Make3D

I went to a really cool talk this morning. It was given by Ashutosh Saxena on his work on Make3D, which is a tool to generate 3d models from a (single) digital picture. Rather fashionably he uses machine learning techniques to build the tool. The results are pretty impressive. You can even upload images yourself to their website!

in rainbows in east london

LAST NIGHT\ SO GREAT

LAST NIG_HT SO GREAT

LA\ST NIGHT SO\GREAT

LAS_T NIGHT SO G_REAT

LAS\T NIGH\T SO GRE\AT

L_AST NIGHT_SO GREAT

Streets of Cambridge

One of the fun things about living in Cambridge (like many University cities) is that you get to see some unusual things. In fact, many "normal" things are probably quite unnormal elsewhere; e.g. walking my daughter to school (in of itself quite unnormal for many schools I guess) we are routinely passed by a chap on a unicycle. (My guess is that he works in the Science Park.)

Anyhow, seeing this on Chesterton Road on Friday afternoon prompted me to jump off my bike and take a snap with the phone.

Asus eee 901

There's a nice review of the new Asus eee 901 on The Register [here]. It seems that the concerns I had when I tried the 701 have been addressed, viz. better screen, and improved trackpad+buttons. However £319 is quite a bit of money as I was thinking of it as a laptop for the kids...

OCaml/F# vs SML

I've programmed in SML for many years. However, readers will know that recently I have been using F# quite seriously. I'm very impressed. It's really a very productive language. One minor problem for me are the subtle (syntactic) differences between the two. For example, I still find my fingers wanting to type [x,y,z] for lists instead of [x;y;z].

[I still don't get why people would want to drop the brackets around a tuple?! This is probably a hang-up from the pain that adding this feature caused me when I wrote the Imperial Hope Interpreter when I was an undergraduate! (Gosh - this was in 1990. Get over it!)]

Anyhow, I just stumbled across Adam Chlipala's OCAML/SML comparison page, which I enjoyed reading. [here] Check it out.

Dynamic Rebinding: journal paper finally out

The paper on dynamic rebinding with Peter, Mike, Gareth and Keith has finally appeared in Journal of Functional Programming (here). [You may or may not be able to view it for free from the JFP page - depends where you're reading it from.]

This is a great relief! Contained within are details of Gareth's mega-theorem (which proves equivalence between CBV lambda-calculus and our funky late-binding calculi (lambda-r and lambda-d)). Gareth sweated blood over these!

Enjoy!