PYCON RUSSIA 2017: video of all reports and presentations
On July 16-17, the fifth international conference of python developers PyCon Russia was held 95 km from Moscow. Under the cut - a lot of videos, presentations and photos. And also look at the reporting video - it briefly describes how PyCon-2017 went.
Python ore developer since 2010, Facebook developer, âchronic perfectionist, pianist, dadâ Ĺukasz Langa (California, USA) spoke on âGradual Typing of Production Applicationsâ.
')
One of the developers of PyPy , the main developer of jitviewer - a tool for analyzing the performance of python programs under PyPy, Maciej FijaĹkowski (South Africa, Cape Town) explained âWhy is Python slow?â.
Developer at Venmo , previously at Reddit and HBO, Nina Zakharenko (Portland, USA) talked about common anti-patterns in python programs and showed practical python solutions for improving your code using tools like Decorators, Context Managers, Mixins and Lambdas .
The performance of the PyPy interpreter is achieved through specialization, as elsewhere. Alexander Koshkin (St. Petersburg) told Senior Python developer at Positive Technologies what it means and how RPython allows building fast interpreters of arbitrary languages.
Elizaveta Shashkova (St. Petersburg) from JetBrains described how the new interface for computing frames in Python 3.6 works, how it can help you create a quick debugger, and why such a quick debugger could not be created in previous versions of Python. For those who have not yet made the final decision about switching to Python 3.6, this report will provide several additional reasons why this is worth doing.
Scrapy is a good example of a modern asynchronous application. Moreover, it combines in itself every functional: Item pipelines, HTML / CSS selectors, Middleware, support for several network protocols, stop / continue from the moment of stopping, and many others. Alexander Sibiryakov (Czech Republic, Prague) from Scrapinghub described how data processing looks from within Scrapy, the architecture of the module's queue for loading content and other components necessary for its debugging: Scrapy shell, telnet console, memory consumption debugger.
Kirill Borisov (Moscow) from Yandex considered the place of Python in the IoT world, how to use it in communication with various pieces of iron and on what it is launched for the sake of great justice.
In 2014, MicroPython appeared - a Python implementation for microcontrollers. How did you manage to get Python to work on something with only 16 kilobytes of memory? Did you have to cut the language hard for this? (spoiler: no!) What interesting things can be done with glands and MicroPython? Different sensors, accelerometers, LEDs, motors - that's all in the report of the PyConRu program director, the PyCharm Community head at JetBrains, Andrei Vlasov (St. Petersburg).
Mikhail Yumatov (Moscow), head of development at TsIAN , told us what tools there are to monitor the performance of web applications. How to use them to collect detailed information on all operations, quickly detect problems and respond to them. Mikhail also explained what statistical analysis is and how to use it for a detailed study of the performance of parts of the application.
Yandex is a search company. But itâs good to find information not only on the Internet. Inside Yandex there is a huge intranet - a large number of internal services, which also need to be effectively searched. The company is actively using Python, they have a rather interesting infrastructure - there are components on both Django and Tornado. They actively use Celery with different brokers. Interact with internal tools and APIs to reuse the big Search developments. Alexander Koshelev ( Yandex , Moscow) told how all these components interact and what interesting features there are.
Although development in Python has shifted towards the server segment, desktop applications in Python are still relevant. Moreover, with an increase in processor performance, python-based applications have become an opportunity to reduce the financial, human and time costs of releasing desktop versions. And the most interesting point in this is the multiplatform nature of such applications. Igor Novikov (Kharkiv-Lviv) from Scalr told about the multiplatform python, tulkits and problems associated with them.
The leader of the Python-projects at Rambler & Co, Oleg Churkin (Moscow), explained the requirements for the development process and the infrastructure of the project, so that micro (services) can be tried relatively quickly, efficiently and completely painlessly. Oleg explained what problems the pitonist will have to face when using microservices in production, how to choose the right framework or utilitarian library for a certain type of tasks, and how to ensure fault tolerance and correctly respond to errors from third-party integrations.
