"A Mars a day helps you work, rest and play." Advertising slogan for a chocolate bar
Congratulations to Mike Kowalchik (and advisor Adam Green) for getting Grazr out the door and open to the public. Dave Winer calls it a 'Javascript-based OPML browser' but I've been calling it a Feed Grazer because its an OPML browser and an RSS reader in one. The whole is more than the sum of the parts.
Along with Optimal Browser and OPod (now Grazr powered) I'm finally spoiled for choice in Feed Grazers so my attention is turning more to grazer fodder - OPML directories and taxonomies. A Taxonomy can be defined as "a system of classification, frequently hierarchical". The top-level classification I've been using for a while now in the Open Irish Directory is -
- Arts, Culture & History
- Business & Economy
- Community, Home & Family
- Computers & Communications
- Entertainment & Shopping
- Government & Law
- Health & Environment
- News & Media
- Sport & Recreation
- Transport & Travel
I didn't re-invent the wheel of course, I merely produced a hybrid structure from several tried and tested taxonomies, including Yahoo! Directory, DMOZ, and niceOne. And a right mess I made of it too! But I couldn't see it until Damien Mulley said -
... all it does is give you more things to click and navigate through and the titles of the nodes become crucial to pique someone's interest. Also there are more than 7 choices which is a no no in UI design terms.
Damien is absolutely right. I'd heard that advice (about 7 choices or less) before but ignored it because if Yahoo!'s professional ontologist thought 15 or more categories were appropriate then surely they were? Not according to Clay Shirky but then again he practically throws out the baby with the bath-water in his ode to folksonomic tagging.
A little extreme perhaps but it did set me wondering about the basic premise behind these online directories - that they can catalog the World Wide Web. Whether or not they can is irrelevant as the World Wide Web is increasingly eclisped by the World LIVE Web. If that seems ridiculous just stop and think about how many hours a day you spend reading new blog posts, email, and IMs and listening to podcasts. Those are all part of the live web, the feedosphere, and they certainly occupy the majority of my waking hours. And it appears to be trending towards 100% as observed by Matt Terenzio in the 'EveryBuddy Manifesto'.
So what does that mean for Web 2.0 directories? Shouldn't they reflect the underlying transience of what they organize. Shirky's core criticism of traditional web directories is that they impose a physical order on a digital world. They essentially ape the categorization scheme of a real world library -
The essence of a book isn't the ideas it contains. The essence of a book is "book." Thinking that library catalogs exist to organize concepts confuses the container for the thing contained.
But..
... there is no shelf. In the digital world, there is no physical constraint that's forcing this kind of organization on us any longer.
Not only that but the World Live Web is no more a library than a road network is a carpark. RSS items are the vehicles of the feedosphere in a constant state of frenetic motion, racing and swerving, overtaking and braking, turning and shooting off in all directions. And these are quantum cars, duplicates existing in a multitude of places at once, made real by observation.
Like Jack Sprat and his wife, OPML and RSS were made for each other. RSS describes action and change. OPML is all about structure and order. RSS is a verb. OPML is a noun. But the verb is what puts the 'live' in World Live Web, and our directories should adapt to reflect this. We must strip them of nouns.
With these thoughts and Damien's criticism's in mind I was wondering how I might re-arrange the top level view of the Open Irish Directory to better reflect the underlying dynamics and mutability of the feedosphere? Can you break down the life of a human being into a simple set of verbs? It was then that the song from the old Mars bar advertisement came into my head (really!) - "A Mars a day helps you work, rest and play". Is it as simple is that? Is that, in essence, all we do? Perhaps it is everything else is a sub-category -
- Work
- Rest
- Play
But what of the overlap and conflict? Think quantum cars! Take the verb 'Shop' for example. Where does that fit in? If you ask my sister its play, my mother calls it rest (incredibly!) and I definitely think its work. But according to the quantum theory of the World Live Web there is no conflict here - we just place the wave/particle in each place at once -
- Work
- Recruit
- Shop
- Travel
- Fly
- Drive
- Sail
- Rest
- Shop
- Watch
- TV
- Cinema
- Eat
- In
- Out
- Listen
- to Music
- to Radio
- Play
- Shop
- For clothes
- For food
- Shop
That's just a five minute brainstorm, a 'write it down regardless' first draft and I'm sure it will require much deeper consideration. But maybe, just maybe, Mars Inc. were onto something....
