[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: The new Google "Go" language - with n=0..N buffered blocking channels
On Thu, Nov 12, 2009 at 12:37 AM, Rick Beton <rick.beton@xxxxxxxxx> wrote:
> 2009/11/11 Teig, Oyvind UTCFS <Oyvind.Teig@xxxxxxxxxxxxxxxx>
> Interesting. A new version of C which is a big improvement on C, albeit
> still with dependence on reference (/pointer) aliasing.
> http://en.wikipedia.org/wiki/Go_%28programming_language%29 :
>> The syntax of Go is close to that of C except for the type declarations;
>> other syntactical differences are the missing brackets around for and if
>> expressions. The language supports garbage collection. The concurrency model
>> of Go is modeled after Tony Hoare's CSP, like previous concurrent
>> programming languages such as occam or Limbo, but also has features of
>> the Pi calculus such as channel passing.
>> Features not present in Go include exception handling, type inheritance,
>> generic programming, assert and method overriding. Of those, Google are
>> still having an open mind about generic programming, while assertions are
>> argued against in the language FAQ, and the choice of no type inheritance
>> defended. Unlike Java, maps (a.k.a. hashes or dictionaries) are built-in
>> like strings.
> The concurrency introduction is here:
Out of curiosity I threw together a quick commstime.go benchmark, and
my initial impression is that its roughly 4 times slower than KRoC for
the same. I'm doing quite a bit of playing with it now, and the most
peculiar thing I'm running into is the way channels get closed.