Olga Senemova (Moscow) from Tinkoff Bank spoke about how a live chat is organized, what load it can withstand, the introduction of predictive answers to reduce the burden on operators, as well as how the chat is integrated into the bank's infrastructure.
Unit manager at Ingram Micro (ex Parallels) Alexander Hayorov (Moscow) introduced one of the undeservedly obscure Google Cloud Platform services - Application Engine for Python. In addition to the traditional review of the service on the device and functionality, Alexander described the two-year experience of use, pros, cons and subtle points. After this report, you will know exactly how to place your projects in the cloud according to the PaaS concept.
Nadezhda Mirgorodskaya (St. Petersburg), a load testing engineer at Yandex , described how they test the performance of services using Yandex.Tank and BFG, a load generator that supports testing scripts written in Python. I analyzed in detail how BFG can be used to create load and other non-functional tests, how it allowed to easily integrate load testing into the overall development cycle and why BFG became more popular in their teams than jMeter.
Why use the server explicitly if you can without them? In the modern world, we have the opportunity not to cut the cloud into separate machines, but to build entire pipelines and deploy applications without caring where and how they will be executed. Even the code can be written the absolute minimum, but in general you can hardly think about support and development. Sounds like science fiction? Nikolay Markov ( Aligned Research Group , Moscow) understood the report.
Probably only a very talented programmer does not need debugging tools. To change the value of variables, to look step by step in the course of the program execution, it will be hard to sort everything out without Pdb. Artem Malyshev, an independent developer from Nizhny Novgorod, talked about how these tools are designed, what internal Python VM mechanisms involve and what limitations the platform suffers from.
This year, the Open Web Application Security Project (OWASP) has published the next TOP-10 of the most critical web application vulnerabilities. Ivan Tsyganov (St. Petersburg) from Positive Technologies told what kind of TOP-10 this is and what has changed in the last 4 years since the publication of the previous version. He explained what types of vulnerabilities are in the area of ââresponsibility of the developer, and which they can not directly affect. Showed how popular frameworks help develop secure applications, and in what situations the framework will not be able to help.
The head of the Rambler / Top100 development department, Alexander Podsoblyaev (Moscow), described how they restarted the Rambler / Top-100 tools available on the market and their experience in moving from the batch-data architecture to real-time data.
Alexander Mokrov ( Positive Technologies , Nizhny Novgorod) said that the gevent library is under the hood and for which it can be useful. He cited architectural solutions for building asynchronous RPC based on gevent, and spoke about the problems that can be encountered when using it.
Ivan Smaller ( RaRe Technologies ) and Lev Konstantinovsky ( WorldSense ) talked about what theme modeling is all about (what âtopicsâ are and how to extract them from many texts), and how it can be applied to the data that we have (and not only). They told about the Gensim framework and demonstrated its capabilities.
Alexey Kuzmin , a technical leader of the machine learning and data analysis group in the central nervous system, described how memory management works in Python and why your applications consume so much.
Technical director Marilyn System and Aori Alexander Shvets (Moscow) conducted a quiz on the knowledge of his beloved Python. The most advanced received prizes, the majority discovered something new about the language interpreter, and no one was exactly indifferent. Try to answer the questions and you :)
Alexander Kalyuzhny, Oleg Churkin, Alexander Kovalev, Alexey Okhrimenko, Ivan Miniaylenko, Artem Malyshev, Ĺukasz Langa, Anton Patrushev and Alexander Sibiryakov spoke at the Lightning Talks this year.
According to the rules of Disney, it is forbidden to film employees, therefore, the video of the Paul Hildebrant report, unfortunately, is not.
All reports can be viewed on the channel IT-People on YouTube.
Presentations posted on Slideshare .
See photos from the conference in the VKontakte group, on Facebook and on Yandex . Photos .
Sponsors did not allow to miss between reports.
General sponsor - the company Adcombo - gave t-shirts, stickers, treated with soda, beer, and at the end of the day played a hoverboard
The partner of good mood TsIAN was treated to sweets, soda and gave spiners, and in the morning of the second day they saved healing mineral water
Silver sponsor - the company Rambler & Co - brought a live python, staged a chic lounge area in nature and organized a thematic toy "Python". All participants of the game received souvenirs, and the main prize was provided by the company JetBrains - an annual license for PyCharm
Another silver sponsor - DomKlik company - organized a company stand with souvenirs, and at the afterparty they were treated to free beer.
Bronze sponsor - the company Mediascope - organized a booth with a souvenir, and at the entrance to the conference gave participants water
Next are fotochki that show why we love PyCon so much.
One of the great things about PyConRu is that the speakers and participants spend two days together. Any speaker can be caught during the day or chat with him after the party
How long have you been on nature?
One of the favorite parts of the conference is after-party. At PyCon - this is traditionally a bonfire with songs to the guitar, free beer and talk until morning
Photo: Alexander Morozov
Evgeny Slezko
Pycon Russia, as always, on top! Thanks to the organizers of IT-People.ru for the warm welcome, the wonderful atmosphere and the stickers :)
#pyconru
Alexander Morozov:
#pyconrussia These were very cool two days! Thanks to the organizers and speakers for the excellent mood and a lot of new ideas!
Ps: I caught one song around the fire :)
Alexey Okhrimenko:
#pycon Incredibly big and cool :) @ Cronwell Park Yahonty Tarusa
irbisx:
Attended the coolest python conference #pyconru Thank you so much @demis_group for the opportunity to attend such events! #demisgroup #pycon
mindhatter:
Everything was great, the organizers know their stuff. #Pycon #pyconru # yakhontytarus
my_another_prototype:
I returned with PyCon Russia :) Feels as if I visited a children's camp for programmers: fresh air, emotional atmosphere, all day reports, and in the evening a disco and a bonfire
#pyconru #matroyshka #vodka
Big review of Sergei Emelyanov with a description of the impressions of the reports:
âThere were only pleasant impressions from the conference, I heard a lot of useful things for myself and if I answer the questionâ will I go next year? â, I have the answer,â if I have the opportunity, I will. â
Traditional general photo of participants with PyConRu
Join @PyconRu_bot bot, subscribe to our channel , to the Facebook page and you will be the first to hear news about the conference.
See you at PyConRu-2018!
Reports in English
Python ore developer since 2010, Facebook developer, âchronic perfectionist, pianist, dadâ Ĺukasz Langa (California, USA) spoke on âGradual Typing of Production Applicationsâ.
')
One of the developers of PyPy , the main developer of jitviewer - a tool for analyzing the performance of python programs under PyPy, Maciej FijaĹkowski (South Africa, Cape Town) explained âWhy is Python slow?â.
Developer at Venmo , previously at Reddit and HBO, Nina Zakharenko (Portland, USA) talked about common anti-patterns in python programs and showed practical python solutions for improving your code using tools like Decorators, Context Managers, Mixins and Lambdas .
Reports in Russian
The performance of the PyPy interpreter is achieved through specialization, as elsewhere. Alexander Koshkin (St. Petersburg) told Senior Python developer at Positive Technologies what it means and how RPython allows building fast interpreters of arbitrary languages.
Elizaveta Shashkova (St. Petersburg) from JetBrains described how the new interface for computing frames in Python 3.6 works, how it can help you create a quick debugger, and why such a quick debugger could not be created in previous versions of Python. For those who have not yet made the final decision about switching to Python 3.6, this report will provide several additional reasons why this is worth doing.
Scrapy is a good example of a modern asynchronous application. Moreover, it combines in itself every functional: Item pipelines, HTML / CSS selectors, Middleware, support for several network protocols, stop / continue from the moment of stopping, and many others. Alexander Sibiryakov (Czech Republic, Prague) from Scrapinghub described how data processing looks from within Scrapy, the architecture of the module's queue for loading content and other components necessary for its debugging: Scrapy shell, telnet console, memory consumption debugger.
Kirill Borisov (Moscow) from Yandex considered the place of Python in the IoT world, how to use it in communication with various pieces of iron and on what it is launched for the sake of great justice.
