This is an idea that’s been buzzing around my head for a while now and Stef’s Inbox Victory post (where he got his email down from thousands of unreads to zero in one day) has prompted me to try and turn into a method for dealing with RSS feeds. Or at least start the process.
When I started subscribing to blogs there wasn’t, to be frank, a whole lot to track. There’d be a few for each of my interest areas and a smattering in or about Birmingham where I live. In the last couple of years though it seems like blogging, or at least publishing online and providing an RSS feed, has exploded. Really exploded. When I started paying attention to local blogs you could count them on a couple of hands. Now there are hundreds. Check out the sidebar on BiNS, and that’s just the ones Jon thinks are worth sharing. It’s become apparent that unless it’s your job to do so reading all of these is impossible.
Email has also become a bit of an issue for me, especially in the case of circulars and the like. While most people have been good about removing me form their mailing lists and adding createdinbirmingham [at] gmail.com (hint, if you haven’t already) I still get a fair amount of arts-related email that, frankly, isn’t really of interest to me. But there are also email notices that are kinda interesting, the odd piece of serendipity where I’m hovering over the delete button and pause, cocking my head to a slight angle and going “hmmm…”. So it’s not a hard and fast rule and in many cases I’m reluctant to ask them to stop.
Similarly with the blogs. While there are those blogs I really want read every post from and those I can easily do without (and have done without now) the vast majority I just like having around just in case. Take Laura Careless. I started following her for CiB when she was doing a lot of gig listings and reviews and stuck with her as she moved on to more eclectic blogging (always a wise move, I find). Lately she’s been writing a lot about homemade clothes which, frankly, bore me senseless. If I was being all ruthless about my blog following I’d unsubscribe, but I don’t because I know that eventually our interests will cross and she’ll write something invaluable, as she did with the homemade peanut butter post.
The same goes for the vast majority of the blogs I follow. I might not be interested in most of the things they write but I like their character and writing style and occasionally they inspire me. The problem is finding a way to get to those moments of serendipity without having to spend hours ploughing through the rest.
Unfortunately RSS readers aren’t designed with this middle ground in mind. They have quite a reproving attitude to unread blog posts. Google Reader goes so far as to not even bother telling you how many unreads you have. Mine has been at 1,000+ for months now and it’s like a passive aggressive sneer every time I open it. “You’re so behind that I can’t even begin to quantify how much you need to do to catch up” says Google and since I don’t know whether it’s 1,001 or 1,000,000,001 posts I assume the worst and just never get around to it. Especially as I know that the vast majority of them won’t be of interest to me.
This worried me for a while, but then Twitter came along. Twitter has become my first port of call for Internet stuff. It tells me what my friends and people I’m interested in are up to, it keeps me in the news loop water-cooler style and it provides enough links to cool things to keep me distracted. And it does all this very ambiently. In fact, with the current technical traumas where you often can only get the 20 most recent messages, it enforces ambiance. If you miss it on Twitter then it’s gone and if you haven’t got time to plough through all the tweets then it doesn’t matter. The attitude that Twitter encourages is a very healthy one regarding the Internet. It’s utterly unimportant, unless it’s important, and if it’s important then you’ll find out about it eventually thanks to the safety net of the community. So don’t worry about it.
So, I wonder, how can this be applied to RSS feeds? A technical solution might be to create a middle ground, say a Google Blogsearch that only monitors the blogs I’m subscribed to. But rather than giving it keywords to track (something which destroys the whole serendipity thing) it looks for patterns across the blogs. Take this woefully inadequate diagram:
I’m the red circle and all the blogs I follow are the black circles. The posts they write that are of interest to me are coloured in dark grey and the stuff that people I follow share is in light grey. I wonder if it might be possible to approximate this based on an analysis of firstly what I’m writing about on my blogs and secondly where my reading pool overlaps. This filter could then be applied to the incoming items in my feed reader and create a serendipity pool of alleged relevance.
For all I know Google is trying this already. After all if you have Google History switched on it affects your search results in a similar way (something to remember if you see your blogs coming in surprisingly high for certain searches) but the Home page for Reader never seems that useful to me – just a random selection of recent posts. Maybe if I starred stuff more often it might learn from that, I dunno.