Technorati Tags: opml, feed grazing
Is there a definition somewhere of what a "feed grazer" is?
Posted by: Dave Winer | March 16, 2006 at 03:31 PM
I guess it all goes back to here Dave -
http://eirepreneur.blogs.com/eirepreneur/2006/01/do_purple_cows_.html
where I was looking for a term that described what an OPML browser + RSS reader does, which is much more than what a feed aggregator does IMHO. Quoting from that post -
"I'm now wondering if it might be better to use the term 'feed grazing' instead [of feed glancing].
Afterall, one definition of 'graze' is "to feed on herbage in a field". Another is "to touch lightly in passing".
And that's exactly what OPML browsing, RSS reading applications like OPod and Taskable allow us to do with RSS feeds. We could even think of a feed as an individual blade of grass. An OPML file (or Reading List) is then a mouthful of grass where the fresh herbage (new RSS items) at the top are consumed.
The important point is that the cow doesn't eat the grass down to the root. She does not subscribe. She does not aggregate. Rather she moves on to another fresh patch for further grazing. She may then return to the first clump at a later date, when there is fresh growth (new RSS items). Of course it will hardly be the exact same cluster of blades so it is in effect a Reading List - a dynamic OPML file.
* Blade of Grass = RSS feed
* Clump of Grass = OPML file (Reading List)
* Cow = Feed Grazer
Posted by: James Corbett | March 16, 2006 at 03:38 PM
World Wide Web is increasingly eclisped by the World LIVE Web.
But the OPML directory is not live. IrishBlogs.ie is alive, it is the live web. OPML is more dead than alive, the content it points to is alive but the effort to get there is far too much work. I see it like a tree and OPML is the trunk and branches. Sure they can grow but the good stuff is in the fruit. Do you really have to crawl all the way to the top to get one apple and then halfway down and out over another branch and get an orange? (this is a magical multi-fruited tree) Would it not be best to have quick access to all the fruits you want? Have a pile and look at it and go "organges yum, I'll keep. Pears, yuk." *poof* All pears are now gone. "Grapes, eugh" *poof* gone too.
I still think we are seeing things in different ways. The idea of a hierarchy is enforcing too much order on something that might not need it. Going from 10 to 7 to 3 only means what is below it gets further and further compartmentalised meaning that a user is going to get quite annoyed having to drill down through 18 layers to eventually click on something that shows them information.
IrishBlogs.ie is a zero click website. You go, you read, you scroll, you read.
I think the OPML structure should be a bit like Digg. All is the default view on the main page. Everything is lobbed in there. You then have different subcategories where everything again in that category is displayed. The more you click the more you filter.
I'm a firm believer of the River of News ideology and I think that if IrishBlogs.ie was actually built like the open directory then we would not have seen the rise of the Blog O'Sphere in the past 12 months. For me it is better to start with a mess and then remove things besides clicking for eternity to find something interesting. It just slows progress.
When a post is shown in the Main river, you should have the option of seeing which tributary it came from and if it came from a stream that built the tributary.
Kind of like:
Post title | Arts and Culture / Music / Irish Music |
With a hover box over "Arts and Culture" saying "see all most recent posts in Arts and Culture" and a hover box over "Music" and Irish Music too. You could have the option of putting a "Do not show posts from this category in the main view"
Posted by: Damien Mulley | March 16, 2006 at 08:23 PM
"Like Jack Sprat and his wife, OPML and RSS were made for each other. RSS describes action and change. OPML is all about structure and order. RSS is a verb. OPML is a noun. But the verb is what puts the 'live' in World Live Web, and our directories should adapt to reflect this. We must strip them of nouns."
