F5: PDD | Books | Quality of Code | Requirements Engineeing | Career

Programmer's Philosophy in Q&A Format (Every Friday at 18:00 UTC+3)

0:00 Start
2:55 PDD (Puzzle Driven Development) methodology.
7:16 Can PDD be used by big corporations to make developers work a lot?
11:00 What if you go back in time, what would you like to change personally for/in yourself?
14:14 OpenAI has presented capability to create custom GPTs. Are such AI-agencies our future?
20:01 When is going to be the next book and what kind of book it's going to be?
21:23 What is your thought about "code freeze" stages in project management?
24:10 Product owner. Is there a place for such person?
27:45 Can the modern AI (neural networks) help us to create linters of the next level?
30:48 How in object-oriented style you can make implement routing, which through regular expressions calls the proper method of the controller by the URL?
32:03 We missed the meetups in Moscow.
32:50 Any books that changed your life except technical books?
33:07 What do you think about Rust language? Which advantages, disadvantages it has?
36:03 What do you do if as a project manager you will see that a full-time lead project developer works less than eight hours in your project in a remote environment?
38:43 You're saying that there is a customer on the other side who gives money. But what if there is no such person and there is a big company with a lot of investors who don't really into tech project?
42:22 In the coming decade demand for software engineers is going to diminish, thanks to AI. What would you do?
45:48 What would you do if you couldn't be a programmer?
46:16 What motivates you in your activity, in your fight for clean code in the world of deadlines and the dirty code?
48:49 What do you think about what Gosling said that if he would be designing Java again, then he would not make classes there?
51:25 Is it worse to stay on a job that I don't like and where I'm underpaid in order to show in my CV that I'm not a job hopper?
54:28 Do you like Clojure approach to simplifying developers work?
56:48 With AI getting better, are we going to care less and less about code quality?
57:46 What skills will be in the mainstream in the near future?

[YaraslauSauchanka]

PDD sounds really great, but there're some obstacles. First, most of the programmers are bad at decomposition (me too sometimes😢) and won't be able to do tasks bit by bit. Second, what about the time the programmer needs to find a solution? Sometimes it's a good approach to create a POC, to look for existing solutions, etc. And the last, a codebase have to be written in a way which supports this methodology. At least it have to be built with small objects/functions, have low coupling, and have good test coverage. None of these is true for the most projects I have worked with.

[svaleryutube]

About approach (4:00-7:00) when tasks are committed with incomplete state after 1h or 0.5h : what if the task is really not easy and to complete the task it could take few days or even more. Should programmer commit incomplete task each time after an hour (write TODO each time before that) and take it back (coz he knows the best how to continue the task) to proceed progress ? Does not look so efficient...