> And if you have technical risk, you'd better be an expert at it, before you start. I can't second this one enough. Our startup heavily involved video recordi

> And if you have technical risk, you'd better be an expert at it, before you start. I can't second this one enough. Our startup heavily involved video recording, which I thought was a solved problem. It turned out not to be at all. I ended up learning a lot of really cool stuff about Flash's crazy edge case bugs (I should attach a microphone to a connection with a null hostname to get the activity levels? Sure. I should wait with attaching a camera until I've detected a greater than 0 activityLevel? Sure. I should sample connection speeds and run a regression analysis on that data to find the right bandwidth/quality settings? Cool, but seriously?) and the various flaws of the media servers out there (So I don't have access to the packet stream with Wowza, but in Red5 I need to rewrite the way audio and video packets are interleaved because when uploading high quality video on slow connections Flash delays video packets until the end of stream? No problem.) In the end even though our startup wasn't about video but about interacting using video, we spent 90% of our time on the video part. That's a lot of energy I wish I could've spent on building the business instead.