Hacker News Reader

Comment by gverrilla | Hacker News Reader

I don't understand it: these companies have enormous funding, an army of employees, and they can't provide the service reliably (both regarding consistency and safety). What all these coders do all day? I'm asking as an uninformed party of course. But it looks to me like these are companies that build bridges, and their bridges are collapsing all the time.

View on HN

What do you mean? You think all coders are security engineers? All code has dependencies, often dozens of them. You might just need a single vulnerability in a trusted third party library to allow this to happen. These are humans creating these products. I would say that SPECIALLY because of the size of these products, vulnerabilities are inevitable.

> What do you mean? You think all coders are security engineers?

Now, imagine using that argument when a bridge falls down. "What do you mean? You think all the bridge builders were safety engineers? Bridge components rely on different dependencies, often dozens at the same time. You just need one point of failure and boom, it collapses. These are humans creating these bridges. I would say that SPECIALLY because of the size of these bridges, collapses are inevitable."

Comparing bridges to a streaming service is nonsensical, frankly.

People die when bridges collapse. People get mildly inconvenienced if twitch is slow or down.

Accordingly bridge construction takes security & safety much more seriously throughout the project. And it's orders of magnitude more expensive to build and check bridges for safety issues, etc.

But it's still true that each individual contractor in a bridge project is not a bridge integrity engineer. I was replying to GP who said

> What all these coders do all day?

Implying that all coders have to care about everything in their stack. Putting the blame of UX devs or data engineers for the platform to be offline doesn't make any sense. Even hiring more platform engineers doesn't necessarily fix this issue. Like when building a bridge, you avoid this problem with good architecture.

Your analogy would be more accurate if the bridges were constantly being blown up by terrorists. Designing perfectly secure online systems is very hard (if not impossible). Software is very complex, and people are trying to break in constantly. It only takes one person to get lucky or find a vulnerability

Uninformed point of view - I'd be curious the split of that army of employees, since the money isn't in keeping the lights on, it's in sales and feature development. Stability is rarely the forethought unless it's there from day one. It probably takes a lot of money and human hours to keep the streamers engaged, and far less to watch Grafana dashboards.

Because so much of programming is written at a high level, most coders don’t know what the hell they are doing. Maybe the level of abstraction achieved makes it impossible to know.

Edit: One of the reasons is that because there are a very few people (probably) who do the low level stuff, there aren’t enough eyes on the code and a lot of vulnerabilities left in production.

Software companies are maybe incentivised to hire a lot of programmers who can start delivering on day 1. This wouldn’t be possible without the convenience afforded by high level languages.

A company cannot out-engineer bad management nor out-manage bad leadership.

A lot of people who worked in giant tech companies can tell you stories of talent being wasted on tight deadlines for unnecessary projects.

Reply on news.ycombinator.com