Krzysztof Kowalczyk's Weblog: Aggregation is not enough.The way you define problem determines how you'll approach solving it. Sometimes that makes a big difference. Let me give you an example: writing RSS aggregators is now en vogue (I see a new one every day). But those programs never go out beyond solving "I want easily aggregate RSS feeds". But is getting RSS feeds really a core problem? No. The problem is bigger (and less well defined): we want to get new information on topics of interest to us. RSS feeds are a partial solution, but it has weaknesses. You have to actively look for feeds that match your interests. RSS feeds usually cover more topics than you're really interested in but the burden of filtering uninteresting news is on you. What would happen if people tried to write "news gathering/filtering software" instead of "RSS feeds readers"? In my opinion we would get better software.

Practical tip: before solving a problem think a bit to find out if there isn't a larger problem wanting to be solved.

Partially I agree, this kind of application can still be significantly improved, but I'm not expecting any breakthrough in the field of automatically filtering relevant contents, and I think that news aggregators offer a huge advantage today.

Just look at what happened with search engines. Since the very beginning all search engines companies tried to develop some kind of "smart" technology that would somehow filter contents in order to distinguish relevant sites and pages from the vast sea of information. Maybe some of them succeeded, but the Google came around.

As we all know, Google ranking system is based on how many links are pointing to a specific page.

What is relevant is that most of those links are made by people. So what Google is doing is actually harvesting people minds to figure out what is relevant.

I think that news aggregators are good today, especially coupled with blogging tools, because they allow users to quickly scan contents (not necessarily news) and let them decide what is relevant and what isn't.

Before news aggregators, every one of us had a set of bookmarks of web site that we were scanning on regular basis. Every day or every hour we were going to these sites and we had developed our own set of "mind filters" that allowed us to quickly pick what was relevant to us.

With news aggregators, instead of having to scan all these pages, we have something that allow us to only check one page, where all contents are displayed in a way that is comfortable to us, and still use the same "mind filter" to pick up what is interesting to us.

If the news aggregator is coupled with a blogging tool, we can then almost instantly share the news bit we found with others, adding some value with a comment.

News aggregators can still be improved letting users better organize their feeds, for example separating contents coming from co-workers from contents coming from general news sites.

Then, we should find more ways to flow more contents trough our news aggregators. As an example, all updates in our intranet-based calendaring application are rendered to an RSS file, so each morning I find updates about who's where in my company right next to my New York Times news.

But still, I want to filter my contents myself, I don't want some piece of software to decide for me what is relevant and what isn't, especially because most of the times I don't know if something is relevant or not until I see it. Of course, I'm still free to decide which feeds are flowing trough my aggregators.