Internet whimsy: I'm looking for more examples of quirky web and computing stuff, here's what I have so far

Part of my day job involves writing about computer science (in its widest sense) for the Computer Science For Fun (CS4FN) blog and print magazine(s). We write about computer science research, history of the topic, themed stuff, how to do something or other - all sorts. 

I'm also a big fan of the Really? No, Really? podcast in which the hosts (Jason Alexander and Peter Tilden) have a guest on each week to talk about something that's amazed, amused or baffled them and sometimes all three. 

There are lots of stories in the world of computing that do the same and I thought I'd write about some of them, but while I know lots (below) I'm sure there are others I've not heard of yet. 

What am I missing in the list below of quirky computer science type things? 

• Lava lamps and security - Cloudflare trains a camera on a wall of lava lamps and converts the images (incredibly random!) into a string of numbers to help with encryption.
  
  
• "Open Office can't print on Tuesdays" - a glitch which meant that if the word Tue appeared at a particular point in a line of metadata then the file was miscategorised and treated as a non-printing file. The fix was to replace all instances of 'Tue' with 'tue'.
  
  
• "Can't send email further than 500 miles" - a university stats team in North Carolina found that they could only email within a 500 mile radius. The culprit was the length of time allowed for a ping message and its response. It didn't time out when emails were sent 'locally' but the message took too long once the distance was over 500 miles so the emails failed.
  
  
• The UK Government website for Bank Holidays displays bunting if you visit it on the day of a Bank Holiday (on Christmas Day it displays tinsel).
  
  
• IP over avian carriers - an April Fool's joke written as a serious document discussing the transfer speed of data sent by pigeon. The real-world equivalent is river rafting tourism which used carrier pigeon to return camera rolls to the head office which could then be processed for guests to collect once the trip ended.
  
  
• Google's Pigeon Rank - another April Fool's joke with Google explaining that pigeons were used to pick websites to display when a user searched for information. The real-world equivalent is the discovery that pigeons are pretty good at discriminating photographs of healthy tissue samples from diseased ones.
  
  
• Broadband over wet string - this one is peak internet whimsy ;) Engineers had a go at getting a signal to transmit over wet string and managed to get an impressive 3.5 Mbps. Althugh wet string is conductive apparently it's more about the medium acting as a wave guide for the high frequency signal.
  
  
• I need to dig out the nice story on Twitter of someone's mum being paid to try and test software to destruction - she first used social engineering to get into the room and the program by carrying an awkward amount of papers and struggling through a door then simply asking people "what's today's password".
  
  
• Optical mice don't always work in bright sunshine.
  
  
• The first webcam was pointed at a coffee machine so that people could check whether there was coffee left.
  
  
• The Utah teapot as the first item rendered digitally in 3D.
  
  
• Ada Lovelace and the first dry run table.
  
  
• HTTP status codes - many fairly dull, one or two demonstrate a bit of whimsy.
  
  
• List of animals awarded human credentials (formerly 'list of cats with fraudulent diplomas') - edging out of remit but computational in the sense of logic - testing the weak points of a system where people can get PhDs without doing the work, as demonstrated by people getting qualifications for their cats and dogs. To make it a bit more computing-relevant I might add something in to this post about web certificates and keychains etc.
  
  
• "What kind of medicine does Dracula take? Con medicine" - this baffling cracker joke can be explained by jokes written in fonts with a ligature then printed on devices with fonts that don't have it, meaning that the word 'coffin' loses the ffi as ligature fonts treat that as a single glyph. Bonus material: non-English orthography and how many foreign names and addresses used to get mangled when people entered them into the old CS4FN mailing list database.
  
  
• Les Horribles Cernettes - first picture on Tim Berners-Lee's interwebs. See also Annie Rauwerda / Depth of Wikipedia's TikTok about them.
  
  
• Mark Rober's dartsboard that uses Arduinos, computer vision / motion capture, sensors and rapid trajectory calculations to ensure the dart always lands you an excellent score.
  
  
• Please add your example of computerish internet whimsy to this form :-)
  

Downloading files via mild geeky skills

Every now and again I remember a song I heard years ago via 3wk.com (http://www.3wk.com) - I don't get to listen to music radio in the background these days as much as I'd like but sometimes I remember that I enjoyed hearing a particular song, and want to hear it again.

Generally I go straight to YouTube as most stuff's there, or Last.fm, or imeem.com (before it was bought by MySpace and shut down - I'm afraid that regardless of what Cory Doctorow has to say about MySpace it's not going to convince me to assault my eyeballs and visit it) or one of the other places. You can nearly always hear a snippet of a tune at amazon.COM although not so often at amazon.CO.UK and iTunes is usually worth a visit.

The track I was after is from - Antarctica - a band that apparently lasted a year, releasing one EP and an album and were last heard from in about 1998. I'm not a particular fan but I really liked the beginning of the song, Hallucinus, which was this electronic synth thing that then went a bit indie. Quite annoying whiny vocals come in a bit later too.

Anyway, bit of googling later and I found that I could listen to it here http://pizza.saur.us/ but what I really wanted to do was to download it. The format of the 'mp3 listening bar' on that page is used on plenty of other pages and usually lends itself quite well to downloading an mp3 with a 'right-click, save as' manoeuvre. However on this occasion I was thwarted by Flash.

More generally some pages don't make the downloading (or 'playing in another window') link obvious - perhaps deliberately but probably not - but you can usually uncover the file's real, individual URL by clicking VIEW / VIEW SOURCE - this usually gives you the full html coded text for the entire page you're looking at, including all the links and sources of files.

I was quite pleased that I managed to download the track by searching within this 'source code' for the word 'Hallucinus', which appears a handful of times, then selecting the relevant string of text that looked most like an URL or filepath, which was this:-

http%3A%2F%2Fpizza.saur.us%2Fdelivery%2F%3Ff%3DAntarctica%20-%20Hallucinus%26e%3D1

Punting this over to a fresh browser window didn't work because all of the %xyz stuff - I don't really understand what this is all about, and notice it only as an annoyance that crops up every now and again that needs trimming out.

I think it's pretty obvious that %3A is going to be : and that the two %2F strings are going to be // but I wasn't too sure about the rest, so I googled %2F and found http://en.wikipedia.org/wiki/Percent-encoding which told me what should replace the %bits. Quite a lot like code really, as I just looked up %3D and found that it should be the = sign.

That gave me this:-

http://pizza.saur.us/delivery/?f=Antarctica+-+Hallucinus&e=1

Once the URL was reconstituted correctly I put that in a new window and it automatically downloaded as the mp3, and is saved to my desktop - listening to it now.

Another thing to be aware of is that the filepath might include only the bit after the domain name, eg /deliver/?f=... in which case you need to manually add in the domain name yourself (http://....) so it'll work.

I've collected a few files in this way, although websites that hide much of their code in Flash etc make this a bit harder.

Probably the only reason I know this stuff is that I played around under the bonnet of html code a bit in the early 90s and picked up / absorbed some helpful skills from doing so, while creating basic web pages. It had really never occurred to me that everyone else didn't tweak URLs to suit themselves until I read this article by Snyder Consulting, called "Seven tricks that web users don't know" http://www.snyderconsulting.net/article_7tricks.htm

Anyway, like Ray Mears I'm hoping to keep the old skills alive :)