Teach a man to fish - Feeds Vs Attention

Piaras responded to my post about reading lists by suggesting that it would be better to build a del.icio.us feed for articles he pays attention to than to publish an OPML Reading List. While one benefit of a del.icio.us feed is more granularity the problem is that he'd be spoonfeeding us instead of teaching us what to pay attention to.

On the other hand, by sharing an OPML based Reading List Piaras would be providing an "attention lense" which could be applied to many services going forward. Take Kevin Burton's TailRank for instance. It allows me to build my own Memeorandum and import any OPML list I wish. So, if I was to import Piaras' Reading List I could effectively see the world through his eyes! That's what's exciting about Attention data and for the moment it looks like OPML is going to be at least a starting point for Attention applications.

"Give a man a fish, feed him for a day. Teach a man to fish, feed him for life." By supplying a del.icio.us bookmark feed you're handing people fish. By sharing your OPML attention data (reading list) you're teaching them how to fish.

Comments

[Simon McGarr]

Just to be clear on this- I have no idea what an OPML reading list is or how to make one. A guide for dummies would be helpful, if it was seriously wanted.

[James Corbett]

Simon, good point - I'm going to eat my own dog food (boy, I'm bursting with culinary metaphors today amn't I!) and produce a reading list on the topic of OPML.

A quick pointer though, to a good starting point -

http://www.opmlmanager.com/

wherein you can build and maintain OPML lists free of charge. The homepage also contains a number of useful links.

[Piaras Kelly]

Meh...That would mean my having to maintain a couple of RSS accounts, one for my PR reading and then my personal reading through in. I presume that anyone that would read my what I'm reading would be interested in PR, but they don't want to read all the latest Premiership news as well though. So unless you've got some way of filtering that out then I'd still prefer the del.ici.ous option.

I actually subscribe to one PR blogger's del.ici.ous feed, but not his blog feed so I'm not sure that your system would accurately reflect my respect for him. I'm interested in articles that are of interest to him because I value his opinion, but I weight other PR blogs higher than his.

As it stands I'm happy with all my subscriptions, I prefer finding new blogs through del.ici.ous or links from other blogs. I'm consuming my optimum amount of content via RSS. Other people may like to go down the route you propose, but I'm happy enough the way I am.

[James Corbett]

"Meh...That would mean my having to maintain a couple of RSS accounts"

Well I'm not sure what RSS aggregator you use Piaras but with most it would just be a matter of separating your feeds into groups or folders (or whatever terminology your aggregator uses). My feed list is about 200 so I couldn't manage it if I didn't separate it into groupings.

Looking in my FeedDemon right now I have 9 different channel groups. One for "Irish blogs", one for "OPML/RSS/SSE", one for "Mobile", one for "Entrepreurship", etc, etc. I can choose to export any of those groupings as separate OPML file. In the next release of FeedDemon (and BlogBridges, etc) I could just as easily publish that OPML listing online for sharing, as a Reading List). So, perhaps it would require a little reorganisation if you don't already have separate RSS groupings.

If you're subscribed to another PR blogger's del.icio.us feed but not his blog feed that's no problem - his del.icio.us feed is the RSS feed you add to your reading list, not his blog feed. But you're probably also subscribed to the blog feeds for other PR bloggers? Then they should also be added to a reading list.

The reading list is the aggregation of your attention in a 'standard' format which can be used in other applications (eg. TailRank). But a single (del.icio.us) RSS feed doesn't really do it. It should all become clearer in the next round of aggregators as OPML is adopted as an attention standard (or staring point at least).

We supply our blogs in RSS format because its easy and it ads value for others. We're effectively sharing our *thought stream*. Because that thought stream is in a standard format all sorts of wonderful searching, remixing and 'mashups' can be done, enabling different usage scenarios which even Dave Winer probably never dreamt.

Likewise, aggregators will soon make it as easy to supply our OPML feeds and in so doing we'll be sharing our *attention stream*. Because that stream will be in a standard format we can expect a similar cross polination of of ideas and new efficincies.

[Ed Byrne]

James, I'm with you on OPML - it's got fantastics potential. I think the problem here is, and no offence to you Piaras - is that he just doesn't understand it. Similar thing happens all the time trying to get people using RSS instead of HTML - and once they bite the bullet the results are astounding!

Attention is on my 'hot list' for 2006 - it's just not easy enough yet to let people look at your 'lense' (I like the term lense ... I think squidoo's premise for 'lenses' is excellent), since exporting/importing OPML is still laborious to a point.

Actually, squidoo's idea is very good when I think about it in this context - it's not just a list of RSS feeds - it's a lense you build on a topic (so pr / personal / it) with refernces to RSS feeds, flickr, delicious, things you've written etc.

[Simon McGarr]

Hello again.
I went to look at OPMLManager.com.
I scratched my head. I went away again. I am still none the wiser. Would you like to know the RSS feeds in my Digital Rights folder? I'm afraid I can't tell you that.

Because I don't keep them seperated by topic like that. I keep them seperated by interest. So I have all my searchy ones stuffed down the bottom, I have the ones by people I know up the top and so on. If they say something about Digital rights, I get them twice, but so what?

Still don't know what you want, but am still willing to help you in getting it if it will be of interest to you. Please restate what you'd like in very, very plain language, such as a dummy might follow. Thanks

[Simon McGarr]

Is this it?
http://www.opmlmanager.com/outliner/smcgarr

you've exhausted my abilities now.

[Piaras Kelly]

Okay I think I need to clarify what I mean.

If I were to give a client a reading list, I'd recommend articles rather than authors. The reason for this is simple, I want to pass on information taht would be of specific interest to them.

If on an ongoing basis they wanted to learn more then I understand the benefits of what you're deeming a reading list.

Think of it like a college degree. In your first year the professor is going to recommend specific books to build your knowledge base, but in your final year he'll recommend authors which you can go off and read at your own leisure to expand your knowledge.

So my theory is recommend specific articles to a client (books to a pupil) = del.ici.ous

If they get the swing of things then I'd pass on an OPML file. I don't think I'll be getting to that stage of complicity for some time yet though.

[James Corbett]

Simon, that's it! Great stuff - can I add it into the Open Irish Directory hierarchy? Here's a link to the top level -

http://snipurl.com/jxh1

I can now create a new 'Guide' in my Blogbridge aggregator and set your OPML file as a reading list - they hey presto I'll be subscribed to all the RSS feeds you indexed and the great thing is that my 'Guide' will keep synchronizing automatically with your list as you update it. Thanks!

Piaras I understand your points and the debate we're having is a reflection of the debates ongoing between aggregator developers as regards what level of granularity is required for Attention aggregation.

To use the book analogy, the way I'd look at it, is that we don't generally ask people what paragraphs they'd recommend, just what books (or genre) they read. Neither do we ask what type of scenes they like - just what movies they watch. Neither do I usually probe friends for what tracks are their favourites, just what albums they're listening to. Similarl we don't ask what journalist you follow, but instead what newspaper you read. The assumption is that we won't necessarily resonate with every sample of a media stream but that it nevertheless makes sense to consume them whole.

But of course these are all generalizations and there's no 'one size fits all'. The level of granularity needed is something that can change as demand requires. However, I guess I keep coming back to the fact that OPML is out there now and is therefore an appropriate starting point for Attention.

Thanks for the interesting discussion.