In 2014, MicroPython appeared - a Python implementation for microcontrollers. How did you manage to get Python to work on something with only 16 kilobytes of memory? Did you have to cut the language hard for this? (spoiler: no!) What interesting things can be done with glands and MicroPython? Different sensors, accelerometers, LEDs, motors - that's all in the report of the PyConRu program director, the PyCharm Community head at JetBrains, Andrei Vlasov (St. Petersburg).
Mikhail Yumatov (Moscow), head of development at TsIAN , told us what tools there are to monitor the performance of web applications. How to use them to collect detailed information on all operations, quickly detect problems and respond to them. Mikhail also explained what statistical analysis is and how to use it for a detailed study of the performance of parts of the application.
Yandex is a search company. But itâs good to find information not only on the Internet. Inside Yandex there is a huge intranet - a large number of internal services, which also need to be effectively searched. The company is actively using Python, they have a rather interesting infrastructure - there are components on both Django and Tornado. They actively use Celery with different brokers. Interact with internal tools and APIs to reuse the big Search developments. Alexander Koshelev ( Yandex , Moscow) told how all these components interact and what interesting features there are.
Although development in Python has shifted towards the server segment, desktop applications in Python are still relevant. Moreover, with an increase in processor performance, python-based applications have become an opportunity to reduce the financial, human and time costs of releasing desktop versions. And the most interesting point in this is the multiplatform nature of such applications. Igor Novikov (Kharkiv-Lviv) from Scalr told about the multiplatform python, tulkits and problems associated with them.
The leader of the Python-projects at Rambler & Co, Oleg Churkin (Moscow), explained the requirements for the development process and the infrastructure of the project, so that micro (services) can be tried relatively quickly, efficiently and completely painlessly. Oleg explained what problems the pitonist will have to face when using microservices in production, how to choose the right framework or utilitarian library for a certain type of tasks, and how to ensure fault tolerance and correctly respond to errors from third-party integrations.
Olga Senemova (Moscow) from Tinkoff Bank spoke about how a live chat is organized, what load it can withstand, the introduction of predictive answers to reduce the burden on operators, as well as how the chat is integrated into the bank's infrastructure.
Unit manager at Ingram Micro (ex Parallels) Alexander Hayorov (Moscow) introduced one of the undeservedly obscure Google Cloud Platform services - Application Engine for Python. In addition to the traditional review of the service on the device and functionality, Alexander described the two-year experience of use, pros, cons and subtle points. After this report, you will know exactly how to place your projects in the cloud according to the PaaS concept.
Nadezhda Mirgorodskaya (St. Petersburg), a load testing engineer at Yandex , described how they test the performance of services using Yandex.Tank and BFG, a load generator that supports testing scripts written in Python. I analyzed in detail how BFG can be used to create load and other non-functional tests, how it allowed to easily integrate load testing into the overall development cycle and why BFG became more popular in their teams than jMeter.
Why use the server explicitly if you can without them? In the modern world, we have the opportunity not to cut the cloud into separate machines, but to build entire pipelines and deploy applications without caring where and how they will be executed. Even the code can be written the absolute minimum, but in general you can hardly think about support and development. Sounds like science fiction? Nikolay Markov ( Aligned Research Group , Moscow) understood the report.
Probably only a very talented programmer does not need debugging tools. To change the value of variables, to look step by step in the course of the program execution, it will be hard to sort everything out without Pdb. Artem Malyshev, an independent developer from Nizhny Novgorod, talked about how these tools are designed, what internal Python VM mechanisms involve and what limitations the platform suffers from.
This year, the Open Web Application Security Project (OWASP) has published the next TOP-10 of the most critical web application vulnerabilities. Ivan Tsyganov (St. Petersburg) from Positive Technologies told what kind of TOP-10 this is and what has changed in the last 4 years since the publication of the previous version. He explained what types of vulnerabilities are in the area of ââresponsibility of the developer, and which they can not directly affect. Showed how popular frameworks help develop secure applications, and in what situations the framework will not be able to help.
The head of the Rambler / Top100 development department, Alexander Podsoblyaev (Moscow), described how they restarted the Rambler / Top-100 tools available on the market and their experience in moving from the batch-data architecture to real-time data.
