In this episode we talk about Twitter and the future of journalism, and the XMPP (aka Jabber) protocol and what it means for web developers.
Twitter and journalism
Newspapers are mass journalism, Twitter is a niche — or at least a collection of niches, if that makes sense. Can the latter replace the former completely, or can they co-exist? Is Twitter the future of news distribution? And what will become of the journalism itself?
XMPP (aka Jabber)
- Here Comes Everybody, by Clay Shirky
- Mindhacks, by Tom Stafford and Matt Webb
- Hyperspace, by Michio Kaku
- A View from the Foothills, by Chris Mullin
- Scottish Screen Archive (not really a book)
Matt: Fourth recording, first episode, North Atlantic Radio. With me, Matt Riggott
Brian: And me, Brian Suda
Matt: I’m based in Glasgow — shows you how long we’ve been experimenting with this podcast cos I actually lived in Edinburgh when we first started this, and now I’ve moved westward to Glasgow — and Brian, you’re based in Reykjavík
Matt: And in fact, you’ve moved too as well, haven’t you?
Brian: Moved apartments since we started as well
Matt: Yeah. So hopefully more than just the two of us will be listening to this recording
Brian: So ideally I think the format for these kind of podcasts are going to be short ten, fifteen minutes long. We’re going to talk about maybe one or two topics, a few other little things, and then just kind of post it online for everyone to listen to
Brian: On this week’s podcast, Matt and I are gonna talk a little bit about XMPP, formerly known as the Jabber protocol, and a little bit about Twitter and journalism and a how things are gonna be shaped in the future
Matt: Yep, so which one should go first? Start with the Twitter one because that got us heated last time we talked about it. So there’s been a conversation going around the web over the last couple months, I guess, about how newspapers a fairing given the ubiquity of the Web these days, and if newspapers … if they can survive in their current form and if they can, how that can happen. And I think conversation’s moved on. I’m sure someone else moved on other than us talking between ourselves about, well, new formats for newspapers, I guess, and we had a brief discussion about well, you think — explain your argument. You think twitter is
Brian: Yes, I mean, I’ve been using Twitter for quite some years now and I, from both my perspective and from a lot of things that have been covered, I would argue that some of the content on Twitter will actually replace a newspaper. I think we have our sticking points in that newspaper is a mass journalism type device, where as Twitter is very niche. and I would argue that with the speed and the abilities with Twitter we’re getting more and more news reports that are actually surplanting and surpassing traditional news media, both newspaper and television, and I think we had the argument on if one can possibly completely replace the other or not
Matt: Yeah, we’ll probably getting a little bit, uh, we’re twisting things — were not twisting — we’re getting things a little confused. I guess I would agree with you that you can get a lot of what you get from, say, BBC news reports on Twitter, because they are very shallow in themselves. You get a half an hour news programme at six o’clock in the evening or nine o’clock or wherever time it’s on, and the reports they give you on each piece is very very — well it’s a very shallow piece, it’s very brief, they don’t go into too much detail at all. And in actual fact you can probably get something similar from Twitter. Possibly where we disagree on the point is, I don’t necessarily think the most important bit of newspapers or TV news is the speed of the reports or the amount of information you get. It’s more about collating that information and investigating stories deeper, and I don’t think that is going to come through Twitter. I think that’s the best resource that newspapers have, and I think that’s what they need to save rather than the medium per se or, you know, being first with news or anything like that. I think what they need to save is the investigative journalism — or in fact journalism itself
Brian: You’re probably right there. My own, my argument that would be the newspaper only has, you know, one hundred people on the staff, if that, whereas Twitter might have several million users, and when you know you have one person that can only spend so many hours digging into the report, whereas Twitter, I mean, you might get the breaking report quick, but you can also then get the follow-ups, the individual stories, and you can get it potentially from hundreds, if not thousands, if not millions of different angles, whereas the newspaper is being written by one person from one angle. Now, I will agree that what a newspaper is good at is the journalistic aspect and the filtering, in that one person sits down, writes a cohesive story, and puts it into, you know, a thousand words, and it gets copy edited. What we don’t get on Twitter is an easy way to pull in those thousands and thousands of short 140-character blurbs into some sort of cohesive single thought. But I think that’s going to be — I mean, those tools just need to be built and if it’s possible, I think it is
Matt: What do you suggest then? Rather than having an editor of a newspaper, you just have an editor of Twitter — not a central editor, that say, someone who took all the tech news or something, and then they collected the stories or made it made it more coherent?