This is genius stuff, James, you're a metaphysician.
Posted by: Dermod Moore | March 16, 2006 at 10:13 PM
Damien, I'm just heading out the door now and will respond in more detail later, but thanks for the great feedback. Just to quickly say - you're not thinking ahead.... irisblogs.ie is but ONE feed (and a great feed too) but what will we do as our cup runneth over? What will we do when every sensor, monitoring system, blogject, and electronic gadget in our lives is transmitting data feeds? How can we possibly manage that through aggregation and subscription? How can we possibly organize those if not by hierarchy? Human beings think in hierarchies - its a natural way to organise the world.
Thanks Dermod [I'm off to check dictionary.com to see if that's good or bad ;-)]
Posted by: James Corbett | March 16, 2006 at 11:29 PM
Very interesting post and debate, I was wondering where the Mars reference would sneak back in. Certainly very informative!
Posted by: Ken McGuire | March 17, 2006 at 09:35 AM
IrishBlogs is not one feed it is an aggregation of feeds. There are 1000 different feeds already contributing to this single view.
If you want to take every feed that is available to a person and put them into a structure then yes, you are going to overwhelm them by displaying them all together but having an OPML directory view will just mean hiding interesting feeds on them.
But I had thought we were talking about Irish feeds since you stated in the previous post:
"What I'm trying to show on the other hand is how to make it easier for newbies to access all these great feeds emanating out of Irish blogs, photo collections, podcasts"
We should build on the Irish Blogs or PoTB aggregators and then add in feeds from photos and podcasts. Just have specific tabs for them.
Look at the Open Irish Directory: How many live feeds are there? What is the average click rate to get to a feed? Compare that to IrishBlogs.ie where they are there all mixed together on the main page.
If we want to make something for newbies and introduce them to feeds then I do think that it is better to remove barriers to entry and I think one of these barriers is making them click way down into directories to find one feed. Sure eventually every single node is going to be populated with lots of feeds but I still think every node needs a preview so they can see what is there at a glance. Kind of like your grazing analogy.
With the current OPML structure it is click click click click and then one final click to look at the content of a feed. It feels like feeds are being isolated besides streaming together.
Surely feeds in one node should be displayed together so that a user/grazer can see them and compare them instantly? Like image thumbnails in Google Images, we need feed previews in the OPML structure.
Posted by: Damien Mulley | March 17, 2006 at 01:00 PM
"Human beings think in hierarchies - its a natural way to organise the world."
Since when? Human beings do not think in hierachies. We don't think "bungalow, part of house group, part of structure group, part of " etc
We think in very loose structures and label things or even possibly tag things. The struture of the human brain and the way we consciously and subconsciously think shows that we manage information in a far more interesting and I suppose not fully understood way. Hierachies were used to figure out evolutionary paths and associations. It is not a natural human trait. We may organise things physically in hierachies but we certainly do not think in such a way. If we did we'd be a lot slower in reacting to stimuli.
Posted by: Damien Mulley | March 17, 2006 at 02:47 PM
"IrishBlogs is not one feed it is an aggregation of feeds. There are 1000 different feeds already contributing to this single view."
You're still thinking in terms of the blogosphere and not the feedosphere. Irishblogs.ie *is* a single feed and the address is -
http://www.irishblogs.ie/feed/atom/
Its an aggregation feed, a feed digest, a feed splice but it absolutely *is* a single feed.
"but having an OPML directory view will just mean hiding interesting feeds on them."
In what way are they being hidden? When we walk into a shopping center is every item available for purchase immediately in view? Not at all. The main foyer is like the top level. The shop front names you see are the top level node names. When you choose a shop and step in you have gone down one level. Now you have a choice of aisles - more nodes. When you go to the frozen food aisle you have choosen to do down that into that node in the hierarchy. You walk along and stop at the icecream section. Looks like an interesting subcategory! Open a fridge door and you have gone down another level. Several shelves, choose one. You've gone down another level in the hierarchy.
"We should build on the Irish Blogs or PoTB aggregators and then add in feeds from photos and podcasts. Just have specific tabs for them."
How scalable is that? At what point does the river of news turn into a deluge? Remember Dave Winer, who came up with the River of News paradigm is also the guy behind OPML, the outline/hierarchical format. How many feed can you throw into the river before everything starts whizzing by you at such a rate that it simply becomes a blur? A thosand feeds? 10 thousand? Wouldn't it be like drinking from a firehose?
"How many live feeds are there? What is the average click rate to get to a feed? Compare that to IrishBlogs.ie where they are there all mixed together on the main page."
Do you have an iPod? How many clicks are there before you start a tune? The 'normal' path to playing a tune from the top menu is Music->Album->CD->Track. Why don't apply just forget about the damn menu system and throw every song title there on the first screen? Because that would be totally unscalable! Imagine scrolling down through hundreds of songs to find and play one. The iPod interface is one of the most successful in history precisely because it uses the same hierarchical view and side scrolling mechanism as Grazr uses to present the Open Irish Directory.
"With the current OPML structure it is click click click click and then one final click to look at the content of a feed. It feels like feeds are being isolated besides streaming together."
But looks what happens when you stream everything together! Inundation. Overwhelming noise. It can't scale. IPod users have no prolems clicking their way through a few hierarchies. How long does it really take to click 5 times? And its not as if you can do anything useful with a web browser without a few clicks, is it?
Now what about Podcasts. How can you just 'stream them together'. Doesn't that take us back to linear programming - ie. Radio! Random access is where its at and that's what you get with hierarchical views. Not with stremaing everything into a single feed.
"Like image thumbnails in Google Images, we need feed previews in the OPML structure."
Damien, you ain't seen nothin' yet - there's a feed grazer on the way that will blow your socks off in this regard. Patience ;->
"Since when? Human beings do not think in hierachies. We don't think "bungalow, part of house group, part of structure group, part of ""
Ok, I concede I should have worded that more carefully but if we don't think in hierarchies we naturally organize in hierarchies. Why is that? Go back gain to the supermarket analogy. How many times have you directed people to a product by shop, aisle, shelf? Why is the shop organised like that if it isn't a natural way to do it? Why did human beings come up with that structure? Why did Apple and Microsoft come up with folder and sub folders and sub folders? Do you store every file on your PC in your "My Documents" folder? I think not. If you're anyway organised you probably use at least 4 or 5 levels of sub folder. Think folders aren't a hierarchy? Use WinKey + E to open the Windows Explorer!
Posted by: James Corbett | March 17, 2006 at 03:18 PM
"In what way are they being hidden? When we walk into a shopping center is every item available for purchase immediately in view?"
You are going to build a massive shopping centre which will force people to walk around an area the size of a giant football stadium across ten floors all to find some icecream? Why not put all the icecream all on one shelf as you walk in the door? There's a reason that people will go to the local shop if they want icecream and a reason they go to a shopping centre to get a load of different things. The answer to information overload is not to build DMOZ. If it was, then Google would not have been needed.
"Wouldn't it be like drinking from a firehose?"
Your directory would have someone seeking water and be given one drop at a time. They'd die of thirst.
You have failed to answer my questions. How many feeds in your directory and what are the average number of clicks to get to the content? IrishBlogs.ie has zero clicks. How many clicks to get to the IrishBlogs category in the Open Directory? Why is the IrishBlogs "All" view more popular than their blog directory?