Alexander Mokrov ( Positive Technologies , Nizhny Novgorod) said that the gevent library is under the hood and for which it can be useful. He cited architectural solutions for building asynchronous RPC based on gevent, and spoke about the problems that can be encountered when using it.
Ivan Smaller ( RaRe Technologies ) and Lev Konstantinovsky ( WorldSense ) talked about what theme modeling is all about (what âtopicsâ are and how to extract them from many texts), and how it can be applied to the data that we have (and not only). They told about the Gensim framework and demonstrated its capabilities.
Alexey Kuzmin , a technical leader of the machine learning and data analysis group in the central nervous system, described how memory management works in Python and why your applications consume so much.
Technical director Marilyn System and Aori Alexander Shvets (Moscow) conducted a quiz on the knowledge of his beloved Python. The most advanced received prizes, the majority discovered something new about the language interpreter, and no one was exactly indifferent. Try to answer the questions and you :)
Alexander Kalyuzhny, Oleg Churkin, Alexander Kovalev, Alexey Okhrimenko, Ivan Miniaylenko, Artem Malyshev, Ĺukasz Langa, Anton Patrushev and Alexander Sibiryakov spoke at the Lightning Talks this year.
According to the rules of Disney, it is forbidden to film employees, therefore, the video of the Paul Hildebrant report, unfortunately, is not.
Materials
All reports can be viewed on the channel IT-People on YouTube.
Presentations posted on Slideshare .
See photos from the conference in the VKontakte group, on Facebook and on Yandex . Photos .
Breaks and aftepati
Sponsors did not allow to miss between reports.
General sponsor - the company Adcombo - gave t-shirts, stickers, treated with soda, beer, and at the end of the day played a hoverboard
The partner of good mood TsIAN was treated to sweets, soda and gave spiners, and in the morning of the second day they saved healing mineral water
Silver sponsor - the company Rambler & Co - brought a live python, staged a chic lounge area in nature and organized a thematic toy "Python". All participants of the game received souvenirs, and the main prize was provided by the company JetBrains - an annual license for PyCharm
Another silver sponsor - DomKlik company - organized a company stand with souvenirs, and at the afterparty they were treated to free beer.
Bronze sponsor - the company Mediascope - organized a booth with a souvenir, and at the entrance to the conference gave participants water
Next are fotochki that show why we love PyCon so much.
One of the great things about PyConRu is that the speakers and participants spend two days together. Any speaker can be caught during the day or chat with him after the party
How long have you been on nature?
One of the favorite parts of the conference is after-party. At PyCon - this is traditionally a bonfire with songs to the guitar, free beer and talk until morning
Photo: Alexander Morozov
Several reviews from social networks
Evgeny Slezko
Pycon Russia, as always, on top! Thanks to the organizers of IT-People.ru for the warm welcome, the wonderful atmosphere and the stickers :)
#pyconru
Alexander Morozov:
#pyconrussia These were very cool two days! Thanks to the organizers and speakers for the excellent mood and a lot of new ideas!
Ps: I caught one song around the fire :)
Alexey Okhrimenko:
#pycon Incredibly big and cool :) @ Cronwell Park Yahonty Tarusa
irbisx:
Attended the coolest python conference #pyconru Thank you so much @demis_group for the opportunity to attend such events! #demisgroup #pycon
mindhatter:
Everything was great, the organizers know their stuff. #Pycon #pyconru # yakhontytarus
my_another_prototype:
I returned with PyCon Russia :) Feels as if I visited a children's camp for programmers: fresh air, emotional atmosphere, all day reports, and in the evening a disco and a bonfire
#pyconru #matroyshka #vodka
Big review of Sergei Emelyanov with a description of the impressions of the reports:
âThere were only pleasant impressions from the conference, I heard a lot of useful things for myself and if I answer the questionâ will I go next year? â, I have the answer,â if I have the opportunity, I will. â
Traditional general photo of participants with PyConRu
Join @PyconRu_bot bot, subscribe to our channel , to the Facebook page and you will be the first to hear news about the conference.
See you at PyConRu-2018!
Source: https://habr.com/ru/post/335684/
All Articles