Google just kills its projects when they’re not big enough. That being said, Reader was maybe the biggest one Google killed, and caused some backlash.
Fortunately, the guys at Feedly are doing a great job to improve their aggregator for the millions of new users.

But as certain federal agencies blatantly disregard every privacy concern worldwide it is quite glum for Google to cite pesky privacy laws for being unable to comply with them and being forced to shut down an “unproftiable” service.

Hypothetically, let’s say that Walter Olson tells Google that he doesn’t want to appear on Google Reader. (And remember, regulations in Europe often state that mere linking to an article is the same as publishing it.)

Assume Density Duck writes a blog post or article on Mr. Olson attending a Red Cross gala. The RSS feed will give a summary of the article which if Olson is the focus or is in the “lead,” may trip the filter on Google Reader. That’s the easy one.

But what if Density Duck writes a post or article that mentions Olson in a much smaller role. For example, Duck writes about a Red Cross gala in Washington DC and amongst 3o or so names that he saw at the gala is Walter Olson. Now you are asking Google Reader to not only read the RSS feed from the article, but scan the entire article and compare every name to the “delete” database because of the “linking” regulation. That is a huge resource hog.

Finally, assume that Density Duck’s article talks about the Red Cross gala, but doesn’t mention Walter Olson. However, aaaa writes a comment saying that he went to the gala and saw Walter Olson. Now if the RSS feed links to the article and the comment is there, GR is on the hook as well.

I can truly see this being a huge computer and manpower issue to try and comply with the ever expanding Euro regulations. While I am not willing to state the regulations were the only reason GR shut down, I believe the regulations may have contributed to the decision.

Even if the story would be true and the regulation played role in killing reader, I doubt that adding delete would by such a big deal for them, if they would want it.

On the other hand…really? It’s *really* impossible to delete all the data about a user? You *really* design your service such that it puts user-identifiable information in places where you can’t get to it?