Archive: October, 2006

The song Lay Low by My Morning Jacket is the eighth track on the album Z. When the song starts, it seems like your typical My Morning Jacket song, ‘though perhaps a bit more upbeat than most. For the first few minutes, Jim James sings away in his usual style.

At precisely three minutes and three seconds, the vocals cease and the purely instrumental portion of the song begins. As one guitar continues to play the melody line, a second guitar begins its solo.

It starts like something from Wayne’s World: a cheesy little figure tapped out quickly on the fretboard. But then it begins to soar. Far from being cheesy, it quickly becomes clear that what I’m hearing is the sound of joy articulated through the manipulation of steel strings stretched over a piece of wood, amplified by electricity.

As the lead guitar settles into a repeated motif, the guitar that was previously maintaining the melody line switches over. At exactly three minutes and thirty seconds, it starts repeating a mantra of notes that are infectiously simple.

For a short while, the two guitars play their separate parts until, at three minutes and thirty three seconds, they meet. The mantra, the riff — call it what you will — is now being played in unison, raising my spirits and pushing the song forward.

The guitars remain in unison until just after four minutes into the song. Now they begin to really let loose, each soaring in its own direction as the rest of the band increase the intensity of the backing.

Two seconds before the five minute mark, the guitar parts are once again reunited, but this time in harmony rather than unison. At five minutes and nineteen seconds, a piano — that was always there but I just hadn’t noticed until now — begins to pick out a delicate tinkling melody in a high register. It sounds impossibly fragile surrounded by a whirlwind of guitars, drums, and bass, but it cuts right through. And it is beautiful.

At five minutes and twenty six seconds, as the piano continues to play, one of the guitars drops down low and starts growling out its solo. From there, everything tumbles inevitably to the end of the song.

The band stops playing at five minutes and fifty seconds, but we’re given another twenty seconds to hear the notes fade to silence. The instrumental break has lasted three minutes. It is the most uplifting and joyous three minutes that has ever been captured in a recording studio.

Now, divine air! now is his soul ravish’d! Is it not strange that sheep’s guts should hale souls out of men’s bodies?

While I had Matthew in my clutches, I made him show me around the API for They Work For You. Who knew that so much could fun be derived from data about MPs?

First off, there’s Matthew’s game of MP Top Trumps, ‘though he had to call it MP Fab Farts to avoid getting a cease and desist letter.

Then there’s a text adventure built on the API. This is so good! Enter your postcode and you find yourself playing the part of your parliamentary representative with zero experience points and one hundred hit points. You must work your way across the country, doing battle with rival MPs, as you make your way towards Sedgefield, the lair of Blair.

You can play a Web version but for some real old-school fun, try the telnet version. This reminded me of how much I used to love text adventures back in the days of 8-bit computers. I even remember trying to write my own in BASIC.

For what it’s worth, Celia Barlow, MP for Hove, has excellent pesteredness points. I made it all the way up to Sedgefield and defeated Tony Blair in battle. My prize was the source code of the adventure game in Python.

Ah, what larks!

There’s another project that Matthew works on that I find extremely useful. He has created accessible UK train timetables using the data from the National Rail site, a scrAPI if you will. This is where I go whenever I need to plan a train journey.

The latest feature is something that warms the cockles of my heart: beautiful, hackable URLs. If I want a list of trains going from Brighton to London, I can just type:

http://traintimes.org.uk/brighton/london

It handles spaces (or pluses or underscores) too:

http://traintimes.org.uk/brighton/london victoria

The URL can also be extended with a departure time:

http://traintimes.org.uk/brighton/london victoria/14:00

My address bar is my command line. This is the kind of design that makes URL fetishists like Tom very happy.

I’m at an event called Take Back The Web. It’s a cosy little unconference aimed at non-profits and activist groups.

