Just because there might be a bubble doesn't mean that all startups will autofail. However, if I ran a startup with ads as my main source of revenue, I would be worried as advertising budgets are the first thing to be axed in a crisis.

For entrepreneurs crashes are the best times to innovate and launch new ventures because while everyone else is angsting about cutbacks and surviving just another day, the entrepreneur can launch lean businesses and products AND usually get great deals on people, equipment, office space, services, etc.

Do you mean crashes are the best time? During bubbles everything gets more expensive.

I agree that a bubble doesn't mean all startups with autofail. I would like to learn from the past and understand what worked.

Yes, I meant crashes. Fixed. Thank you.

I seriously doubt that modern democratic-capitalist societies as a whole can avoid bubbles. Individual business owners OTOH can not only survive but also thrive in the aftermath given that some basic conditions are meet.

Advertising. People are already bemoaning the network effect in the lack of content on color - I think maybe the plan is to buy their way into users.

Developers are not cheap.

True and false, assuming color has one tech cofounder. What they need at this point is an Android and IOS client that can take pictures and record the current position of the user then you need a backend system that can handle the massive concurrent upload of pictures and do concurrent geo-spacial lookups. This is by far the most difficult part but MongoDB has a filesystem component and there is a plugin for Nginx that allows you to serve files directly from it.

The geospacial indexing can be accomplished by creating a 2-dimentional index of picture buckets where each bucket contains the pictures taken within a small geographical area (say 2000x2000 meters) then all you only have to iterate over the pictures in the buckets that are within the current distance from your location. The clients can easily sort their images before their are uploaded.

It may sound complicated but games do that kind of thing all the time, and often with bigger constraints (skipping a picture or two isn't going to be a big deal).

So yeah devs are not cheap, but a technical co-founder should be able to develop this.

but could you make Color.com in a weekend?

The android client capable of taking pictures and communicating with the service? Yes (minus the graphics, I draw like a retarded goat).

The service behind? No, but I highly doubt they have been spending only a weekend on it.

In two months? Properly, though it assumes that there is nothing more to it than has been described so far.