Saturday, December 31, 2011

PL Predictions for 2012

Summary: another year of broken lexical scope and broken type systems.
  • There will be a new language with broken lexical scope (remember, it's a criterion for success).
  • A GWT-like, 80's style, object-oriented GUI framework will make Dart a major player in web apps.
Happy new year!

Wednesday, December 21, 2011


Call-by-push-value has a seducing PR message: Science is reductionism. Once the fine structure has been exposed, why ignore it?

Unfortunately, I'm too unsophisticated to be able to gain much insight from the CBPV papers. There's an introduction to CBPV for more mortal-like folks. Other material is hard to find.

What caught my eye was the following code sample from Paul Blain Levy's (rhymes with "all plain Stevie") PhD thesis:

thunk(print "hello1"; λz. print "we just popped" z; ...)

So, apparently you can do stuff in a function before you're even taking arguments. Interested!

Maybe one of you can provide more information about this interesting new thing?!

Looking back at predictions for 2011

Let's take a look at my PL Predictions for 2011:

▶ JSON will be the default format for new internet APIs.

I'm not sure how to corroborate this one, but I think it's true. The most significant piece of evidence I have is that IBM even developed JSONx, a standard for representing JSON as XML.

▶ As more people use JSON, we'll see a XML renaissance, as we - for the first time - discover that XML actually gets some things right.

Didn't really happen, AFAICS. But there's this post: Why XML won’t die: XML vs. JSON for your API which says: "not everything in the world is a programming-language object". +1 to that.

▶ GHC will get typed type-level programming and end this highly embarrassing state of untypedness.

▶ We'll have a verified compiler-and-OS toolchain for compiling and running some kinds of apps. It won't be x86-based.

Unfortunately, I didn't really track this issue in 2011. There's SAFE, and maybe the O'Caml folks have something to show.

▶ All kinds of stuff targetting JavaScript.

Yeah, well, that was easy. My favorite is Emscripten.

▶ Split stacks and maybe some scheduler improvements will be shown competitive with green threads in the millions-of-threads!!1! (anti-)scenario.
Nobody did this AFAIK, but I still think it's true. Maybe next year.

▶ some insanely great new PL for 2011

So, although most of the predictions were rather cautious, I'm satisified with the outcome.

Monday, December 19, 2011

what distinguishes a good program from a bad one

Bugs and Battleships by Edward Z. Yang:
One might even say that the ability to generalize behavior of specific tests to the behavior of the program is precisely what distinguishes a good program from a bad one. A bad program is filled with many, many different cases, all of which must be tested individually in order to achieve assurance.
(HT @dyokomizo)

Tuesday, December 13, 2011


One of my all-time favorite quotes:
Uniform Resource Locator is just the result of squeezing the term object reference through the IETF standardization process.