On Thu, Mar 6, 2014 at 1:36 PM, Jochen Theodorou blackd...@gmx.org wrote:
But if and when you have such cases, you better have a
language that produces java source code, as John said.
Maybe that's an option for Frege, but not for Groovy.
Is there a technical reason why not? Goto can
On Thu, Sep 15, 2011 at 10:43 AM, Ingo Wechsung
ingo.wechs...@googlemail.com wrote:
Frege is not a Haskell dialect, though I put some effort to make it look and
feel like so.
It looks very much like a Haskell dialect to me, just judging from the
number of Incompatibility with Haskell notes in
On Wed, Aug 10, 2011 at 10:29 AM, Robert Fischer
smokejumpe...@gmail.com wrote:
Can you define contra- and co- variance for me in the context of
structural (not nominal) typing?
Whether a type system is structurally or nominally typed is orthogonal
to whether it has subtyping. Algol 68
On Wed, Aug 10, 2011 at 12:16 PM, Robert Fischer
smokejumpe...@gmail.com wrote:
What does subtyping mean in a prototype language with structural
types? There is no such thing as a parent or child type, so the
type hierarchy context — the only context where subtyping means
something to me —
On Tue, May 17, 2011 at 9:15 AM, Alexander Bertram
a...@bedatadriven.com wrote:
My question for the JVM-niks out there is how feasable is it create a
new jvm class for each thunk or function value?
When you write a function in scala or clojure, does this get compiled
into a new, separate
On Tue, May 17, 2011 at 1:46 PM, Charles Oliver Nutter
head...@headius.com wrote:
Optimal for both size and performance is to generate the bodies into
static methods as above, but use either unique stub classes as the
function objects [...].
Does this mean that there are Foo, Bar, and Baz
On Tue, May 17, 2011 at 2:38 PM, Charles Oliver Nutter
head...@headius.com wrote:
Yes, that's what it means. If the stubs were generated ahead of time,
then you're right: it's no better than a class-per-function. However
in JRuby we generate the stubs at runtime, so you can batch-compile a
On Wed, Feb 2, 2011 at 5:24 PM, Brian Hurt bhur...@gmail.com wrote:
This isn't the right venue for this question, but I'm not sure where else to
ask, and I'm hoping that someone here might be able to help. Basically,
I've got a program that, for long involved reasons, creates lots of sockets
On Wed, Feb 2, 2011 at 5:47 PM, Robert Fischer smokejumpe...@gmail.com wrote:
Wait—let me get this straight. Even if you close a socket, you've
still consumed a port number, and there's no way to free the port
numbers for re-use?
No, not at all. If the socket has been closed long enough,
On Fri, Nov 19, 2010 at 12:54 PM, Charles Oliver Nutter
head...@headius.com wrote:
On Thu, Nov 18, 2010 at 11:02 PM, John Rose john.r.r...@oracle.com wrote:
Suggestion: Make the close behavior optional, per wrapper.
Only set the mustClose bit on channels created via Ruby APIs.
Yes,
On Sat, Sep 11, 2010 at 11:55 AM, Jim Baker jba...@zyasoft.com wrote:
Remi, you may be right - maybe we just need to continue to generate naive
bytecode and let the JVM sort it out. Just somewhat better naive bytecode,
that's all.
It's always good to remember that Java JITs are designed to
On Mon, Sep 6, 2010 at 11:33 AM, Per Bothner p...@bothner.com wrote:
As an old curmudgeon, I still prefer a hard-written recursive-descent
parser. It's simple, efficient, and much easier to debug.
I strongly agree. Lua, where code compactness is almost everything,
started out with a yacc
On Sat, Jun 5, 2010 at 12:27 PM, Charles Oliver Nutter
head...@headius.com wrote:
I also see something else I find interesting: EObject apparently
implements dec() and is_ge()...I'm guessing you have implemented a
full set of numeric operators on EObject so that when they're
explicitly
On Thu, Jun 3, 2010 at 4:57 PM, Charles Oliver Nutter
head...@headius.com wrote:
Sorry, I meant BigInteger above. Perhaps that changes the situation?
At any rate, I'll reply as though we both said BigInteger.
Right, that's what I meant.
Wow, I'd be very surprised if that's true.
