OPML Reading/Listening/Viewing ‘Lists’ and ‘Trees’ = ‘Treedia’

OPML webDave on Scripting News and Mike over on Techcrunch post their views of OPML used for ‘reading lists’.

This is an important discussion and one close my heart. I commented over on Dave’s (wordpress) blog , but I thought I’d add to it over here:

Re: OPML lists and ‘trees’ – etc expect some along these lines soon:

The system developed behind podcast.com enables just that, it’s just that naturally with podcast.com we are only interested in feeds with audio or video enclosures.

So, podcast.com is a ‘client’ of this larger system which will be able to support all types of ‘types’ – all supporting OPML inclusion.

I don’t think you should call them ‘lists’ – I think ‘trees’ is more descriptive. ‘Reading trees’. People can either ‘climb’ up then to find the ‘fruit’ (content) they want, or they could wait for the most popular to drop into their lap. Also through looking through ‘leaves’ (tags) they can be sure of following the right branches, as they ‘climb’ up the tree, deep and deeper into the directory/tree. (’directree’? ;) )

OPML is the tree
FEEDS are the vines
TAGS are the leaves
ITEMS/ENCLOSURES are the fruit.

The coolest thing about OPML inclusion, is you could think of it like climbing up a tree, going along a branch, and then ‘zap’, you’re in another tree. Like magic. Painless. Rewarding. ;)

Telepordata?

Semantics. Taxonomy. Folksonomy.

Now that’s what I call a ’social media network’ ;)

I hope to be able to let some users in to try out the system in the next 4 weeks.

Then people could create something like : http://my.podcast.com/kosso , which is my ‘listening tree’/ podcast folders/ ‘tree of sound’

One reason why I think there are more OPML ‘lists’ than ‘trees’ out there is that there are too few tools out there to effectively manage an OPML file. Your apps have been the best to date (once we got the rss attributes sorted)

I mean managing OPML folders/nodes effectively, over import/export of ‘flat’ lists. So many systems eith dont ‘do’ folders or destroy them on import/export.

Also, I think we really need to evangelise the concept of ‘OPML subscription’ .
OPML files are so often used as static files to import/export bunches of feeds. It is indeed great for backup in that sense. But I believe that ‘OPML Subscription CONNECTS’ all these trees/branches out there. When one branch changes on one tree – if there is an ‘inclusion’ (think of it like a shortcut/symbolic link to a folder) then the other will be able to reflect that if one was to climb that branch.

The trees are ALIVE! So subscribe to them! Watch them grow ;) Feed them, prune them and ‘graft’ ‘cuttings’ to their branches.

We like to call it ‘treedia‘ 🙂

Clearly for this to work, we also need well defined standards to make sure the data out there is well-formed – or we’re screwed ;) but let’s not make the same mistake the web browsers did, by just displaying html if it was broken or not.

I feel like doing another podcast ;)

Is Microsoft Trying To Kill Us?

What on earth are Microsoft trying to do here? This is the US Software Patent office at its most ridiculous ever!! It has filed a patent on the Windows RSS Platform. Does this apply to Europe?

Part of the application appears to be describing the auto-discovery methods of detecting feeds in a page, but what about the rest of it? Is it for desktop Windows applications? Is it web aggregators? Time to call the lawyer.

If granted, one proposed patent would cover “finding and consuming Web subscriptions in a Web browser.” The invention, for example, could allow a user to “subscribe to a particular Web feed, be provided with a user interface that contains distinct indicia to identify new feeds, and…efficiently consume or read RSS feeds using both an RSS reader and a Web browser.”

I’d like to know what effect this is going to have in Google Reader, Bloglines, Netvibes, PageFlakes, etc, etc and OF COURSE, podcast.com (which is nearing the end of a major overhaul with regards to feed reading and feed folder organisation – and ‘amplification’) !!! The whole podcast industry (yes, there is one) relies on this so-called invention.

The application was made in July 2005, so surely there’s plenty of prior art? I don’t get it?

Nick Bradbury, creator of FeedDemon apparently does not want to call Microsoft ‘evil’ just yet, but personally I feel like we just had a major shot across our bows. Friendly fire? I think not.

Is this just a way for lawyers to make money? I sometimes I think these departments just sit around working out ways to make eachother money while screwing consumers in the process. A little bit like a nation’s defence department working out ways to keep us on the brink of mass destruction in order to keep their budgets up and brass polished.

Foul!! Microsoft, I denounce thee!!

I just had an idea : Get every web developer in the world to write an RSS reader and post it up online. Let’s see how they deal with that. 🙂

The Well-Formed Web?


With well formed data, there’s no reason why not. 😉

Now, can we please work out the right way to *do* OPML please?

Basically:

If you want to point to a WEBSITE: use type=link and url=http://your.web.site.com

If you want to point to an RSS feed use type=rss and xmlUrl=htt://www.your.rss.feed.com/rss.xml

If you want to point to MORE OPML use type=include and url=http://www.you.opml.file.com/hoopla.opml

OK. So, you could argue that there might be better ways of doing/specifying it.
And what about htmlUrl, opmlUrl, atomUrl, foafUrl, rdfUrl, type=rdf, type=atom, type=application/pdf or whatever… I don’t know. Just DOCUMENT how you do it.

But what I do know is the sooner we agree and what is the right way to ‘do it’ – the better. 😀

IMHO.

Hey! Is my OPML Icon out of date?? :p

3D News with RSS

For Vista, when it comes out, here’s an app/widget from Microsoft called UniveRSS.

UniveRSS is a showcase application that demonstrates the use of several WPF features, 3D animations, data binding, and data visualization. Currently UniveRSS uses the RSS Feed Store managed through Internet Explorer 7. Later versions will allow you to manage the Feed Store from within the UniveRSS application.

It looks very nice and very much the type of thing I envisage in Second Life, when (and if) we get html on a prim surface. With all the available data bindings and communications we have at our disposal there, it should be faily easy to build this in the Metaverse.

Thanks to Dave Winer for the link.

Podshow redirecting?

Hmm.. All my web requests to anything dot podshow.com are being redirected to btpodshow.com

I know they have just done a deal with BT. But hiking the stats like that is a bit rum.

Is this just a UK thing BT are forcing them to do with IP address blocks (as the BBC do). If you are outside the UK, what happens when you try?

Also, I wish they’d fix up the OPML over on podcastalley. OPML lists of RSS feeds should have entries which are type=rss and the feed url is set to the xmlURL=  attribute 🙂

Expect to see some new OPML tools to manage lots of folders full of feeds in the very near future at podcast.com ;p

An A-Z of Podcasts

If you head on over to index.podcast.com , you will find a browsable A to Z index of all the podcasts in the podcast.com system.

Soon, you’ll be able to easily add feeds from this or the search (or anywhere!) to your own podcast directory and share the shows you like to listen to and organise them into feed folders, all with lovely OPML for you to build nice widgets out of 😉 – and we will to!

Yay! Validated by C-NET News!

Ha! Wow! Just as I was looking for a bunch of RSS feeds from C-NET to play with, I came across their page listing their feeds and saw they were also linking to an OPML feed of feeds too! The ‘validated’ button they’re using is one I created at Dave’s request, a while ago, when he knocked together the validator (beta) for OPML.org!

Heheh! Kewl.

PS: Attn CNET – there’s an error in your OPML – so, it’s not quite ‘valid’ 😉 You have an ‘xmlUrl’ pointing to a web page for downloads.com – tsk tsk – naughty naughty 😉 If you want to point to a website, use ‘htmlUrl’, Thanks. The validator doesn’t go to check to see if your ‘xmlUrl’ is indeed an XML file.