Google programming language
Mister2u
Why Tea
> http://golang.org/
Interesting. It's still an experimental language that
runs on Unix only. For a quick view of the language,
the FAQ is a good place to start:
http://golang.org/doc/go_faq.html.
/Why Tea
Zombie Elvis
<Yrra...@secret.cult.invalid> wrote:
>Mister2u <mist...@gmx.com>:
>
>> http://gola...
>
>And your point is what?
>
Let me try to guess. It's a new, fast-compiling, free programming
language that promises to be easier to write than traditional
languages like C and C++ while avoiding a lot of the gotchas that
those languages have which promote sloppy code. And it's from Google
so you know they won't get bored and abandon it if it's not successful
-- uh oh.
--
Obama: Hey Ballmer, you mind if we borrow 90% of the world's computers for a quick cyber war?
Ballmer: Finally, the moment I've been waiting for! *Throws ceremonial war chair at wall*
- Seen on Slashdot
Roberto Castillo
roberto...@ameritech.net
http://mind-grapes.blogspot.com/
http://zombie-gulch.myminicity.com/
Craig
> And it's from Google
> so you know they won't get bored and abandon it if it's not successful
> -- uh oh.
<grin>
--
-Craig
Mark Carter
I think it's difficult to get excited about the language. Forth, Lisp
and Smalltalk are ground-breaking languages that are really cool to
try out. Go doesn't really seem to have anything that make you go
"wow".
I'm slowly getting to grips with Microsoft's .Net framework, where I
can create UIs very quickly, and write code. You can do a lot with it
- but like all things Microsoft, seems to be a boatload of bloat. So I
can see why Go might be attractive. They've also got some nice
libraries going for doing networking and suchlike.
Why Tea
It's early days yet for GO, and it doesn't look like running
any time soon on Windows. From the FAQ, it seems to me that
the language wasn't designed to be groundbreaking. Instead, it's
meant to be a lighter OO, especially in the "type" usage. It's
probably closer to C than C++, but with OO features.
Mark Carter
> It's early days yet for GO, and it doesn't look like running
> any time soon on Windows. From the FAQ, it seems to me that
> the language wasn't designed to be groundbreaking. Instead, it's
> meant to be a lighter OO, especially in the "type" usage. It's
> probably closer to C than C++, but with OO features.
I think it's a bit of a shame that they decided that they needed to
invent yet another syntax. Couldn't they have tarted up something like
Obj-C to suit whatever it is they were trying to achieve? Imagine
having a language that could be compiled as a system language, but
also run interactively through an interpreter. You could then do stuff
like write and test drivers interactively, make the debugger part of
the language, and so on. I know you're going to say, "congratulations,
you've just invented Smalltalk"; but it seems that the guys doing Go
are missing a great opportunity, and giving us yet ANOTHER programming
language.
Why Tea
Debugging system software via an interpreter?? That's something
new to me.
Have you ever wondered why Smalltalk has been used mainly as
an OO teaching tool and hardly used for anything in the real world?
Mark Carter
> Debugging system software via an interpreter?? That's something
> new to me.
The advantage of having a programming language that can understand
itself and have a REPL is that you can do these neat kind of things.
Lisp will drop in to the debugger if it encounters a problem. You can
do things like examine the stack trace. I've tried it in CLISP, and
it's a bit of a confusing mess, to be sure. It also tends to optimise
some variables away, which hampers debugging. Python, on the other
hand, is much friendlier. Most of the time it will just crash out -
but it does give much more meaningful descriptions as to where the
thing went wrong. You can set breakpoints by using Python code itself.
If you don't want to do that, you can just let it run and crash, and
use the post-mortem debugger. The important point is that all this
stuff is in the language itself, you don't have to run separate tools
like the GNU debugger.
> Have you ever wondered why Smalltalk has been used mainly as
> an OO teaching tool and hardly used for anything in the real world?
Perhaps a good answer to the question would be: for the same reason
that people use Windows over Linux. Linux gets you most of the way
there, but it too often feels like a kit of parts. With missing
pieces. And sharp edges ;) The free Smalltalks generally lack good
documentation - which is an absolute killer for a language. Libraries
are often incomplete or abandoned. The same kind of problem exists
with the Lisps, with weird incompatabilities, bit rot, and so on.
Compare that to something like Ruby or Python, where if I want to do
client-side web fetching I can Google around, and come up with simple
working examples. It gets the job done with minimum fuss.
Surely, though these aren't language issues, but implementation
issues. That's why I say that it might be better to fix whatever it is
that is perceived to be wrong with current implementations, rather
than invent yet more new syntaxes. I doubt that Smalltalk or Scheme
takes particularly long to parse. If anything, I'd expect they'd be
relatively fast to parse, on account of the simplicity of their
syntax.
Why Tea
I'm a bit lost here. I think you have mixed up the scripting
(Ruby, Perl, Tcl, Python, Javascript), interpretive (Java, Smalltalk)
and other high level compiled languages (C/C++). Although
the lines are a bit blur over time, but they are different classes
of languages that serve different purposes.