Brian: Yes. I mean, you’ve seen some of the, some of the more useful Twitter applications out there, do threading — even the search, where if you reply to somebody it starts building the kind of conversation tree. And if there was some smarter stories out there, you know, if the Kindle 2, when that came out, a lot of people mentioned it in Twitter. If there was an easy way to say, this person is talking about this feature, and then this person is also contributing their thoughts about this feature and build up … I don’t know, I’m not 100% sure how it would look, but I would argue that through the use of the millions and millions of people — you know, many eyes, with many eyes any bug is just shallow
Brian: It’ll be the same for – my argument would be it would be the same for the story. You know, maybe one editor at CNET might review a tech product; is that better than having a thousand people give there voice an opinion about it? There are some websites, like EveryBlock and Outside In, are attempting to geolocate some of the Twitter messages, so if you say ‘I am at XYZ Deli’ it tries to look that up in the database, and then you can start browsing your news geographically. And that’s — when you start clumping those sorts of things together, I think that is how the sort of article, cohesive article of the future might might look from thousands of different people
Matt: Technically it sounds fantastic. I’m still not convinced that there’s the useful data on Twitter. People’s messages don’t tend to be … they’re certainly not as descriptive or as interesting as a news article
Brian: No, but that’s also maybe that’s our myopic world. I mean, we’re in a bit of an echo chamber because we probably only listening to our friends. If there was a way to look and listen to everybody at the same time, that editor, you know, whoever that editor is, be it automatic or human, would break out of that … the crap, more or less
Matt: Have you got an example then of something that you found out through Twitter, and purely through Twitter you’ve found out, you know, an incredibly in-depth amount of information about it?
Brian: Well, this is also the … I think the other thing we forget is that Twitter, much like newspapers, I mean, any, any newspaper on the web has links off to other stories, so there’s no reason why some sort of Twitter information couldn’t also contain a link. I mean, the big example more recently was — is it Monrovia?
Brian: Moldova. The riots in the capital. Those were, as far as I know, not even really covered by traditional news sources. And one day I was looking on the Twitter search and it was trending topic, and that was shooting up the list. So when you click on that, it was telling you, you know, police are in the square, they’ve broken into the, the parliament building, all the things are on fire. Now, those are just individual little snapshots, and you can build a timeline off of that, and then hopefully fill in the gaps with either links to people’s blog accounts, or maybe supplemental information that’s not necessarily directly on Twitter, but is found via Twitter
Brian: I think along those same lines, one of the other topics we’re going to talk about this evening is XMPP
Brian: Which is formerly known as the Jabber protocol and, XMPP is an interesting one because it’s a push service rather than a pull. And I guess the difference is we’re mostly used to you … visit a website and request the information, and you’re basically pulling the web page back. And the same works for RSS, you know, every thirty minutes it makes a request, and it pulls the data down to you. Whereas things like people on the BlackBerry get push email or push calendar to where it’s not checking every thirty seconds or every minute. When a new message comes in, it gets pushed directly to your phone
Brian: So XMPP works in a similar manner
Matt: Yeah, the obvious way to look at it is, you know, RSS feeds, you can check them every half an hour and download anything new, whereas if you sit on an instant messenger as soon as that person’s typed a message — another person’s typed a message to you, the reply comes to you. You don’t check for replies every half an hour and then send your replies
Brian: Right, it’s synchronous rather than asynchronous
Brian: And the way, much like — your IM analogy is a good one — XMPP allows you to hold an open connection, and then any time something happens it just comes directly down that connection
Brian: So it also enables services to scale a lot better because otherwise, if you’ve got a million users, those million users request your website every thirty minutes, and you say ‘No, nothing’s changed’, you’ve got to reply to those million people, whereas if you just have all those as open connections, if something changes, you just push it to the five or ten people that it might affect
Matt: Yeah. And you have a hell of a lot less to do on your server. You get a lot less load and, of course, everyone else gets to instant updates as well, so it benefits everyone. It’s a fascinating idea. I think it’s something that a lot of websites would like to have a go out — even just tinker with — but the technology seems to be, it’s not complicated, but it seems to be hard enough to get to set up — or it is hard enough to set up on your own server to play with. You haven’t seen … you don’t see in action anywhere really
Matt: Flickr said, you know, wouldn’t it be great if we could do this? But as far as I know, they haven’t actually set this up. They haven’t said here you go, here is an XMPP API for you
Brian: No. I think the biggest proponent I know is, again, Twitter, and that’s how we kind of segued on to this. They have, it’s just called the firehose, where every single message that comes in, they just broadcast down the XMPP pipeline, and you just subscribe to that — if you want to — and then you just basically get the firehose worth of data
Matt: So Twitter does use XMPP for their firehose?
