a review report on modern approaches to caching in different areas of development in python, will outline the current state of the ecosystem and share interesting cases from practice.

Python Lead Development Engineer at Rambler & Co , Rambler project team lead. Video Pavel Petlinsky (Moscow) will give a talk on “Making of external DSL for Django ORM” on PyConRu.

The story goes about how we implemented in Django the same mechanism as Advanced Searching in JIRA, about the internal structure of such solutions (writing DSL, parsing grammars, etc.), why and how we use them on our project.

For those who are interested in scientific calculations and data analysis, Ekaterina Tuzova (St. Petersburg) , PyCharm developer, lecturer and curator of student projects at CompSciCenter and at the Academic University, will tell you about the Numpy library.

NumPy is the basic library for scientific computing in Python. On the basis of NumPy arrays, many libraries have been built for various fields of science: machine learning, symbolic computations, and work with biological data. However, in order to use NumPy effectively, you have to slightly change your approach to writing code. We'll see why Python loops work slowly and how the use of vector operations from Numpy improves performance.

Yandex authorization systems developer Kirill Borisov (Moscow) will teach you how to avoid a routine for a programmer.

One of the most unpleasant things in a programmer’s work is routine. Unfortunately, it cannot be avoided - the work on maintaining the existing code forces us to make the same changes to several files at once. As the project grows, this task grows to dozens of edits in hundreds of files, forcing even the most patient programmer to think about the meaning of life. But what to do? Automate it! In this report, we will talk about different ways to work with source code from Python, and explain why this is fun.

The developer of the Yandex.Taxi service, leading the Key Technology column (Core Technology) in the Linux Voice magazine, Valentin Sinitsyn (Yekaterinburg) , a teacher at the Ural Federal University, will talk about memory management in python.

In this report, we will review the memory management subsystem in the most popular Python implementation, CPython. Procedures for allocating memory and garbage collection, ways of interacting with them from the Python code, as well as typical errors that may interfere with their normal operation will be considered. We will also demonstrate a series of simple experiments that prove that working with memory in Python really happens this way and not otherwise.

DataArt Senior Python Developer, organizer of the Python community in St. Petersburg, developer, free software apologist Sergey Matveyenko, will give a talk on how Python is used in the most popular and actively developing IT direction “Internet of Things”.

I'll talk a little about the available methods to lower the IoT entry threshold, talk about the capabilities of the “alljoyn” protocol, one of the most popular IoT protocols, and show with an example of this device how to use the DeviceHive D-Bus Framework from any managed devices that you have at hand, devices that operate on the "alljoyn" protocol.

Product Director at Marilyn Alexander Shvets (Moscow) will tell useful recipes with Celery.

• The secret composition of celery or an informal introduction to the architecture of Celery;
• We select the ingredients for the sample. How to log and debug queues correctly;
• Take care of a long finish. How to ensure stable operation of queues without downtime and loss of tasks;
• Preparing for the culinary competition. How to avoid racing when performing tasks;
• Several professional chef recipes for setting up celery for large projects.

Alexander Kozlovsky and Alexey Malashkevich (St. Petersburg) , the authors of the object-relational mapper Pony ORM , PyCon 2014 speakers, will teach how to create a one-page ReactJS application using PonyORM / PonyJS objects as models this year.



Ilya Beda (Krasnoyarsk) - developer of web applications with 7 years of experience, team leader bro.agency , will tell you about caching web applications.

In most web applications, we somehow interact with the database. It stores information, and we, as programmers, write code that converts this data into html pages or API interfaces. And since the single point of truth is the database, then we need to cache the WEB applications based on its state. In my report, I will talk about the methods of caching, based on the ORM. How you can cache database queries and functions that depend on them. And most importantly, about how such a cache is correctly invalidate.

Dmitry Vakhrushev (Omsk) , web developer, pythonist, linuxoid, author of a number of open source projects and technical blogs , will give a talk about building a RESTful API on Pyramid.

Pyramid is an extremely flexible framework ideally suited to those developers who know exactly what they need. The other side of its flexibility is the absence of the only correct, approved by the authors, ways to solve problems. In this report, I will share my time-tested recipes for building a RESTful API on Pyramid.

Alexander Shchepanovsky (Krasnoyarsk) , the author of several popular (funcy, cacheops) and a little more unpopular libraries, will tell you how to write your own rules for linters and not experience the painful pain of their deeds.

To maintain our standards for both coding style and its semantics, we often use linters. However, adding your own rules to them is quite difficult. Therefore, in the majority of teams, the means in no way or almost do not adapt in any way to either the specific project or the team as a whole. The code review allows you to partially close this flaw, however, sometimes you just want to poke a finger and say "never do that again." I will talk about my research project on how to make the automation of such situations simple and therefore flexible and practical.

Graduate MGTU them. Bauman (Computer Science and Management) and the Higher School of Economics (Business Informatics), Konstantin Ignatov (Moscow) , programmer in the Qrator Labs research department, will tell PyConRu about creating setup.py and setup.cfg files.

In the report we will look at the creation of the setup.py and setup.cfg files. It will be useful to anyone who develops> 1 projects interacting with each other, copies into each new project a folder with favorite functions, installs projects using git clone, copy, rsync and etc., does not write extensions due to the complexity of deployment, writes console scripts using def main () ... if name == 'main' ..., avoids packaging because projects are too dependent on each other (changes in one lead to changes everywhere).

Full abstracts of the reports read here .

Student discount

We want as many people as possible to attend our conference. At the same time, we understand that students cannot always pay for a ticket. Therefore, traditionally, we offer a discount to students - 50% of the price. That is, a ticket to the conference for students costs 5,000 rubles.

To get a discount, send a student scan to om@it-people.ru, in response we will send a code for registration.

Teachers associated with the python disciplines participate in the conference for free.

check in

Sign up here . We saved the price of last year - a ticket costs 10,000 rubles.

The ticket price includes: participation in two days of the conference (at this point you need to look at the program again), lunch 18.09, dinner 18.09, lunch 19.09, coffee breaks and afterparty, a bus from the center of Yekaterinburg to the venue and back. Paid separately: the place where you will live (for example, on the Oriole - from 2800 for a double room per day with breakfast), and alcohol after the party.

PyCon Russia - 2015, September 18-19, Yekaterinburg, Oriole Training Center.