"Do you have an iPod? How many clicks are there before you start a tune? The 'normal' path to playing a tune from the top menu is Music->Album->CD->Track. "
With your directory you're making it Music->Genre->Sub-Genre->Sub-sub-Genre->Album->CD->Track
"Why don't apply just forget about the damn menu system and throw every song title there on the first screen? Because that would be totally unscalable! Imagine scrolling down through hundreds of songs to find and play one. "
There would still be a "damn menu system" James, except it wouldn't be so structured and it would allow you to play all from the very start if you wanted it.
There an option in the main "damn menu" called the random option James. Or maybe it's called shuffle. Shuffle even got its own mp3 player from the people that brought us the "most successful interface in history". It is one of the most popular options in every mp3 player that allows it. You have to wonder why.
"The iPod interface is one of the most successful in history precisely because it uses the same hierarchical view and side scrolling mechanism as Grazr uses to present the Open Irish Directory."
Calling the iPod interface the most successful interface in history and comparing it to the structure of the directory is pure arrogance. I'd question it being the most successful interface ever. It kicked off better interfaces for sure but I think some of the other UIs in mp3 players have superceded it.
As for good interfaces, why did apple bring about software to create cross-genre and genre playlists that monitors what people like? Why did they bring about the shuffle option? The average number of songs played in an iPod is less than 500 songs even when people have 10,000 songs on the device. And this is with *their own* songs and albums that they know about. Can you explain why this is? Not a very successful interface if it was designed to get people playing many songs surely?
A user can organise their categories and directories with the iPod. Your directory foists your structure on them. To use your iPod analogy and compare it to the open directory you'd have someone that likes Indie having to go all the way down a few directories to find Indie, then choose an album and then choose a song.
But the MAJOR MAJOR difference is that we are talking about people who do not know ANY of the bands in the directory or the sub-directory or the sub-sub-direcotry. iPod owners put THEIR music onto THEIR structure. The open directory is in reality all about NEW content being introduced to NEW people. Making them work for the content and not mixing it with other content from the same genre is restricting choice in a way by adding more steps. UI history has shown as you add a step to process you'll lose users. More steps will mean less usage. Very much like has happened with the iPod. 20,000 songs and only 500 get played.
"Now what about Podcasts. How can you just 'stream them together'. Doesn't that take us back to linear programming - ie. Radio! "
Well no because the user has the option to say "don't like this" and that will not appear again. It's listening to radio and then changing it as you listen to it by removing what you don't like and adding in what you do. It is inheriting a radio station and being given access to the playlist. This is how Pandora and last.fm works. It is how the likes of the personalised Yahoo! radio stations work. Radio stations are in effect grazing whereas the open directory is taking niche radio stations and having you move the dial, listen in, move the dial again, listen in etc etc For someone new to listening to music ,and this is what this debate is all about I had thought, then someone needs to get a range even in one genre so they know what they like and as with most people when they start out they'll only know what they like when they hear it.
Seasoned feed professionals might enjoy the ridgidity and extreme compartmentalisation of the feedosphere but the person starting off will just get impatient and move on and enjoy IrishBlogs.ie or just use Google.
"Random access is where its at and that's what you get with hierarchical views. Not with stremaing everything into a single feed."
You poopooed the idea of random earlier on, you can't have it both ways. Hierarchy does not give you randomness. It's an oxymoron. Random is the opposite to structure. You can't randomly select a food item if you are in the baked beans section of Tescos. You are picking something in the same node. That's not random.
"Damien, you ain't seen nothin' yet - there's a feed grazer on the way that will blow your socks off in this regard. Patience"
You're right. What I currently see is nothing and I'm sorry but a vapourware solution is not a solution. What's the point in currently discussing something if the actual answer is somewhere in the future and can't be discussed? I have no time for vapoureware sellers and those that tell me "coming real soon".
Posted by: Damien Mulley | March 17, 2006 at 05:44 PM
"You are going to build a massive shopping centre which will force people to walk around an area the size of a giant football stadium across ten floors all to find some icecream?"
