|
|
|
|
|
|
by xytop
3932 days ago
|
|
|
That's analog of http://embed.ly/docs/api/embed API. I've done it in few days. Server can handle 1000s of requests. Does smart things like de-shortening urls, and fetching image sizes based on first 1024 bytes. It doesn't need to download full URL content in order to get oembed info. For HTML pages it fetches just first 50KB of data, for binary streams - just 4KB (although 1KB should be enough too for most of data). Also, guys, poll! Guess what hardware is powering the server? :) You'll be surprised. I'll answer in few hours in the thread! |
|
|