🧘🏼 💞 😲 How we were wrong with monoliths, udalenka, legacy, refusal from REST API and online mitaps 🎅🏼 ☂️ 😈

It’s great when people who tried, made mistakes and found working solutions are not afraid to talk about it. And I really do not want the exchange of knowledge froze during the quarantine period. Therefore, on Saturday March 28th, we are trying to build a virtual and interactive PHP mitap .

To make it pass plus or minus exactly, over the past weekend we learned to feed people pizza remotely, arrange distributed holivars by voice, do not overheat laptops that stream from Youtube, and catch other format bugs. And in parallel, they found their advantages on the Internet: for example, the ability to attract experts from different countries and cities and play the rarest elephant.

What kind of elephant?

Here is such a plush handsome man from the author of the PHP digest Roman pronskiyPronsky will receive the main winner of the quiz - it is planned for the middle of the online meeting . There will be other activities in the program: for example, drawing a ticket for PHP Russia (online in May and offline in September). The winner will be determined not by random, but by knowledge + typing speed.

But the main thing to come for is the reports - and their discussion by voice in Zoom during breaks. To make discussions even more interesting and useful, they will be moderated by those who also have something to say on the topic.

What kind of reports and who will be the first to throw a stone at them ?

Speakers are presented in order of priority. Each hour is one report in the format “35 minutes on air, 25 minutes for discussion in the zoom with the speaker and invited expert”.

11 a.m. Ivan Matveev, Skyeng: about the mistakes that the developer makes on a remote site during the day

What will you tell about?

About how, after working a couple of months in the office, I got to work remotely - but I stayed here. For the tenth year now. I’ll try to tell you about key things that I wouldn’t do in half an hour and share the practices of international and Russian teams in which I worked.

Experience is the sum of our mistakes. Tell me about the main shoals that he made while he was saving his own.

Figachil all night. I had lunch only at 6 pm. He was expelled from the first distributed team for stupid jambs. There were a lot of things: half of the report was about this) But I will share the practices “it was possible”, of course.

Who will moderate the discussion after the talk: VictorTyvikTyschenko, chat admin for remote workers.

12:00 p.m. Anton2ballZhukov, Manychat: about the fight against Legacy. Again)

What will you tell about?

Someone has already heard the previous version of the report on the meeting from Badoo, but there is always room for improvements over Legacy)

I’ll tell you about the sensitive deploy approach, safe refactoring and code restructuring. And it doesn’t matter what we refactor: cut out a relatively new feature that did not pass the A / B test, or a scary procedural code that has grown with commits and has grown into a huge component that cannot be read, but needs to be supported.

Tell me about the main schools that he made while accumulating his experience.

I didn’t populate the queue metrics, got a failure in the message flow ... For a long time then I looked for where they disappear.

Who will moderate the discussion after the report: SergeyseregazhukBeetle, author of the podcast “Between Brackets” .

13:00. AdeleAdelfFayzrakhmanov, Laravel Idea: About CQRS.
And he has 5 2 reasons for this.

What will you tell about?

As project complexity grows, developers often come up with the idea of completely separating code to read data from code to modify data. I will show two reasons for this: infrastructure and business logic. With examples.

Tell me about the main schools that he made while accumulating his experience.

All my shoals are boring and dim. But come to discuss the report, it will suddenly seem different)

Who will moderate the discussion after the report: Dmitry Simushev, author of the report “ Half ORM or How to Stop Being Afraid of CQRS ”.

2 p.m. Julia Nikolaeva, iSpring: on the benefits of a modular monolith

What will you tell about?

How to solve the problems of a monolithic PHP application without switching to a microservice architecture.

Tell us about the main shoals that you and the team completed while accumulating your experience.

They combined the processing of legacy monolith architecture with the introduction of a new framework and the development of a large feature with a tight deadline. Yes, yes, all at once. Do not do this)

Who will moderate the discussion after the talk: Leonid Korsakov, author of the holiv talk about the code for commercial projects.

3 p.m. AntonamorevMorev, WormSoft: On the pros and cons of GraphQL. Again)

What will you tell about?

I will continue the topic of abandoning the REST API in projects with microservices, which was raised at the first PHP Russia. Since then she has overgrown with new details.

Tell me about the main schools that he made while accumulating his experience.

No comment)

Who will moderate the discussion after the report: CyrilSerafimArts Nesmeyanov, evangelist of GraphQL, member of the program committee of PHP Russia.

4 p.m. Mikhail Samarkin, SimbirSoft: DO NOT talk about Psalm, Phan, PHPStan. But…

What will you tell about?

About control over your code: how to keep tests up to date, identify architectural errors, comply with the Single Responsibility Principle, maintain a single code-style in the repository, collect qualitative and quantitative metrics, identify copy-paste ... And I’ll tell you about git-hooks and how help in code control.

Tell me about the main schools that he made while accumulating his experience.

I didn’t run the tests after and at the time of the implementation of the new functionality)

ps “Come” to the online mitap without leaving home.

How we were wrong with monoliths, udalenka, legacy, refusal from REST API and online mitaps