But that’s a technical fix that will need the next generation of RSS readers to solve. Right now I need a solution and that solution seems to be to not worry about it.
While I haven’t put this completely into practice my plan looks a little like this.
1) Locate the feeds you really need to follow. Good friends, essential information, work-critical stuff, current projects, online comics, that sort of thing. Give these specific tags so they can be checked methodologically. But be strict. Some good friends aren’t worth following online. (This is a lesson learned from Flickr – most of your friends aren’t good photographers so you don’t need to follow their photos.)
2) Do the same for admin and pure info feeds. eBay alerts, Google news / blog alerts, comment monitoring feeds, event listings. The sort of things you should keep and eye on but don’t have to worry about every minute of the day.
3) Optionally, create an “Important” tag where you put the things you really really really need to follow from 1 and 2.
4) Finally, put everything else in an extremely vague tag or two. I have “Blogs” and a subset of that, “Birmingham”, a hangover from the CiB days. I have no idea how many feeds are in there but at a guess I’d say 4-500. Give or take.
Now, when you come across a blog you think might be worth following add it to the general “Blogs” tag unless you’re really sure it’s an essential read. And feel free to demote stuff. And above all, feel free to unsubscribe.
The next thing you need is a system and for that I’m going to switch to letters.
a) Aim to clear your essential feeds once a day and certainly the Important tag.
b) Admin tags should be cleared every few days and certainly once a week.
c) Everything else should be skimmed. Switch to Expanded View and scroll for a few minutes looking for things that catch your eye, then Star or open them in a tab or whatever system works best. Expect most of the items in your feeds to be of no interest to you just as most of the books on someone’s bookshelf or songs on their hard drive will be. You’re looking for the common ground, those moments of serendipity where you cock your head slightly to one side and go “hmmmm…”
d) Mark everything as read on Sunday. I’m not 100% sure about this. In some ways it doesn’t really matter if those posts from 2 months ago are read or unread. But it might make you feel better about things to no longer have the 1,000+ hovering over you like a stern teacher.
e) Finally, don’t worry about missing stuff. If you miss someone’s writings then it doesn’t matter. If it’s important then you’ll hear about it eventually and the posts are always there if you want to dig through and catch up. (Say, before meeting them for a drink. Not that I do that ever. Oh no.)
Like I saids at the outset, this is me figuring out a plan as much as offering advice. Got any tips of your own for managing the flood while keeping the serendipity?
In the meanwhile I’ll crack on with putting this into practice, which I’ll have to do now I’ve announced it so publicly. Which, of course, was the whole reason for writing this post. I’ll let you know how I get on.
Interesting post Pete, I’m not sure I have either a feed-reading or email strategy — expect not letting things build up. That said, I’m going to write mine down anyway…
I tend not to worry about the “unread items” score as it’s so often photos of cats, or search feeds I monitor. Things I’d like to see, bit I’m not so bothered if I miss them.
In Google Reader I do have some folders that I read intently and won’t miss. Some that I skim and ’star’ for later (mostly long articles and videos that I need to concentrate on).Some that I never read and just periodically “mark as read” — I’m using GReader’s search facility here, so I can find them if I need them.
In this instance I’m thinking the solution could be partly at least technical. Google Reader could kill all the friendfeed type sites if it added a few features some I wished for here.
And I guess you’re right – like twitter it’s best not to think you have to read all the small stuff.
Ah, and the BiNS blogroll is dumped straight from my “Birmingham Blogs” Google Reader folder – one I very much skim-read. Skim-read so you don’t have to ;)
Hmm yes this is very interesting. I use Bloglines Beta and at this moment it has 16080 unread. I am not too worried about it though as I drag any blogs that are essential reading up to the top of my list to check daily (about 25) and then skim read the rest when I need a hit of inspiration, and as they are mostly creative or culinary that is quite easy to do as it is mostly image based. I also advocate a one-in-one-out policy wherein when I add an RSS then I delete an older one that is no longer relevant.
Technorati keyword search does something like a peripheral aggregation, but not as well as I’d like. What would be really nice is a two stage search, where it finds material based on an explicit keyword, then looks for tags or text using a glossary I provide, and ranks the results according to the match rate. I can imagine Delicious having this sort of search function, where it uses a bundle to search the web for other bundles or similar groupings of text.