Friday, February 19th, 2021
Monday, April 20th, 2020
Thursday, April 9th, 2020
On Monday, I linked to Tom’s latest video. It uses a clever trick whereby the title of the video is updated to match the number of views the video has had. But there’s a lot more to the video than that. Stick around and you’ll be treated to a meditation on the changing nature of APIs, from a shared open lake to a closed commercial drybed.
It reminds me of (other) Tom’s post from a couple of year’s ago called Pouring one out for the Boxmakers, wherein he talks about Twitter’s crackdown on fun bots:
Web 2.0 really, truly, is over. The public APIs, feeds to be consumed in a platform of your choice, services that had value beyond their own walls, mashups that merged content and services into new things… have all been replaced with heavyweight websites to ensure a consistent, single experience, no out-of-context content, and maximising the views of advertising. That’s it: back to single-serving websites for single-serving use cases.
A shame. A thing I had always loved about the internet was its juxtapositions, the way it supported so many use-cases all at once. At its heart, a fundamental one: it was a medium which you could both read and write to. From that flow others: it’s not only work and play that coexisted on it, but the real and the fictional; the useful and the useless; the human and the machine.
Both Toms echo the sentiment in Anil’s The Web We Lost, written back in 2012:
Five years ago, if you wanted to show content from one site or app on your own site or app, you could use a simple, documented format to do so, without requiring a business-development deal or contractual agreement between the sites. Thus, user experiences weren’t subject to the vagaries of the political battles between different companies, but instead were consistently based on the extensible architecture of the web itself.
I know, I know. We’re a bunch of old men shouting at The Cloud. But really, Anil is right:
This isn’t our web today. We’ve lost key features that we used to rely on, and worse, we’ve abandoned core values that used to be fundamental to the web world. To the credit of today’s social networks, they’ve brought in hundreds of millions of new participants to these networks, and they’ve certainly made a small number of people rich.
But they haven’t shown the web itself the respect and care it deserves, as a medium which has enabled them to succeed. And they’ve now narrowed the possibilites of the web for an entire generation of users who don’t realize how much more innovative and meaningful their experience could be.
In his video, Tom mentions Yahoo Pipes as an example of a service that has been shut down for commercial and idealogical reasons. In many ways, it was the epitome of what Anil was talking about—a sort of meta-API that allowed you to connect different services together. Kinda like IFTTT but with a visual interface that made it as empowering as something like the Scratch programming language.
There are services today that provide some of that functionality, but they’re more developer-focused. Trys pointed me to Pipedream, which looks good but you need to know how to write Node.js code and import npm packages. I’m sure it’s great if you’re into serverless Jamstack lambda thingamybobs but I don’t think it’s going to unlock the potential for non-coders to create cool stuff.
Cables is a tool for creating beautiful interactive content.
It isn’t about making mashups, but it does look something that non-coders could potentially use to make something that looks cool. It reminds me a bit of Bret Victor and his classic talk on Inventing On Principle—always worth revisting!
Monday, April 6th, 2020
Tom’s videos are so good! Did you see his excellent in-depth piece on copyright?
This one is all about APIs and the golden age of Web 2.0 when we were free to create mashups.
It pairs nicely with a piece by another Tom from a couple of years back on the joy of Twitterbots.
Tuesday, November 13th, 2018
Jigsaw puzzle companies tend to use the same cut patterns for multiple puzzles. This makes the pieces interchangeable, and I sometimes find that I can combine portions from two or more puzzles to make a surreal picture that the publisher never imagined. I take great pleasure in “discovering” such bizarre images lying latent, sometimes for decades, within the pieces of ordinary mass-produced puzzles.
Wednesday, June 13th, 2018
Refresh for a new design challenge.
Saturday, March 31st, 2018
Tuesday, May 2nd, 2017
Absolute genius! I’ll never hear Sgt. Pepper’s quite the same way again.
Wednesday, August 24th, 2016
I giggled at quite of few of these mashups.
Friday, March 11th, 2016
The thesis: any film is improved by playing Walk Of Life by Dire Straits over the ending.
The proof: this website.
(this is absorbing and brilliant)
Friday, October 4th, 2013
Radio Free Earth
Back at the first San Francisco Science Hack Day I wanted to do some kind of mashup involving the speed of light and the distance of stars:
I wanted to build a visualisation based on Matt’s brilliant light cone idea, but I found it far too daunting to try to find data in a usable format and come up with a way of drawing a customisable geocentric starmap of our corner of the galaxy. So I put that idea on the back burner…
At this year’s San Francisco Science Hack Day, I came back to that idea. I wanted some kind of mashup that demonstrated the connection between the time that light has travelled from distant stars, and the events that would have been happening on this planet at that moment. So, for example, a star would be labelled with “the battle of Hastings” or “the sack of Rome” or “Columbus’s voyage to America”. To do that, I’d need two datasets; the distance of stars, and the dates of historical events (leaving aside any Gregorian/Julian fuzziness).
For wont of a better hack, Chloe agreed to help me out. We set to work finding a good dataset of stellar objects. It turned out that a lot of the best datasets from NASA were either about our local solar neighbourhood, or else really distant galaxies and stars that are emitting prehistoric light.
The best dataset we could find was the Near Star Catalogue from Uranometria but the most distant star in that collection was only 70 or 80 light years away. That meant that we could only mash it up with historical events from the twentieth century. We figured we could maybe choose important scientific dates from the past 70 or 80 years, but to be honest, we really weren’t feeling it.
We had reached this impasse when it was time for the Science Hack Day planetarium show. It was terrific: we were treated to a panoramic tour of space, beginning with low Earth orbit and expanding all the way out to the cosmic microwave background radiation. At one point, the presenter outlined the reach of Earth’s radiosphere. That’s the distance that ionosphere-penetrating radio and television signals from Earth, travelling at the speed of light, have reached. “It extends about 70 light years out”, said the presenter.
This was perfect! That was exactly the dataset of stars that we had. It was a time for a pivot. Instead of the lofty goal of mapping historical events to the night sky, what if we tried to do something more trivial and fun? We could demonstrate how far classic television shows have travelled. Has Star Trek reached Altair? Is Sirius receiving I Love Lucy yet?
No, not TV shows …music! Now we were onto something. We would show how far the songs of planet Earth had travelled through space and which stars were currently receiving which hits.
Chloe remembered there being an API from Billboard, who have collected data on chart-topping songs since the 1940s. But that API appears to be gone, and the Echonest API doesn’t have chart dates. So instead, Chloe set to work screen-scraping Wikipedia for number one hits of the 40s, 50s, 60s, 70s …you get the picture. It was a lot of finding and replacing, but in the end we had a JSON file with every number one for the past 70 years.
Meanwhile, I was putting together the logic. Our list of stars had the distances in parsecs. So I needed to convert the date of a number one hit song into the number of parsecs that song had travelled, and then find the last star that it has passed.
By the end of the first day, the functionality was in place: you could enter a date, and find out what was number one on that date, and which star is just now receiving that song.
After the sleepover (more like a wakeover) in the aquarium, we started to style the interface. I say “we” …Chloe wrote the CSS while I made unhelpful remarks.
For the icing on the cake, Chloe used her previous experience with the Rdio API to add playback of short snippets of each song (when it’s available).
Here’s the (more or less) finished hack:
Basically, it’s a simple mashup of music and space …which is why I spent the whole time thinking “What would Matt do?”
Just keep hitting that button to hear a hit from planet Earth and see which lucky star is currently receiving the signal.*
*I know, I know: the inverse-square law means it’s practically impossible that the signal would be in any state to be received, but hey, it’s a hack.
Wednesday, July 3rd, 2013
A superb piece by Marco Arment prompted by the closing of Google Reader. He nails the power of RSS:
RSS represents the antithesis of this new world: it’s completely open, decentralized, and owned by nobody, just like the web itself. It allows anyone, large or small, to build something new and disrupt anyone else they’d like because nobody has to fly six salespeople out first to work out a partnership with anyone else’s salespeople.
And he’s absolutely on the money when he describes what changed:
RSS, semantic markup, microformats, and open APIs all enable interoperability, but the big players don’t want that — they want to lock you in, shut out competitors, and make a service so proprietary that even if you could get your data out, it would be either useless (no alternatives to import into) or cripplingly lonely (empty social networks).
I share his anger.
Well, fuck them, and fuck that.
Wednesday, June 26th, 2013
This is a really nice and simple idea: view photos from a specific place taken at a specific time. Voyeuristic fun.
Sunday, November 18th, 2012
A nifty little mashup from Music Hack Day London 2012.
Sunday, December 25th, 2011
Steven Johnson describes the beautifully chaotic way that ideas collide and coalesce. Oh, and this bit…
Listening to Cerf talk about the origins of the Internet — and thinking about the book project — made me wonder who had actually come up with the original idea for a decentralized network. So that day, I tweeted out that question, and instantly got several replies. One of those Twitter replies pointed to a Wired interview from a decade ago with Paul Baran, the RAND researcher who was partially responsible for the decentralized design.
Monday, September 26th, 2011
I never expected to see a cross between responsive design and AR, but here ya go:
A silly mashup of HTML5 technologies: We use the canvas to capture the contents of a video element. The canvas then identifies the blue markers and overlays an iframe on top of it. The iframe contains our website (upperdog.se) which has a responsive design.
Sunday, September 25th, 2011
What if Mario had a portal gun?
Tuesday, November 23rd, 2010
A visual representation of each track on the new Girl Talk album.
Monday, October 18th, 2010
A great write-up of the latest additions to the Guardian's Open Platform API including a lukewarm assessment of Semantic Web technologies like RDF.
Monday, October 4th, 2010
An inspiring presentation by Tom Armitage on the value of open data.