Peter Williams

Everyone else is talking about Google Reader so I think I will too. Personally, I think they got it almost exactly right. The reading style Google Reader encourages (requires?) is exactly how I want to read my subscriptions. I do a pretty close emulation of the style in Rojo but it is much easier in the Google Reader.

Note, though, that I said “almost”. Unfortunately is does not support segmenting my subscriptions. I do not even know now many feeds I subscribe to these days. I do know that is too many to read them all, all of the time. So I have segmented my subscriptions into basically three groups:

1. friends and the comment feed from my blog
2. feeds which I as least scan practically every post (right up until I see “Sox” or “Yankees”, at which point I seriously consider unsubscribing)
3. everything else tagged by major subject area

I read posts by friends and comments on my own blog first. Then I read every thing that is new in the the middle priority group in chronological order. I think order is very important in conversations and the canonical newest first order of most aggregators mostly just annoys me. I only get to the third group when I have time. When I do, choose a subject that seems interesting that day and read the entries in reverse chronological order until I run out of time. In the lower priority bucket the newest first order works because I have already conceded that I am not going to read it all so I cannot really follow the conversations.

As far as I can tell, there is really only one big bucket of posts in Google Reader. That fits their M.O. but does not really work for me. I would never get any work done if I actually looked at all the feeds to which I subscribe every day. You can limit the articles to those with a particular tag which might work except I really think I want to seen if there are any unread entries in the higher priority buckets with out actually going to a different URI (the only easy way to filter on a tag, that I could find, is to go to a URI that looks like <http://www.google.com/reader/lens/user/03449734039335303102/label/friend>).

I suppose I could unsubscribe from the feeds I do not read every day but I do read those lower priority feeds regularly and I do not want to loose them as they often have good posts. Anyway, if I did that then I would just be bored when I am procrastinating. I just occurred to me that the “relevance” ordering provided by the Google Reader might be really useful in this lower priority bucket.

I doubt that Google Reader will ever meet my segmenting/privatization and ordering needs which is quite unfortunate because the article reading model is pretty much perfect as far as I am concerned. So I guess what I am saying is that if anyone knows of an aggregator that works like Google Reader but has multiple buckets and can sort the entries in chronological order, I would really like to know about it.

Just so there is no confusion, Google Reader is really buggy, just like everyone else says, but I tried to ignore that and focus on what appears to function as design.