Here’s a screenshot of a link being previewed in the new Interclue 1.5 Beta, which you should install if you haven’t already.
See the hint next to the magnify cursor you get when you mouseover a thumbnail in a the new clueviews? When you mouseover the comics in XKCD you can usually see a little in-joke in the title hint. Sometimes, the titles are pretty long, and Firefox the tooltips don’t wrap. This has been a subject of some complaint, but thankfully, with the landing of the infamous reflow branch (count the dependencies!), this will be resolved in Firefox 3.
Yep, in the 1.5 Clueviews, images of sufficient size are thumbnailed, you can click to see the image full size, and then click again to go back from there. It’s pretty cool, and we think it’s a good compromise to showing no images, or a full size window with all the images, like Cooliris does (frankly, if I wanted to see the whole page, I’d just open a new tab).
I love this cartoon, and I wish he’d put it on a T-shirt, because Turing featured in one of my favorite assignments [1] while I was at UC, where I majored in Philosophy and Computer Science.
I was never a great academic, far too scatterbrained most of the time, and being up till 3am most nights running my BBS, playing Civilization, or online backgammon (FIBS 1700+, back when that was actually pretty good), sure didn’t help, but I do have my name in the credits of one academic paper, “On Alan Turing’s Anticipation of Connectionism”, because our logic lecturer, who later went on to become co-director of the Turing Archive, discovered a mostly ignored paper by Turing from 1948, in which he definitely did anticipate “Neural Nets“, which is pretty amazing given that at the time there was only one non-specialized computing machine on the entire planet, ENIAC. [2]
Unfortunately, Jack couldn’t figure out how to make Turing’s “B Type” networks actually compute something, so in a brilliant combination of laziness, sadism and cunning, he gave his 3rd year students the option of skipping one our 20% take home assignments for the year, and instead figuring out how to make Turing’s type B networks work.
It was fun trying. In the end, I was one of about 3 students who instead succeeded in making a logical proof that these networks couldn’t work, (as specified by Turing, anyway). As I recall, I had to prove it 3 times, the last by structural induction, because Jack couldn’t accept that Turing might have got it wrong. He gave in eventually and I got full marks.
I put this achievement at the top of my list of 10 things I’ve done you probably haven’t, back in Feb 2005. Number 10 on that list was in fact Interclue, which was in the prototype stages even back then. It took a bit longer than expected to actually get it out, to same the least. I blame Hofstadter’s Law. But I’m really happy with the new beta so far, although it’s going to need some testing and tweaking, and we’d love your feedback. So if you’ve got firefox, go for it!
[1] And one of my favourite books, Cryptonomicon by Neal Stephenson
[2] Although others were under construction, and Turing had a hand in two of them – see Jack’s “A Brief History of Computing” for details).
