3 ms·
I was once a "fast" developer, like thousands of lines a day. I could knock things out at an amazing pace but they always had problems and were rarely testable.
by donatj 12y ago
I was once a "fast" developer, like thousands of lines a day. I could knock things out at an amazing pace but they always had problems and were rarely testable. That actually worked out OK where I was where we basically built things and ideally never touched them again.
Now, 8 years later I write maybe 50-100 lines a day. I can see all the vectors of things that would go wrong and take the time to mitigate them. I work with "hotshot" young developers who write my old maybe a thousand+ a day. I'm the grumbly old man in code reviews who forces them to break them into multiple smaller pull requests. The difference is my 100 lines will stay in the codebase for years, whereas their thousand will likely be rewritten several times over in the lifetime of my code. There's tradeoffs here, to be certain.
I see so many younger devs thinking just because their code is tested that its "good". This is certainly not the case. There is so often so little consideration of how their code fits into the "big picture". We end up with 5+ "widgets" where one general widget would have sufficed if they'd thought things out ahead of time. Sigh. Shakes Fist. Get off my lawn. Old man grumbles... I'm 28.
- Raphmedia 12y agoHotshot here. Oh, how I wish I could slow down. But how could I? My task must be done in time for the sprint. Everything I do must be logged into the right ticket, I need to time everything I do. The enemy is time, and I must defeat it. I wish I could output better quality code. Right now, I'm working on a project that was finished at 90% and I need to finish it. The code is horrible, but I have no time to fix it. So I'm simply going to hack at it as fast as I can and make my agile manager happy.
- donatj 12y ago- Well it sounds silly but there will be fewer tickets when you slow down. - If you work for a company where you're not free to question the legitimacy of tickets, get out now. - Not all tickets should need to be done by the end of the sprint, but rather what can be done well or what legitimately NEEDs to be done for buisness reasons. I know your manager mught say everything needs to be done but that is a bold face lie. - If you're throwing together garbage just to get the tickets done by the end of the sprint you are doing Agile wrong and just building technical debt rather than a product.
- e28eta 12y agoIt's your (shared) responsibility to determine what done is, and you should have a hand in estimating how long tasks will take. In my opinion, you as the engineer are ultimately the one who decides what quality you will accept. Instead of saying/thinking "it's done, but I'm not happy with the code quality yet", just say "it's not done yet". I also firmly believe you should review all code before you commit it (would you skip proof-reading a paper before turning it in?), and that's an excellent time to spot easy improvements, places that need documentation, code that doesn't make sense, etc.
- raintrees 12y agoThis conversation reminds me of an electronic relay assembly line I once observed over a period of time. The workers hand-soldered relays, their numbers were being measured/judged, so many cold welds were shipped. QA sent them to soldering classes, where temperatures and flowing of solder were discussed/practiced. They came back with nicely soldered joints, and production numbers dropped to the floor. Management came back in, exhorting higher numbers, so out came the cold welds again to make the numbers.