There’s been plenty of education and discussion going on all day, mostly around things like blogs, wikis, RSS and podcasting. I followed up the RSS talk with a little spiel about APIs and how they can be used to pull in data from other places on the web.

I’m used to attending geekier events where everyone is fairly tech-savvy, but the crowd here is mostly made up of people on the ground who want to be able use technology but who aren’t necessarily from a technological background. It really brought home to me just how far we have to go in making this stuff less geeky and scary-sounding.

Just about everyone gets blogs, and it’s pretty easy to get started with them. Wikis are a little bit trickier, but still attainable. RSS becomes harder again: it’s still too hard to subscribe, and even the term “subscribe” is itself misleading, implying payment. As for APIs, that’s still all pretty much rocket science so I just gave a basic overview of the benefits without really discussing the nitty-gritty of programming.

Notice how the terms change in complexity along that scale: from the word blog to the term API. We’re using way too many acronyms and technobabble for this stuff. Of course, we can’t change the names without upsetting the geeky programmers.

I got a lot of food for thought from the day so far, even though I already know about the technologies. It’s been fascinating to see how people are using the web now and also how much more they could be doing.

The guys from mySociety/They Work For You are talking through their services now and I’ve just found about this nifty API. I’ll have a play around with that. I’ll quiz Matthew about it later; he’s staying over with me. More grist for the bedroll.

Right before I set off on my antipodean adventure, hell froze over and I opened up comments. There was a twist, of course. I wanted to solicit feedback but discourage conversation so, to foster independence of opinion, comments were not immediately published. Instead, all the comments were published together once the ability to add comments had been switched off. In this case, that was a period of one week.

The post received quite a few comments, some more interesting than others.

Quite a few people were sidetracked by the issue of spam and assumed that I would simply be inundated with comment spam. Spam was something I deliberately didn’t mention in the original post as it is tangential to the matter at hand. As it happens, the Akismet API — which I’m using to filter comments — does a great job of keeping comment spam out.

The original post was prompted by Mike mentioning the lack of comments on my site at the Future of Web Apps summit. The audio of Mike’s talk is now available and I highly recommend you listen to it. It’s an excellent presentation chock full of ideas about community, participation and social interaction. Mike also weighed in on my post, saying:

In my view, comments can significantly raise the quality of experience on most sites but cannot significantly lower it.

Now, this is interesting because I disagree almost 100%. In my experience, coming across pointless, lame, nasty or redundant comments attached to a blog post can lower the quality of the document as a whole. I think Mike has managed to crystalise why others are willing to put up with the downsides of comments, while I find the trade-off unacceptable.

Bradley Wright says:

I’m curious as to how comments that no one else can see will foster good feedback.

Well, that goes back to the whole wisdom of crowds thing I was talking about: independence is one of the defining factors.

He continues:

Sometimes the best leaps I’ve seen in online “conversations” come from the back and forth, which is helpful for some people to help crystallise their thoughts.

I agree. But I haven’t found those conversations happening very often in comments. Instead, I find them happening blog to blog. People tend to post more well-thought responses on their own blogs than they do in the comments to someone else’s blog post.

Guy Carberry brings up an important issue:

I have got some very useful information from the comments of a post.

This is true. This is also something that Andy said to me when we were discussing blogs and comments. Well, with my system of comments, this will still be true. Once comments are displayed, there is a greater chance that they will contain only useful information rather than a mixture of information and conversation. The google dance usually takes a while anyway, so if someone is looking for information on a topic and they end up at one of my blog posts, the ability to add comments will probably be disabled and all comments will be displayed. That person gets the information in the original post along with any feedback sent in via a comment. In fact, I would argue that my comment system will prove more valuable for long-term information search.

But how long is long? For how long should the ability to add a comment be available? For the inaugural post, I chose one week. I figured that would be a nice length of time to allow everyone to have their say. As it turned out, the vast majority of comments were submitted in the first few days. Clearly, a week is a long time on the web. I think two to four days might be a better length of time to keep the comment form available.

Pete Lambert commented on the need for some way of tracking when comments get published:

My only issue with it is that I would have to remember to come back and check the comments after whatever arbitrary quarantine period you define.

This is clearly something I need to work on. I still think some form of email notification would work best. Another option would be to actually publish the comments in my RSS feed as well as on the site. With a little jiggery-pokery, this could be flagged up as unread in most RSS readers. But this could just end up pissing off the people who don’t care when comments are published.

Here’s another point:

I suspect what you are going to wind up with is a long list of the same comment over and over again.

Surprisingly, this didn’t turn out to be the case. I actually received a fairly wide range of ideas from people.

That’s not crowd wisdom at all. If your blog was a multiple choice quiz, then fine, but it’s not.

That’s a good point: the wisdom of crowds does seem to work best with multiple choice questions (or “guess how much this weighs” questions). I opened up comments on another entry, where I was soliciting opinions on things to do in Melbourne. I personally thought that this would be a better test of my commenting system. The aggregate results should have given me a good independent overview of recommendations.

As is turned out, I shot myself in the foot by

1. restricting the crowd to people who know anything about Melbourne and
2. also encouraging feedback via my contact form.

Still, my point above about the added value of comments via findability still stands: somebody googling for restaurants in Melbourne will not just find my original post, but also the (one) useful comment about places to eat. The fact that the comments was originally hidden during the submission process doesn’t affect the long-term usefulness. I can see the commenting system working well for less tightly focused questions.

Based on the inaugural post with comments, I’m pretty pleased with the results. Clearly I need to work on a better notification system and I also need to figure out the sweet spot of how long to allow comments to be submitted but overall, this little experiment has yielded some pretty tasty fruit.

I was all set to start allowing more comments when I came across a post by Robert Nyman entitled A blog without comments isn’t really a blog.

Oh, brother! Here we go again.

After all these years of trying to define what a blog is, who knew that someone blogging for a year and a half would stumble upon the answer?

Forgive my sarcasm, but I get pretty fed up of being judged based on how I set up my personal site. I said as much in a comment on Robert’s post.

In his post, Robert has this advice for people who don’t have comments enabled:

Then you shouldn’t blog.

At the same time, he claims to love the openness of the web. I’m seeing a disconnect here. In his reply to my comment, Robert says:

I’m glad that I have comments here so everyone can read your and my opinions on this, and also have the possibility to contribute with what they think. That’s what I like about the web: the openness.

That’s a pretty narrow view of openness. I don’t think it’s very open to have all responses in a single document.

Here’s what I like about the web: openness through hypertext. I can link to any resource on the Web and comment on it here on my own website. Conversely, anyone is free to link to this document and comment on it on their own site (or a third-party site like Newsvine). That’s openness. Demanding that everyone post their thoughts together on the originating document is a closed system.

So, just as I was ready to start experimenting more with feedback mechanisms, I’m confronted with a “my way or the highway” ultimatum about what I’m allowed do on my own site. It’s enough to make me give up on the whole idea.

But I won’t. I’ll keep opening up the occasional post for comments, although I’m also really interest in using tags and pings for tracking responses, hence the Del.icio.us, Technorati, and Flickr links at the end of each post.

I was thinking of opening comments on this very post, but I think most of the salient points have been gathered from the initial post. If you would like to respond to this, you can write a blog post of your own and it will appear through the Technorati link (I really wish Trackback worked better but spam has effectively killed that off). If you have any feedback, there’s always email and IM.

I talked a while back about how I was attempting to add some extra context to my posts by pulling in corresponding tag results from Del.icio.us and Technorati, and then displaying them together through the magic of Ajax.

It struck me that there was another tag space that I had completely forgotten about: Flickr. Now at the end of any post that’s been tagged, you’ll find links entreating you to pull in any of my Flickr pics that have been likewise tagged.

This is all possible thanks to a single method of Flickr’s API. I’m reusing the same method to search for other pictures too…

A had a little epiphany in the pub the other night, chatting after the WSG meetup. I was talking about geotagging and I mentioned that it probably won’t be too long before just about every file will be geotagged in the same way that just about every file already has a time stamp. Then I realised, “hey, all my blog posts have time stamps and so do all my Flickr pics!”

So I added an extra link. You can search for any pictures of mine that were taken on the same day as a journal entry. I like the extra context that provides.

While I was testing this new functionality, I couldn’t figure out why some pictures weren’t being pulled in. Looking at the post from the Opera event written on Tuesday, I expected to be able to view the pictures I took on the same night. They weren’t showing up and I couldn’t understand why not. I assumed I was doing something wrong in the code. As it turned out, the problem was with my camera. I never reset the date and time when I came back from Australia, so all the pictures I’ve taken in the last couple of weeks have been off by a few hours.

Keep your camera’s clock updated, kids. It’s valuable metadata.

Hmmm… I guess I should take a picture today to illustrate the new functionality. In the meantime, check out this older post from BarCamp to see the Ajaxitagging in action.

Shel Israel and Rick Segal are doing a bit of a world tour. They’ve just been to Ireland and now they’re winging their way to Brussels. In between, they made time for a whistle-stop visit to Brighton.

Why Brighton? Ben gave ‘em the heads-up that it’s a happening place.

Andy reserved a table for lunch at Carluccio’s and rounded up a bunch of local geeks and entrepreneurs. Much excellent conversation ensued, much of it about social media and, more importantly, people.

It was a real pleasure to meet Shel and Rick. It’s a shame they couldn’t stick around longer: they were literally in town for just a few hours. We didn’t even have the opportunity to take them out on the pier for a stick of rock.

There’s always next time.

After that, we convened to the pub where many excellent conversations were to be had. I had a great chat with some of the Multimap gang, and I got to meet Alex Robinson, yes that Alex Robinson. As usual, I had to run away to catch a train back to Brighton. Also as usual, it took me hours to get home.

I’ve made a PDF of my slides but, as is always the case, they don’t make too much sense without the context of the presentation. Hopefully, there will be a podcast of the evening’s talk released. Head over to Stuart’s WSG page where you can subscribe to the feed.

It’s seems like I’m going up and down to London like a yo-yo this week.

On Monday, I gave a day of Ajax training at Framfab LBI. They’re a smart bunch. Normally, I have to sell developers on the concepts of progressive enhancement and unobtrusive JavaScript but these guys were already walking the walk. I felt kind of bad: for at least the first half of the day, I must have been preaching to the converted. Nonetheless, they were all very gracious and said they got a lot out of the day anyway. Too kind.

Yesterday evening was the Opera event which, despite the technical hitches, ended up being an enjoyable night out. Once the sales pitches and PowerPoint were out of the way, everyone was able to relax and enjoy the free booze, the mysterious canapés, and of course, the company.

After my talk and my hasty blog post, I spent the remainder of the evening explaining to people that no, I don’t have any connection to Opera, and wishing I had introduced myself before I started spouting my little after-dinner speech.

I had the chance to hang out with some of the gang from Last.fm, which was a lot of fun. It turns out that Hannah is a fellow believer in fighting the good fight for liquid layouts.

Today is the one day I won’t be getting on a train to the big smoke. Band practice takes precedence. Tomorrow, though, I’ll be returning to the bosom of mother London.

The second ever Web Standards Group meetup in London is taking place at the New Cavendish Street campus of Westminster University. The theme of the evening is microformats. Norm!, Drew and I will be covering the past, present and future of the single coolest thing happening on the Web right now. Why not join us for an evening of entertainment and education? The event is free. You can find more details on the Upcoming page. Try to get there for around half past six. Afterwards, we’ll decamp to the Bricklayer’s Arms for drinks ‘till late.

Be there or be square.

I’ve just come offstage, having spoken at the Opera Backstage event in Leicester Square in London.

I was very pleased to be asked to speak. I decided to do a really pretentious over-the-top talk, which I hastily prepared on the train the day before. Forgoing slides, I settled on using the same JavaScript teleprompter that I used at Reboot.

The best laid plans of mice and men…

In between walking from my seat to the stage, my iBook died. It just shut down and wouldn’t start up again. I was left standing on stage with no slides, no script, nothing. It was like that dream where you show up for school without your pants.

“Screw it”, I thought. I decided to wing it. I think I managed to recall most of what I was going to say. It was mostly about science-fiction and Irish poets.

Now I’m done talking, my laptop is behaving just fine. Typical.

I’ve posted the script of what I was planning to say — which is, more or less, what I ended up saying — over in the articles section.

I flew back from Australia at the start of the week… and boy, are my arms tired — ba-doom!

The day of traveling went surprisingly smoothly. That was probably due to the fact that Jessica and I were flying with Virgin Atlantic’s “upper class” service — John and Maxine sure know how to take care of their speakers. The best part was getting a complimentary ride home from Heathrow airport right to our front door in Brighton.

Actually, the best part was probably on the outward journey, getting to hang out in the Clubhouse at Heathrow. It was like the space station from 2001: A Space Odyssey.

Now that Web Directions South is well and truly wrapped up, I’m gripped by the usual tug of post-conference emotions. On the one hand, I’m feeling very inspired by the excellent presentations (and even more excellent people) to start hacking and coding some new stuff. On the other hand, the time I spent in Australia means I’ve let a lot of stuff slide. Now that I’m back, I need to catch up with a whole slew of commitments. That leaves me with no time to put any grand schemes into action.

Cameron sums up the dilemma of maintaining the post-conference buzz nicely. He even illustrates the point with diagrams.

Speaking of Cameron…

I feel I should explain some of the more, um… “unusual” pictures that began showing up on Flickr during Web Directions South.

While I was preparing my first presentation, Explaining Ajax, I needed a screen shot of a typical page on Flickr. What better page to use than this photo entitled “Topless Cameron Diaz”? Don’t worry: it’s safe for work. It’s a picture of Cameron Adams and Dustin Diaz with their shirts off, see?

Now, I felt perfectly justified in using this photo. Cam is infamous for using his presentations as platforms for ridiculing others. I thought it would be fun to put him at the receiving end for a change.

“You’re a dead man!”, he shouted from the audience when the photograph blazed across the screen.

He was not a happy camper.

The next day, Cameron was presenting together with Kevin Yank. I think it would be fair to say that everyone was waiting for the other shoe to drop.

Right in the middle of a superb talk on APIs, this slide appeared.

In the interest of full disclosure, I feel I must point out that that is not my body. It’s a Photoshop job!

I had the opportunity to get in the last word. My second presentation came after Cam and Kevin’s. I could have used this as a platform to get in one last dig, but I decided to be a bigger man than that. I called a truce. Besides, I was in mortal terror of what revenge Cam would wreak… perhaps not today, perhaps not here, but at some future date, in a dark alley, years from now, when the whole incident has long faded from my memory.

I did, however, point out that only cheats use Photoshop. So there.

All’s well that ends well. We kissed…

…and made up.

My time in Melbourne is almost at an end. Thanks to everyone who sent tips on places to go and things to see here. Most of my activities, as evidenced by my Flickr pics, have revolved around food. I must get around to writing it all up on Principia Gastronomica.

I took some time out from my culinary explorations to give a talk at the local Web Standards Group meetup. It was fun. I recycled my talk from d.Construct, The Joy of API. People seemed to enjoy it and there were a lot of great questions asked afterwards.

The audio from the talk at d.Construct is now available through the podcast. I’ve had the audio transcribed — using Casting Words — and I’ve posted the results here.