Comment by noduerme | Hacker News Reader

Comment by noduerme | Hacker News Reader

Parent Full thread

I think anyone who starts to get serious about writing software quickly ends up chafing at the constraints imposed by the free-to-paid pipelines. You get to a point where you're paying $500/mo for some project and you realize, well, I might not migrate that thing but I can definitely do it a lot cheaper for the next one if I learn some new chops. And when you face a client you can tell them how much you'll save them (or you can pocket half of it). The easiest hosted solutions are always good for trying things out, but would I commit a client's software to being backed by one? I don't even know if they'll be around. At least if I know my own infrastructure I know I can chill without being updated without warning or someone going out of business.

OSS in the end is not less technical debt, but it's less urgent and less centralized and better to monitor. Less volatile than any end to end solution.

This is probably harsh to say, but if a $500/mo expense is what breaks the bank, you can't seriously claim to be operating commercially successful software. Even small businesses spend $XX,000 a month on labor and transaction fees and vendors and rent...etc. If the option is $500/mo for some other guy's work or $3,000/mo to hire a new developer to work on recreating the already existing system full-time, the choice is pretty clear.

Small businesses are definitely not casual about 500$ mo. It is also unlikely you need to rewrite small business shop entirely to move it or that it would take too long time.

I'm not doing commercial software for end users. My projects are usually in-house apps for midsize businesses. The largest one is a local hotel chain whose entire server infrastructure I've gotten to under $400/mo. That serves their employee-facing apps (front desk and back-of-house), their reservations system (~1k bookings per day), their website (~100k uniques per month), email systems and training software. If I told them it was going to cost $3k a month, they wouldn't balk. But they've compensated me well for being able to shave it down considerably over time. It's a great company that appreciates owning their own code. If I can write software once that they would otherwise have to pay a subscription for, they'll happily shell out $100k for their own version if it saves them $2k/mo in SaS. Much of what I've saved them over time has gone into my own bank account, and they've ended up with a pretty amazing ecosystem. But also, it has to do with hard lessons I've learned from not relying too much on SaS platforms. So many have come and gone since I started writing software. In the end each one becomes a single point of failure.

Reply on news.ycombinator.com