Wrapping Up

[00:00:00]
>> Jem Young: That was fun. We still have more to go, but how was it? System design. Easy, harder than you thought? This was harder than I thought. But did it make sense in the end though? Yeah, this probably is not optimal. We're probably missing something in the way, but it does work. This would work. There's probably some edge cases here we're forgetting, but like it's okay. The thing with the thing to remember here is how did we get here?

[00:00:30]
We talked it out, we wrote down our assumptions, we built it step-by-step. We didn't try to over optimize and say like, oh, let's create a database here and all this. We said cool, video process service, that doesn't work, let's create more services. And that's the thing you have to remember, create that high-level architecture first and then get into more of the details. Otherwise your temptation is like, oh, I know this pattern.

[00:00:52]
Let me just jump right into it. And then you forget something and it doesn't work anymore. So yeah, take it slowly, step-by-step. Reason out your thoughts, it doesn't matter if they're right or wrong, directionally correct. As long as you have a vague idea, I think this is probably good enough. Hey, we made it to the course. Oh, what did we learn? We learned a lot. We learned a lot. We learned about trade-offs.

[00:01:22]
We learned about thinking about, hey, you can either be consistent or available and you can't be both. It's really close, we can't do both. We learned about functional and nonfunctional requirements, the importance of asking these questions, and one thing we've made from where we started to where we got now, we're much sharper about those questions. I can just see it. Hey, what's the problem we're trying to solve?

[00:01:45]
How do we make our estimations? Why do estimations matter? How accurate do we need to be? We covered every technology just about that you need to build anything of any scale. Async workflows, message brokers, load balancers, reverse proxies, all the flavors of databases, we covered a lot of ground here, literally, you know, a lifetime's worth of designing systems, and you can go deep into any one of these.

[00:02:14]
But we don't need to, we have approximation, we know enough to be dangerous, we know enough to reason our way through any problem now. So congratulations, you made it to the end. Thanks for sticking around.