I toyed with ideas such as this in the past. One incarnation is at http://www.dotnetclipper.com/

I previously has a version that anyone could use to set up clips.

Problems I saw were…

- Copyright. Who owns the data? Copyright owners can object.
- Complexity. My experiments have shown that what can be easy for us tech people to set up can be impossible for the average user. i.e. Most people can only cope with pre-created source streams.
- Understanding. If it’s so new and there’s nothing like it most people just don’t understand what it is.
- Coping with source changes. Source data formats/URLs/content formats change with time and over the longer term it’s difficult to keep sourcestreams working.

Simon

Thanks for your response Simon. You have good points that need to be addressed:

- Copyright. Who owns the data? Copyright owners can object.

In the end, sourcestreams really isn’t any different than, say, dashboard widgets for OSX, or other web clipping apps. The destination and use are what will be important to copyright holders. So if someone uses sourcestreams to publish NFL videos to a ton of people without them subscribing, Sourcestreams will need a way to handle that complaint. But if someone is taking copyrighted Flickr photos and making themselves a collage for their phone’s home screen, is this a problem? Maybe displaying them on their home page as their own is a problem, but at that point, shouldn’t it be handled the same way as if they copied the image outright from Flickr? Sourcestreams really just gives copyright holders another means of cutting these kinds of things off.

I’d love to see a version of this with Creative Commons permissions built in too. An information pipeline that obeys CC rules would quickly become a CC poster child I think.

- Complexity. My experiments have shown that what can be easy for us tech people to set up can be impossible for the average user. i.e. Most people can only cope with pre-created source streams.

Exactly! That’s why you’ll see in the PDF that I have three levels of complexity. The easiest is really just a description of what the pipeline will do and some forms to enter relevant information. It’s these types of straightforward pre-packaged streams that will attract new users – both novice and expert.

If a proper community forms, experts will take pride in their streams and make them easy to use so they get a lot of users. Make a top 10 list on the front page and experts will bend backwards to get the masses to adopt their cool stream.

- Understanding. If it’s so new and there’s nothing like it most people just don’t understand what it is.

Right again. That’s why the more popular of these pre-packaged streams can be realeased as standalone services. You can brand very useful streams as completely separate, easy to use services and draw in new users. Put a “Made with Sourcestreams” at the bottom of the external site to draw people to the full toolkit.

- Coping with source changes. Source data formats/URLs/content formats change with time and over the longer term it’s difficult to keep sourcestreams working.

A valid concern. But services with true APIs like amazon, AIM, ebay, Flickr… those APIs are part of their bread and butter and shouldn’t change too frequently – especially not in a way that breaks all the 3rd party apps that use them. Sure, the more obscure or hacked-together data source adapters might need maintenance, but if a group of people really want them, the community will take care of maintaining them. We just have to provide a feedback/rating/alert system to let people know when components are broken or unreliable.