4 ms·

An experienced programmer can cobble together a moderately buggy, YouTube-esque site which kind-of-works for whatever video formats that can be converted by the

An experienced programmer can cobble together a moderately buggy, YouTube-esque site which kind-of-works for whatever video formats that can be converted by their local ffmpeg install, and allows a basic commenting system and upvotes/downvotes, running on a handful of servers capable of storing a few terabytes of data. YouTube, on the other hand, stores so much video that they can't actually tell you how much storage they have online at any one time. They have to process 300 hours of newly-uploaded video per minute, which they then have to serve up to millions of simultaneous visitors through a network of data centres with many thousands of servers spread around the world. On top of this, they have to deal with analytics, advertising, account management, copyright compliance, and everything else they have to do, on the huge scale on which they operate. That's not a minor technical achievement.

jacquesm 11y ago
Youtube did not start with the tech infrastructure they have today. They started with that moderately buggy site which kind of worked for those formats that their local ffmpeg could handle without comments and votes and could maybe store a few terabytes of data. Don't look at the situation today (or even 3 months post launch) to see what a youtube like MVP could look like.
- ajdlinux 11y ago
  Well of course that's correct, but it's wrong to take that and then make the conclusion that the GP does - that building a technology business these days doesn't actually involve much technology because everything's so ubiquitous. Scaling anywhere past MVP requires extensive and deep technical development. Any competent web dev might be able to reach something like YouTube's MVP, but it takes a lot more than that to build anything from then on.
  - jacquesm 11y ago
    > Any competent web dev might be able to reach something like YouTube's MVP, but it takes a lot more than that to build anything from then on. That goes for all web start-ups. An MVP can be duct-taped together but for major scaling you'll need to have all your ducks in a row. But that's fine and totally expected. It would be entirely wrong to build your MVP using the tools you'd use to scale through 6 orders of magnitude, your competitor would be able to cycle much quicker in the first 6 months or so which is when you'll be adapting to market, business model and opportunities suggested as feedback from early adopters. If you're going to do that with a fine-tuned crazy scalable solution you'll simply be dead.

jacquesm 11y ago

Youtube did not start with the tech infrastructure they have today. They started with that moderately buggy site which kind of worked for those formats that their local ffmpeg could handle without comments and votes and could maybe store a few terabytes of data. Don't look at the situation today (or even 3 months post launch) to see what a youtube like MVP could look like.

ajdlinux 11y ago
Well of course that's correct, but it's wrong to take that and then make the conclusion that the GP does - that building a technology business these days doesn't actually involve much technology because everything's so ubiquitous. Scaling anywhere past MVP requires extensive and deep technical development. Any competent web dev might be able to reach something like YouTube's MVP, but it takes a lot more than that to build anything from then on.
- jacquesm 11y ago
  > Any competent web dev might be able to reach something like YouTube's MVP, but it takes a lot more than that to build anything from then on. That goes for all web start-ups. An MVP can be duct-taped together but for major scaling you'll need to have all your ducks in a row. But that's fine and totally expected. It would be entirely wrong to build your MVP using the tools you'd use to scale through 6 orders of magnitude, your competitor would be able to cycle much quicker in the first 6 months or so which is when you'll be adapting to market, business model and opportunities suggested as feedback from early adopters. If you're going to do that with a fine-tuned crazy scalable solution you'll simply be dead.