Brian: I’m pretty sure, yes
Matt: Ah, OK
Brian: And then there are a few services, like visual – is it the visualisation services? — that build on top of that
Matt: I didn’t know Twitter actually was using XMPP, but that’s good if they do
Brian: Don’t necessarily quote me on that
Brian: I’ve seen it in presentations
Matt: I don’t know if you, but we are recording this, Brian
Brian: Okay, well … don’t take that verbatim
Brian: Don’t take it as truth. I remember sitting in presentations where people were showing that. It’s probably not available to the public
Brian: So I guess this is the bit where we talk about what we’ve been reading, and it doesn’t necessarily have to be anything techie or technical or geeky, just some of the books that we’ve had on our bookshelves and we’re looking at. Just a kind of, maybe, again, bit of serendipity, show some other interesting sides rather than always talk hard-core tech. Although sadly this week, the two or three books that have been reading are techie
Brian: I’ve been working on, I’m just finishing up the Clay Shirky Here Comes Everybody, which is an interesting book. I’ve seen his, both his presentations on the Web and a lot of the same sort of material is covered, but it’s a couple really good resources in there. I’ve also been working on the Mindhacks book, the O’Reilly Mindhacks, which is a fascinating book all about how the brain works and how your eyes and ears. But the other great thing is it’s basically a hundred hacks, so it’s only like two or three pages each, so it’s really easy to pick up and put down. And then the other book that I’m just starting in on is Hyperspace: the Tour of Ten Dimensions by Michio Kaku. It should be interesting. It was written 1993/94, so probably a bit of it is outdated by now, but it’ll still be interesting to see what his thoughts were
Matt: What have I been reading? It’s a good question. I’m actually standing in front of a bookshelf now, full of about one hundred twenty books. What I should be doing is choosing the most intelligent-sounding book from that list, from the shelf. I’m actually in between books at the moment. The last book I finished was A View from the Foothills, which is a political diary by a British MP called Chris Mullen, and it’s a brilliant book. If anyone’s interested in, well, parliament, and the goings-on in parliament, and the daily life of an MP, I would heartily recommend it, it’s a thoroughly good read. Just recently out in hardback, I think. Anyway, that was an excellent book. And I’m gonna dismiss tradition — even though this is only episode one — and instead of mentioning the book I’m reading, as I’m not reading a book, I’m going to mention a website I found today. So we’ve both failed miserably at moving away from nerdy stuff in this little section, but it’s too good, too good not to mention. The National Library of Scotland has a website called the Scottish Screen Archive website — we’ll have show notes that will have any URLs we’ve mentioned in the podcast, but ssa.nls.uk — the Scottish Screen Archive, and it’s brilliant. It’s got clips of videos going back, or clips of films going back to 1890 or so. Literally just street scenes in Scotland — almost as if people have, you know, were just trying out their cameras, and they walked out of their house and were just filming people going about their daily business. Some silly things, some interesting things. There’s a video of Laurel and Hardy making a visit to Edinburgh in 1932, and they just get off a train amid this massive crowd, walk up the train, er, walk about Waverley Station to Princes Street, and go up through the castle. It’s fascinating to watch. So have a look at that. And next week I’ll promise that, or next time we record this I promise I’ll be reading a book, so I can actually talk about a book
Brian: Yeah. No promises that it won’t be a techie book though