Hacker News Reader

Comment by DeathArrow | Hacker News Reader

I think that you should assess costs/benefits. If you're a small startup and you can only afford to pay five programmers for a year, that means you will concentrate on coming up with a MVP as fast as possible, even if that means having to do a complete rewrite soon. Otherwise, you won't survive.

If you are a big corporation and can afford more resources, I guess it pays off to think about code quality from the start.

If you are doing some project for fun or as a POC, it doesn't pay much to spend time on writing quality code.

I think that writing modular code is paramount. That way you can replace it little by little if needed.

View on HN

Modular code is quality code. Separation of concerns and ability to rewrite parts of it independently is paramount to whatever else code quality metrics your team decides to use.

Yes! You pretty much sum up my own thoughts on this subject too.

I'd add that for doing projects for fun or POC, it still pays to pull in something like TypeScript if it's a JavaScript project, even if you're not overly fussy with the typings -- the benefits will outweigh the drawbacks, and you never know when a POC will turn into something bigger.

Choose some sensible tools with a good balance of structure and pragmatism and write some damn code. Keep it modular. Don't spend weeks obsessing and agonising.

Personally, when I’m writing for fun, I especially try writing quality code. There, I can reason about it, learn it and do it without management breathing down my neck due to business constraints like time and money.

If you only have 5 devs, you can't afford to deal with bugs and crisis all the time.

I'd argue code quality matters a lot more than in a company with already a lot of inertia, for which a week or two of delays on a new feature matters a lot less.

Yeah how many stories have we heard about startups that were killed when everything crashed just as they were getting popular?

I don't think you need super high quality code, but everyone uses "we don't have the time or resources!" as an excuse to completely ignore code quality.

Putting a reasonable amount of effort into code quality isn't a lot of effort anyway and pays for itself in a very short time. Like a year or two.

> you will concentrate on coming up with a MVP as fast as possible

You better hurry up too, because you have a problem if you don’t finish before your debt catched up to you.

Reply on news.ycombinator.com