I figure you could "build a Steam" in a couple of years, with the right engineers hitting the main features. There's very little magic at the technology level, and you can make life simpler and forget about minor things like the hardware survey or the pretty graphs. I'm not saying this is trivial, but it's definitely doable.
This is a far different statement than "You can build something and compete with Steam in a couple of years". Most of the really hard problems are not technical. Success ain't gonna happen without a bunch of pain, sweat, and strategic stumbles on the part of the competition.
Steam was built since I was in FUCKING high school. Im old now, well over 30.
Apples, and blueberries.
Bluebarry, Drewbarry, tomato, ToMaHtoH.
Fuck their stupid ass streaming code, it’s a giant crud app, only their devops team can take credit for scaling, everyone else is not worth a shit, sorry, thats life, I gotta Leetcode too, and ur code isn’t worth me reading it, leaked or not).
A lot of the secret sauce of such things are not that secret but just take a lot of work.
Building and maintaining infrastructure simply takes a lot of people, time, relationships and whatnot.
They get good at it over time which I guess could consider some secret sauce but there isn't like some secret code that makes the whole thing way better that now you'll see tons of competitors.
"Building apps is easy as long as you don't have millions of users. For that you have to actually think about bottlenecks, the larger architecture etc."
(I agree with that)
What I wanted to express is that lots of engineers I personally know instead say
"Building apps involves thinking about every bottleneck in advance and optimizing for every possible user scenario and a global user base, regardless if the number of users is only ~100."
"Building apps involves thinking about every bottleneck in advance and optimizing for every possible user scenario and a global user base, regardless if the number of users is only ~100."
I would advocate the exact opposite. If you need to scale to X users focus on making a great platform for X users, even if it’s only 100. If you try to over-engineer instead you’ll prematurely optimize and will make poor decisions that’ll come back to bite you when you actually DO need the scale and the requirements change.
I’m so misread, Twitch is a lot of luck, so is all of these companies. Show me the the source code for luck. I don’t give a fuck if you leaked a video streaming crud app code lol.
You're missing the _hard work_ part. Sure there's always an element of "luck" in any story of success, but that mostly has to do with timing, and is much less weighted than the perseverance and hard work of the people building it.
Twitch is a full-featured, very mature application with many moving parts outside of just the video streaming, and building all those parts took an incredible amount of time and effort.
It’s just luck. I mean, if I was a storyteller, what story would I have to tell if there was no story.
They hit.
It’s sort of like we all hold Golden dice, so we marveled, by our own eyes, at the gold.
Dealer: You rolling those?
Us: no, it’s gold.
They fucking risked it. It’s not a engineering feat, we’re all a bunch of pussies.
Twitch is easiest site to build, you might as well show me a todo app (which will be sieged and dismantled), scale is solved, we will eat your applications, the barbarians.
You don't need a genius. You need a few good people, and a lot of hands. I think the best way to look at things like Twitch is to compare them to cathedrals, bridges, things like that. You might be able to have the idea and sketch the plans by yourself, but it's physically impossible to build it yourself.
Like all things web, the problem is scaling the platform and moderation/security. It wouldn't be hard to build a toy Twitch clone no. But it takes tons of people and money to scale it / secure it. And even with all the security, they still got hacked...
This reminds me of the Albertsons guy on Blind who inadvertently created a meme when he said that Facebook could be rewritten with a small cluster of Oracle dbs. The meme is that Albertsons people are so elite, they work and think in a higher level of existence, way above the scalability bs us commoners are accustomed to.
This is a far different statement than "You can build something and compete with Steam in a couple of years". Most of the really hard problems are not technical. Success ain't gonna happen without a bunch of pain, sweat, and strategic stumbles on the part of the competition.