No, its much worse that! The shopping center is one of a number in a city which is in a country which is in a continent which is on the planet which,.... we've got several more layers in the hierarchy, click, click, click...... Like the little old lady said - its turtles all the way down ;->
So if you were running a supermarket you'd use no shelving - just toss everything out onto the floor? Sure, you could paint a grid on the tiles and give people a map - "The icecream is at x4227, y38930 sir, go on, off you go... don't slip on the bananas!" And that's exactly how Google feels for half the searches I do. I can't remember the last time I went to the second page of results, can you? Is it really such a panacea or have we bought into the hype? Even Google recognise that their main search engine is useless for searcing the world LIVE web - thus Google Blog Search. Why was that needed if the the main search engine is so perfect?
DMOZ was and is a perfectly good idea but it was before its time. The issue with it was one of centralized control. With OPML the directory nodes can be (a)distributed internet wide, to the edges and (b)duplicated so that there is not just one version of a 2nd generation DMOZ but infact a multitude. A search engine is a central resource but the network wants to be distributed. And the OPML directory is a P2P directory.
"Your directory would have someone seeking water and be given one drop at a time. They'd die of thirst."
Better than drowning! But you exaggerate surely - why is it one drop at a time? Why is it so difficult for you to click down through a few hierarchies? Is your mouse in need of oil? Do iPod users die of boredom before the music starts?
"You have failed to answer my questions. How many feeds in your directory and what are the average number of clicks to get to the content?"
Millions! Maybe billions! Lets say each node has 10 subnodes. 10 x 10 x 10 x 10 x 10 x 10 = 1 million. Only 6 clicks to get to any one of 1 millions feeds. If your mouse can't handle that it needs to go to the gym.
Now, you've been ignoring my question - how scalable is a River of News? How many blogs can you add to it before it becomes overwhelming?
"Why is the IrishBlogs "All" view more popular than their blog directory?"
And how long do you think that will remain so? When there are 10 thousand Irish blogs will it be so? No way Jose! Then the directory will be used! Why did they build a directory if there is no use for it?
"Shuffle even got its own mp3 player from the people that brought us the "most successful interface in history"
You control all the music that goes onto your iPod don't you? You'd hardly want to listen to a track if you didn't put it on there in the first place! But do you also control what blogs go into IrishBlogs.ie? Er, nope! Fine now but what happens as volume goes sky high and signal to noise goes down?
"Calling the iPod interface the most successful interface in history and comparing it to the structure of the directory is pure arrogance."
Hey, I'm the one who told Apple to go screw themselves a few days ago so don't think I give that praise lightly. Have you heard of Russell Beattie? An employee of the Yahoo! mobile division who is probably the most authoratative blogger on the subject of mobile. Here's what he had to say -
####
"There’s not a lot of people who have problems or get lost using the iPod’s hierarchical user interface. Instead of navigating buttons, menus, tabs, and sections, you simply move from screen to screen in a clean, easy to remember way. It’s genius. We need to think outside the current pardigms of user interface design and figure out what makes the most sense for users to learn. They need one and only one way of navigating the hierarchy of options available to them on a mobile computing device, not several. They’ll learn something once, and apply it to many places.
What’s my point? That we need to do like Apple did with the iPod and review how our UIs work. We need less widgets, not more. We need more than simplicity, we need consistency. And since *all* data is a hierarchy, using that as a base for all UI elements would be a good thing. Teach a newbie: “This is how a hierarchy works. Now, anytime you need to find or edit information - whether it’s the MP3 you want to play or the settings on your phone, now you’ll know how.”
#####
Now Russ can get rather carried away with this mobile lark but he really does know a thing or two. I'd listen to him!
"A user can organise their categories and directories with the iPod. Your directory foists your structure on them."
Again you're only concentrating on the short term. There is a Feed Grazer on the way that lets one mix and match the directory nodes of others. Just like we pick and choose what feeds to syndicate in our aggregators today we'll pick and choose what OPML nodes to place in our hierarchical trees tomorrow. Think of it like grafting the branches of tree. Don't like the one of the branches on my tree? Then just go grab yourself one from some other tree and graft it. We have to start somewhere with Feed Grazing and the directory structure I'm 'foisting' on you is an attempt to bootstrap this new paradigm by encouraging others to build their own nodes, which they are thankfully doing. By the way, doesn't IrishBlogs.ie foist blogs on you?
"But the MAJOR MAJOR difference is that we are talking about people who do not know ANY of the bands in the directory or the sub-directory or the sub-sub-direcotry. iPod owners put THEIR music onto THEIR structure."
And as I said, so too will Feed Grazer users reassemble their own hierarchical structures, building them like lego blocks to suit their own needs.
"Well no because the user has the option to say "don't like this" and that will not appear again."
Huh? In IrishBlogs.ie? Where do you have the control to decide that?
"Radio stations are in effect grazing whereas the open directory is taking niche radio stations and having you move the dial, listen in, move the dial again, listen in etc etc"
You're getting all mixed up! Radio is linear, like a River of News. A hierarchy is random access! So if you don't like what you're hearing on one channel you can just push a button and try the next station. That's the kind of control we all like.
"Seasoned feed professionals might enjoy the ridgidity and extreme compartmentalisation of the feedosphere"
There is no rigidity in our feed aggreators - we subscribe to what we like. Neither willl there be rigidity in Feed Grazing - we reassemble our own trees through node grafting.
"Hierarchy does not give you randomness. It's an oxymoron. Random is the opposite to structure."
You're confusing randomness with random ACCESS!!! Read it again. Your River of News is like a cassette tape where I must fast forward and rewind laboriously through stuff I'm not interested in. An OPML tree is like an MP3 player where I can directly access what I want to consume.
"What I currently see is nothing and I'm sorry but a vapourware solution is not a solution. What's the point in currently discussing something if the actual answer is somewhere in the future and can't be discussed"
How long has the term "River of News" been in use? A few months? That hasn't stopped you adopting it with ferocious speed. The beautiful thing about the blogosphere is that us lowly users can describe to the developers how we want our software to work - its what Adam Curry and Dave Winer christened "Users and Developers partying together" and its a thing to celebrate. Its what jumpstarted the world of podcasting at breakneck speed. Adam Curry spent ages describing what he wanted a developer to build to help him transfer audio from the net to his iPod. In the end he had to do it himself because he couldn't convince anyone. In your view he was wrong to even write about it before he built it. Sheesh, we'd be waiting another 10 years for podcasting if we'd to follow that line of reasoning. All software is vapourware at some time and if I've to bullshit about it to encourage a developer to make it then that's what I'll continue to do. Hey, guess what.... it's already working.... there are three feed grazers (OPod, Grazr and Optimal Browser) already available and a blow your socks of multimedia grazer on the way. Hmm.... vapour is good :)
Posted by: James Corbett | March 17, 2006 at 08:17 PM
Turtles all the way down -
http://members.tripod.com/TheoLarch/turtle.html
;->
Posted by: James Corbett | March 17, 2006 at 08:20 PM
"How long has the term "River of News" been in use? A few months? That hasn't stopped you adopting it with ferocious speed"
I don't really appreciate that comment James.
Posted by: Damien Mulley | March 17, 2006 at 11:25 PM
Why not? It's not an insult! From my point of view its a compliment! Remember I'm all in favour of the "users and developers partying together". *And* I'm a fan of the River of News view for plain old feed aggregation. My point was that you saw a new model of feed aggregation which appealed to you, which you understood intuitely and thus adopted in your everyday terminology and now evangelize readily. Which is terrific! I don't know what's not to appreciate about it? :(
I'm doing the same thing with Feed Grazing for the same reason.
But you still haven't answered my question - how scalable is the River of News? Put a number on it.
Posted by: James Corbett | March 18, 2006 at 08:02 AM