Aggregation Aggravation

As I was upgrading my blog I was looking at the release notes around support for APML (you can see my APML here).

I feel the need to subscribe to many blogs to keep myself abreast of developments across the varying industries and technologies I look at. Adding one or two too many news aggregators can provide an extra level of depth about a subject, but means that blog reading can become difficult. (The same phenomenon occurs in email too of course once you’ve added yourself to too many distribution lists.)

There are further challenges now with personal aggregators such as FriendFeed. FF does mean that I can unsubscribe from a couple of blogs and even certain Twitter streams, but:

• I become saddled with redundant feeds. e.g. I like person x because of their interesting thoughts on subject y, but I don’t really care about their music tastes or photographic interests. I notice that FriendFeedSpy is heading that way. I could of course just create ‘imaginary friends’ with only the services I want, but that sort of defeats the point, and doesn’t inform me of changes and new feeds.
• Many people currently cross-post between one stream and the next leading to redundant information. e.g. person x tweets “I just blogged…”

APML is interesting then in that it may be able to provide the necessary filters to ensure I get the information I’m interested in, but it won’t take into account the data duplication from various sources – that’ll have to be constructed.

I think APML could be useful for comprehension in dynamic organisations. A simple example.

In my team, there are a bunch of us who essentially do the same work (not all of us blog), but if you met us individually you wouldn’t necessarily realise it as we have a lot of competing interests and focus areas. We’re also hired as a team for our breadth of knowledge, and individually for our depth of knowledge. Knowing where interests lie and who’s best at what is a matter of networking. Publishing an APML feed would be useful for our colleagues (in the large) to understand what we know and what we’re interested in. So the automated construction of APML depending on activity (like in this blog) would be very useful. Maybe they should build it in to Xobni.

One step closer to my Human Hive Hypothesis (which needs a more formal definition).