Google's got a new language... Go...

[Lennart Sorensen]

On Wed, Nov 11, 2009 at 11:12:19PM -0500, Christopher Browne wrote:
> I couldn't care less that it's not "abjectly object oriented."
> 
> Evidence seems to be growing that that's a design "dead end,"
> encouraging both details for the sake of details and preventing use of
> intelligent approaches to concurrency.
> 
> It seems to me like it's got something of an admixture of:
> a) Icon-ish coroutines
> b) Objective C-ish "protocols" (aka interfaces)
> c) It *isn't* filled with the irritating "OO" thing of "oh, you need
> to define an ontology of the types you intend to use before you can do
> anything!" combined with "oh, everything needs to inherit from
> something!"
> d) Elimination of the irritations of cpp
> e) Like Modula-3, it has full fledged garbage collection built in by
> design, so you're not rewriting reference counters everywhere (ala
> C++)

So why not use modula3 or maybe ocaml or something?  Why invent something
new, and especially something with syntax that is almost as ugly as perl.
Perhaps uglier in fact.

> f) channels (which are "streams/queues", a synchronization mechanism
> particularly useful for communicating between coroutines) look MIGHTY
> interesting, and were a Limbo thing
> 
> One of the claims is that the design of the language makes dependancy
> management easy.  That's a mighty interesting claim; it's not
> self-evident how this is made so.
> 
> Keep in mind: The people that designed this are the same people that
> originally designed Unix.  I trust their sense of "taste" quite a bit.

Ehm, well.  Decent internal structure, not so nice an interface perhaps.

-- 
Len Sorensen
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists