2 ms·

Response #1: By which you mean, one makes the other possible? In the middle ages, can you imagine the trouble it was to make your own ink, collect and maintai

by neilk 19y ago

Response #1: By which you mean, one makes the other possible? In the middle ages, can you imagine the trouble it was to make your own ink, collect and maintain quills of the appropriate type, train oneself to wield the quill skillfully and not cause inkblots, and finally deal with the fatigue of manipulating such a large object? Nowadays pens are so cheap and easy to use you never even have to think about it. So you move onto other questions, like what to write. Response #2: This statement is idiotic. Computer science is essential to the collection of technologies people call "Web 2.0". The most advanced Web 2.0 sites allow for quick search across enormous databases, and efficient, many-to-many messaging. These problems are hard. Using the appropriate algorithm is essential to scalability. Look at Facebook; they deploy Stanford whiz kids to invent their own infrastructure (hmmm, that sounds familiar) and they're growing faster than anybody else in this space.

amichail 19y ago
It's true that scalability is hard once you get users. But by then you will have VC funding to hire experts who could do that stuff for you. So it's not essential knowledge for entrepreneurs.
- neilk 19y ago
  I checked out your blog. Someone who blogs about Scala doesn't need schooling from me on CS. ;) I think I see what you're doing now. Ok, to tell you the truth -- this is something I've wondered about recently myself. I happen to know about the codebases of several successful Web 2.0 sites post-acquisition. And in many cases, they used the naive solution and got away with it until acquisition. It's difficult to blog about these things in public, but I've circulated similar speculations privately, that maybe all CS knowledge is a huge waste of time for entrepreneurs. Using naive solutions has definitely worked for some startups. But it comes at a fairly heavy price. You are essentially betting everything on being acquired before you get so slow that your own users start hating you. Once acquired, you then get a reprieve from death, and then you spend about one year building your 2.0 version that actually scales decently (all the while fighting fires with the clunky 1.0 version). Meanwhile your users scratch their heads wondering why, once you got all that money, you stopped building any new features. This pattern has been repeated many times for successful web 2.0 startups. Contrast that situation with a company like Google or Facebook. One of Google's main advantages was scaling, from their algorithms to their ability to create vast computing infrastructures. Facebook acquired this sort of talent a little bit later in their history, but they are definitely there today. So when they got offers to buy their company for millions of dollars -- even a billion dollars -- they were able to say no. Because they knew they could grow faster by themselves. Last year Zuckerberg was considered an idiot for turning down a billion dollars. But he knew that F8 was in the pipeline, and he was very smart to say no. Being able to scale means freedom. Ignoring scalability is a high-stakes bet. But either strategy might work, depending on what the startup's business model is, and the talents and inclinations of the founders.

by neilk 19y ago

amichail 19y ago

It's true that scalability is hard once you get users. But by then you will have VC funding to hire experts who could do that stuff for you. So it's not essential knowledge for entrepreneurs.

neilk 19y ago
I checked out your blog. Someone who blogs about Scala doesn't need schooling from me on CS. ;) I think I see what you're doing now. Ok, to tell you the truth -- this is something I've wondered about recently myself. I happen to know about the codebases of several successful Web 2.0 sites post-acquisition. And in many cases, they used the naive solution and got away with it until acquisition. It's difficult to blog about these things in public, but I've circulated similar speculations privately, that maybe all CS knowledge is a huge waste of time for entrepreneurs. Using naive solutions has definitely worked for some startups. But it comes at a fairly heavy price. You are essentially betting everything on being acquired before you get so slow that your own users start hating you. Once acquired, you then get a reprieve from death, and then you spend about one year building your 2.0 version that actually scales decently (all the while fighting fires with the clunky 1.0 version). Meanwhile your users scratch their heads wondering why, once you got all that money, you stopped building any new features. This pattern has been repeated many times for successful web 2.0 startups. Contrast that situation with a company like Google or Facebook. One of Google's main advantages was scaling, from their algorithms to their ability to create vast computing infrastructures. Facebook acquired this sort of talent a little bit later in their history, but they are definitely there today. So when they got offers to buy their company for millions of dollars -- even a billion dollars -- they were able to say no. Because they knew they could grow faster by themselves. Last year Zuckerberg was considered an idiot for turning down a billion dollars. But he knew that F8 was in the pipeline, and he was very smart to say no. Being able to scale means freedom. Ignoring scalability is a high-stakes bet. But either strategy might work, depending on what the startup's business model is, and the talents and inclinations of the founders.