Hacker News new | past | comments | ask | show | jobs | submit login

it's the write-only culture of software development.

People only care about how long it takes to write something from scratch. That dictates everything, from which frameworks or languages are "hot", to which methodology people use, how projects are managed, and....who gets hired.

There isn't a big argument that someone straight out of college with low experience who happens to know the latest new trendy thing will probably pump out a semi-functional MVP faster than someone more experienced. More energy, more willing to work late hours at home because of less responsabilities, but also less experience leading to less time spent trying to think about real problems that could come down the road and how to prevent them.

That lets someone code really, REALLY quick. It will blow up down the road and then thats just "normal" and goes in the bug queue and tech debt goes out of control.

I was once in a meeting where someone was like "Well, of course this code is buggy and sucks: Its at least a year old!!!"

I was floored.




This.

Many SF tech startups overtly believe that actually planning your code before you write it is just, well, not very "agile" -- that's old school, that's waterfall, proven to be deficient, that's not how we do things in the new world. Planning an architecture before coding it?? Why, you may as well be wearing a pocket protector and hating on women and minorities from some massive IBM cubicle farm in the suburbs while chainsmoking and punching your code into paper cards.

In the new world, you break things and move fast -- when you're assigned a massive new project, you just sit down and start typing. When your unplanned, non-architected code turns out to have serious structural issues that make it unextensible and unmaintainable, why, you just layer another level of crap code on top of that to plaster over the cracks, and another on top of that, and so on.

I wonder how much of this has to do with the churn-and-burn VC culture, where they fund hundreds of startups with just enough money to build an MVP so they can "test for market fit," then burn 99 of them and move on to the next batch.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: