Scripting News for 9/21/07

How to avoid sounding like an monkey 

A few weeks ago a well-respected developer wrote a blog post about something he called the “social graph.” A graph, to most people, is a diagram like the one on the right, which plots the value of a stock over time. For 99.99 percent of the people this is what a graph is. For a very small group of people, a graph is also something like this:

This is the kind of thing you study in a branch of mathematics called Graph Theory. I know a bit about this because when I was an undergraduate, getting a degree in math, I studied this stuff. I proved theorums about how many edges you’d have to traverse to get from one point to another. There are many types of Graph Theory graphs, directed and undirected, for example. Some that you’d need two colors to paint, or three, but none need more than four (a theory that has been proven since I left school, thanks to computers).

Graphs are useful for modeling stuff that goes on in computers. They are also part of a field of math called combinatorics that’s related to statistics, and also related to a highly theoretical area of math called topology.

Now if you showed that diagram to most educated people, they probably would call it a network, and before we talked about social graphs we called them social networks, and you know what — they’re exactly the same thing, and social network is a much less confusing term, so why don’t we just stick with it? (Answer: we should, imho.) So if you don’t want to sound like an idiot, call a social graph a social network and stand up for your right to understand technology, and make the techies actually do some useful stuff instead of making simple stuff sound complicated.

PS: This Google search illustrates. Most of the definitions of “graph” are what you’d expect if you weren’t a math major.

PPS: Copy editors, just change “social graph” to “social network.”

I’ll be writing about podcatchers 

In the coming weeks and months you’ll probably see me writing about issues of podcatchers here, because I’m working on one. It’s the third one I’ve written, so this time maybe I’ll get it right. 🙂

A lot of things have changed since I wrote my first podcatcher back in 2001.

1. Back then there were no podcasts, so it was a proof of concept, a chicken without an egg (or an egg with no chicken), a step in a bootstrap. Today there are lots of podcasts. An embarassment of riches.

2. Back then implementing a podcatcher was simple, there was exactly one format to support, RSS 2.0 with enclosures. Today, luckily, it’s still fairly simple, as far as the format goes. The only variability is the iTunes namespace, which complicates things, just a little.

3. Today there are enough users to make it possible to support lists of podcasts published by fans, and instead of just subscribing to the podcast feeds, you can subscribe to lists of feeds. I will publish one of these lists, in OPML 2.0 format, as a proof of concept.

4. The first version of this new podcatcher will run in the OPML Editor because that’s where all my software runs at first. But the goal is to port it to run in other environments, some with millions of users. I want to provide a popular alternative to the one that Apple publishes which currently dominates the market. (Note: I’m generally pleased with the way Apple dominates, they’ve been very fair about allowing users to export their subscription lists. But if we want to create the opportunity for others to innovate in the area of podcast players, there has to be choice at the podcatcher level. That’s my main motive for revisiting this area.)

There probably are some other changes, and I’ll write about them as the project moves forward.


To people who say that Apple has the market sewn up, I say Bah!

I think iPods are great, but they’re designed to play music, not podcasts.

Every bit of music is something you want to keep forever, a podcast loses almost all its value after you’ve listened to it once.

You have to pay for music (in theory at least) but podcasts are free.

Podcasts beg to have a player that can download them without synching with a desktop computer. Okay that’s something podcasts have in common with music. 🙂

I buy Apple products all the time. I’ve gone from resenting Apple so much that I wouldn’t buy their products, as recently as 2005, to today when not only do I only use Macs, but I’m constantly telling people why they’d be better off using Macs. I can’t help but evangelize the products, I think they’re that much better than Windows PCs.

But as much as I love Apple (can’t believe I actually said that) I still don’t trust them with a whole medium. We need them to have competition. The rest of the tech industry seems to think they’re immune to it, that creates a huge opportunity with someone with enough chutzpah to think they can do it.

Yours truly,

Dave Winer

PS: Here’s my first bit, on the subscription problem, and how it could go away.

Marc Canter 

We could solve the subscription problem 

Okay I’ve been writing about OPML reading lists here for years. I’m now on my second implementation, so maybe this time I’ll get it right. 🙂

But there’s something cool that happens when (hypothetically) the entire installed base of podcatchers supports OPML reading lists. All of a sudden the subscription problem goes poof!

Ask anyone who’s worked on a RSS reader, for that matter, ask anyone who’s used one, what a PITA it is to subscribe to a feed. All those little buttons, or copying and pasting, and looking at urls, and trying to figure out whether you want this format or that format. It’s a miracle anyone actually subscribes to feeds it’s so damned complicated.

Before you blame anyone, it’s not actually anyone’s fault. It’s a result of the market not being a monopoly. The only way to solve the problem is if everyone uses the same web app to manage subscriptions. And we know that’s not going to happen any time soon. Or, if every reader supports OPML reading lists. Now that might actually happen, even though it’s not very likely.

But podcasting, that’s a whole other story. According to many people there’s only one podcatcher, iTunes. So that’s simplified the problem. For example, look at this page of NY Times podcasts, and how they handle it.

See the Subscribe button? Nice. Except for one thing. It really should say “Subscribe in iTunes” because that’s what it does. And it works, because in many people’s minds, iTunes is the only way to subscribe to a podcast.

And it could stay that simple if Apple would do one thing, offer the option of publishing the OPML automatically to a publicly accessible web address, so the user could continue to use Apple’s server to handle subscriptions, even if they’re using a different podcatcher (for example one that runs on a Nokia N800). It would be the mark of a truly great company if they did that. Maybe they are that great.

Otheriwse at some point we’re going to ask the NY TImes to change their page. And they may not be too happy about that. Wouldn’t blame them if they were.

Moral of the story: If we can centralize the subscription process, and move it out of one reader or another, and get the readers to all support subscription to reading lists, the awful ugly issue will go away for users. It’s one of the oldest tradeoffs in the tech business, to make it simple for users, the vendors have to give up some power.

18 responses to this post.

  1. GREAT to see you’re thinking/writing about podcasts and podcatchers again! *high five!*

    You’re absolutely correct about the fact that now we could support ‘lists’ of podcasts people create and listen to or watch.

    Oddly enough, just before reading your post, after seeing your ‘tweet’ I had just made the members section of public at

    You’ll see that every member’s podast directory data is supplied as OPML – also every folder in the whole system has a link to the OPML for that folder at the bottom of the list.

    Every folder also has a fresh podcasts ‘river’ feed (the green rss icon)

    I’m also working on cleaning up a system I have for members to export their podcast subscriptions from iTunes as OPML and import in to their folders. pretty cool!

    Loads of lovely data for you to play with. Drop me a line if you’d like a beta account to have play yourself. 🙂

    cheers! and glad to see you’re feeling better 🙂


  2. Oh and I see you mention the Nokia N800. I have one. The very cool thing about this device (apart from it having linux under the hood) as that its web browser supports the latest desktop version of Flash.

    This means that it would be so easy to create an OPML rendering, RSS reading, podcast downloading interface for that device using the built-in browser and existing tools and skills.

    hurrah! 🙂


  3. Posted by David on September 21, 2007 at 12:07 pm

    I’d be interested to see where you go with this.

    I’m basically an ipod/itunes bigot. I don’t actively look for podcasts if they aren’t in itunes. I’m just too damn lazy to do the whatever extra steps I’d need to add them to my itunes library just so I could synch.


  4. Dave,

    I’ve been hesitating to direct you toward my personal project, an online music player, even while I thought you’d enjoy it, because you haven’t posted on the podcatcher topic in a while. is focussed around music, not podcasts, but we have a very of-the-web goal: to give URLs to music (a media type that has resisted webification until recently.) Once music is a first class citizen of the web, many potential success models for musicians will be unlocked. But it might not be a good time to start a record label – as Bruce Sterling puts it: “That’s just not a business anymore.”

    Podcasting was pretty much the first concerted effort to treat audio as a first-class web resource. Expanding the scope to include music is very exciting. Do you think your podcatcher will be music-friendly?


  5. Cool it’s 2004 again. I still think my enclosure client is the best .. 🙂


  6. Cool it’s 2004 again. I still think my enclosure client is the best .. 🙂


  7. I’ve been looking at how I could use the FOAF format to generate graphs like this from ‘buddy/friend’ lists in a few social networks.

    I think that’s more or less how they did the facebook and twitter ones, but using an API rather than FOAF (Friend Of A Friend).


  8. RE: “Social Graph” vs. “Social Network”

    I think there is good reason for the separate terms. When most people think “Social Network” they think: Facebook, MySpace, LinkedIn, Orkut, Friendster, etc. They don’t think “the online representation of my social connections.” A “Social Network” (in it’s common usage) refers to a product, a package of functionality that’s marketed and that you sign up for.

    I do agree that for anyone outside of the sciences a “graph” brings to mind a different image than intended, but I think simply swapping in “social network” is equally confusing.


  9. Posted by scott on September 21, 2007 at 7:57 pm

    Perhaps proponents of the term graph are interested in framing and solving the problems of normalizing social networks using graph theory? It looks to me like they might be heading towards developing a data interchange format for social networks or establishing ways to form networks of networks.

    Perhaps the term is not for end user consumption? It could be that users are not required to understand what a social graph was to use a service like a social network for the same reasons that they should not be required to understand what RSS or OPML is to subscribe to something.


  10. Posted by scott on September 21, 2007 at 8:34 pm

    Also, if you really want competitor to Apple’s iTunes then you probably will have to move to another playing field. The desktop app space is dying and you will never match the ease of use of iTunes if you start with your OPML Editor. Never.

    A better use of your time might be to assemble and script a mashup of Google’s Reader, Gadgets, and Gears. Gadgets can talk to each other. Gears has a CMS. Talk to the Reader team and come up with the functionality and API for the missing piece(s). There’s also the much rumoured gPhone that is expected to be more open and would provide the means to compete on the iPhone front. If Schmidt can convince Jobs to put Gears on the iPhone then this solution is primed for infiltrating Apple’s premiere device as a benevolent virus.

    If you can’t be David then I think the only option is to find another Goliath to help open up the playing field to more competitors.


  11. Like it or not, “graph” is the correct term for what Brad discusses – a mathematical model, in this case of social networks. There is a big difference between a real-world phenomenon (the social network) and its computer model (a graph). As he actually points out in his piece, this isn’t end-user language. 99.99% of the users of software based on the modelling will use terms like “friends” instead.


  12. dave

    i wrote the same thing (but without explaining myself)

    and after reading the comments, i came to the conclusion that social graph is ok



  13. Dave: the social graph is the common data structure (people and the connections between them) underlying all of the different social networks (web sites and their collections of community features) — they’re different things, existing at different levels of abstraction, and deserve different names. If you don’t understand that, *you’re* the one who sounds like an idiot.


  14. Fred, your post was one of the posts that inspired me to write my piece.

    Buzzwords and phrases are useful if they describe something new.

    For example, I remember when “platform” was new, but I didn’t object to it, because it explained a concept that we needed a word for.

    I was doing audio blog posts before we had the term podcast, and I totally got behind it because we needed a word for what we were doing.

    But social graph is not needed, it makes something simple sound complicated, and we totally need it to sound simple if the problems are going to get solved. They’re not trivial problems, they’ve been there since the Internet outgrew academia and started being used for commercial purposes.

    Another problem with new names for old things is that it tends to push aside the pioneers and makes it sound like newcomers are not also-rans. You had a good gripe as a backer of Wasabe when Mint started getting credit for being a first mover. At least they didn’t have the chutzpah to try to make it a trend and give it a buzzword.

    Someone is being pushed aside with the term “social graph” likely some competitors of Facebook like MySpace and LinkedIn, and some pioneers are going to lose credit for their innovation if it takes root. It may still take root, but I felt I had to say something.


  15. hehe that Marc Canter pic makes me chuckle every time 🙂 hehe

    random. nice


  16. If it’s not “social graph”, use “social diagram” instead. The difference is that is owned by FaceBook while is not.


  17. Just noticed that RSS links on the page do not work. One points to an inexistent rss.php file and another one to loopback host.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: