4 ms·

> Now a someone who's purely back-end can actually take software he's written, and just throw it on some EC2 instances and sell it directly to consumer internet

by 8ren 16y ago

> Now a someone who's purely back-end can actually take software he's written, and just throw it on some EC2 instances and sell it directly to consumer internet companies. There's no need to take a job with these companies now and...there's no need to release your software. http://news.ycombinator.com/item?id=1788558 http://news.ycombinator.com/item?id=1788558 That was in cloud vs. opensource... but I hereby humbly propose 8ren's amazingly wonderful Open Service concept: The cloud is the future, not only for the backend, but also components within it - including middleware. Theoretically, arbitrary functionality can be a service - literally, any "function" (in the math sense - or stateful). That is, anything that was previously open source. It might even get to the point where the economics make sense to use an "open service" instead of hosting it yourself, analogous to conventional open source. This would require hosting that is: (1) practically free; (2) very high performance; (3) very high availability; (4) security/privacy mechanisms. We may be closer than you'd think - eg. if EC2 hosted "Open Services" for free, which would naturally be faster for EC2 paid services, making those paid services more attractive. Then the competition needs to match it... Google might do it too (already does, within its app engine). To qualify for this amazing free hosted Open Service? Be open source.

franck 16y ago
You might be interested in Jeff Lindsay's Public Open Source Services idea : http://blogrium.wordpress.com/2009/10/29/public-open-source-services/ http://blogrium.wordpress.com/2009/10/29/public-open-source-... http://poss.gliderlab.com/ http://poss.gliderlab.com/
- 8ren 16y ago
  Thanks, I like his market payment model esp how demand modulates supply - but I believe that just the fact of payment adds admin overhead/transaction costs, and slows adoption - you might as well charge a little more and make it a business. I think my suggested payment model makes Open Services easier to adopt, because it has no admin overhead. The API is there, you call it, done. It is paid for by amazon/google, because it benefits their paying customers - who indirectly pay for it. Note that they would be paying for its hosting anyway, if they deployed that open source themselves. - There would be admin savings from having just one deployment (more efficient for everyone). - And it would be just as fast, if deployed in the same cloud (theoretically - I'm assuming a coarse-grained IPC world, where apps are built by assembling services anyway, which is where we're headed. I'm thinking of it like static method calls in Java, that so get JIT inlined/stitched into one big lump at runtime, even though they are in separate modules (class files)). - Plus, these free Open Services act as an advertisement and enticement to non-EC2 users (or free users) - and once you've started using EC2-hosted Open Services, the easiest step up is of course to move to becoming a paid EC2 yourself. There's a bit of lock-in. So it is in Amazon's interest, too. So it's economically a little bit like advertiser-subsidized TV. And technologically, it's like Google's many free API's (eg. map API), but the case is more compelling for the supplier, because it costs the same to provide, but more directly pulls in paying customers. - I agree there would sometimes be issues with how much processing power it gets (Dos attacks, abuse in general), but this is a complex issue, and I think you'd need to actually run it for a time to discover the real issues and experiment with solutions. But the vast majority of open source projects are not used by anybody - these aren't using any resources (apart from memory, which is close to 0 cost), and so their cost is negligible. - once one hoster provides it, it gives them a competitive advantage, and other hosters need to match it to stay in the game. - over time, with Moore's law, the hoster's cost of providing these Open Service approaches zero. I think it's inevitable for Open Services to begin and spread (or something functionally similar).

by 8ren 16y ago

franck 16y ago

You might be interested in Jeff Lindsay's Public Open Source Services idea : http://blogrium.wordpress.com/2009/10/29/public-open-source-services/ http://blogrium.wordpress.com/2009/10/29/public-open-source-... http://poss.gliderlab.com/ http://poss.gliderlab.com/

8ren 16y ago
Thanks, I like his market payment model esp how demand modulates supply - but I believe that just the fact of payment adds admin overhead/transaction costs, and slows adoption - you might as well charge a little more and make it a business. I think my suggested payment model makes Open Services easier to adopt, because it has no admin overhead. The API is there, you call it, done. It is paid for by amazon/google, because it benefits their paying customers - who indirectly pay for it. Note that they would be paying for its hosting anyway, if they deployed that open source themselves. - There would be admin savings from having just one deployment (more efficient for everyone). - And it would be just as fast, if deployed in the same cloud (theoretically - I'm assuming a coarse-grained IPC world, where apps are built by assembling services anyway, which is where we're headed. I'm thinking of it like static method calls in Java, that so get JIT inlined/stitched into one big lump at runtime, even though they are in separate modules (class files)). - Plus, these free Open Services act as an advertisement and enticement to non-EC2 users (or free users) - and once you've started using EC2-hosted Open Services, the easiest step up is of course to move to becoming a paid EC2 yourself. There's a bit of lock-in. So it is in Amazon's interest, too. So it's economically a little bit like advertiser-subsidized TV. And technologically, it's like Google's many free API's (eg. map API), but the case is more compelling for the supplier, because it costs the same to provide, but more directly pulls in paying customers. - I agree there would sometimes be issues with how much processing power it gets (Dos attacks, abuse in general), but this is a complex issue, and I think you'd need to actually run it for a time to discover the real issues and experiment with solutions. But the vast majority of open source projects are not used by anybody - these aren't using any resources (apart from memory, which is close to 0 cost), and so their cost is negligible. - once one hoster provides it, it gives them a competitive advantage, and other hosters need to match it to stay in the game. - over time, with Moore's law, the hoster's cost of providing these Open Service approaches zero. I think it's inevitable for Open Services to begin and spread (or something functionally similar).