On Thu, Jun 3, 2010 at 2:17 AM, Charles Oliver Nutter
head...@headius.com wrote:
Another data point for JRuby.
JRuby has only two integer data types: Fixnum (implemented by
RubyFixnum, always containing a long) and Bignum (implemented by
RubyBignum, using BigDecimal). All Fixnum math
On Wed, May 26, 2010 at 7:35 AM, Attila Szegedi szege...@gmail.com wrote:
Out of professional curiosity: how do you implement
1. Ruby's here docs
2. Ruby's %Q, %q, %x and %r constructs
with any LL(k) or LALR parser generator's grammar language?
That's what lexers are for.
--
GMail doesn't
On Fri, Dec 18, 2009 at 4:29 AM, Miles Sabin mi...@milessabin.com wrote:
On Thu, Dec 17, 2009 at 10:36 PM, Jochen Theodorou blackd...@gmx.org wrote:
Charles Oliver Nutter schrieb:
[...]
I'm not even sure a common AST is required. Maybe I'm not making
myself clear enough?
but then what of
On Fri, Dec 18, 2009 at 3:28 PM, Per Bothner p...@bothner.com wrote:
Tail-call optimization is not a pure optimization, as it changes the
semantics in a fundamental way: It removes stack traces, which are
part of both observable behaviour (stack traces) and the Java security
architecture.
On Fri, Dec 18, 2009 at 7:19 PM, Charles Oliver Nutter
head...@headius.com wrote:
I figured I'd do a feasibility study on whether JDI could be used to
implement ObjectSpace (Ruby's feature for walking all objects on the
heap of a given class).
Wouldn't maintaining a weak set initialized by the
On Fri, Dec 18, 2009 at 6:21 PM, fo...@x9c.fr fo...@x9c.fr wrote:
I do wholeheartedly agree ; indeed for (direct) tail recursion, we need
nothing new: a goto with the first instruction of the method as the
destination
does the trick.
Unless you are generating Java source rather than
On Fri, Dec 18, 2009 at 7:31 PM, Charles Oliver Nutter
head...@headius.com wrote:
That's what we do now, but I'd like to just rip that code completely
out so we don't have to check a flag should I add myself or not.
The WeakSet should do that for you, no? That's the point of sets.
SetE
On Fri, Dec 18, 2009 at 8:46 PM, Charles Oliver Nutter
head...@headius.com wrote:
That's not quite what I meant...I meant we already have logic in JRuby
to avoid adding transient object constructions to ObjectSpace, so all
other constructions have to check a flag to know they should add
On Thu, Dec 17, 2009 at 12:34 PM, Charles Oliver Nutter
head...@headius.com wrote:
You bet your bippy I would. I've been wishing for something like LLVM
for JVM bytecode for a long time. I've been meaning to look at Janino
to replace our current compiler backend, but have been spending more
On Thu, Dec 17, 2009 at 1:24 PM, Charles Oliver Nutter
head...@headius.com wrote:
On Thu, Dec 17, 2009 at 12:17 PM, John Cowan johnwco...@gmail.com wrote:
I proposed a long time ago to add goto to Janino, which would remove
a huge amount of the incentive to generate bytecode directly
On Thu, Dec 17, 2009 at 3:49 PM, Matt Fowles matt.fow...@gmail.com wrote:
I am confused... The fact that I don't use Janino's parser doesn't
theoretically stop me from bootstrapping the compiler. In fact, Janino's
parser is only useful if the language I am compiling is actually Java. If
my
On Tue, Dec 8, 2009 at 9:53 AM, Jon Harrop j...@ffconsultancy.com wrote:
If you want JVM+TCO then you should already be able to use OpenJDK. Another
forthcoming option may be VMKit, which is a JVM and CLR built upon LLVM
(which provides TCO so it should be relatively easy to add if it has not
On Mon, Dec 7, 2009 at 2:19 PM, Jon Harrop j...@ffconsultancy.com wrote:
The JVM is incapable of expressing structs so no JVM-based language can
support them. CLR-based languages supporting structs (e.g. C# and F#) simply
delegate the work to the CLR.
The obvious approach, though I haven't
On Mon, Nov 16, 2009 at 5:27 PM, Charles Oliver Nutter
head...@headius.com wrote:
* Dynamic languages on the JVM have to use boxed numerics most of the
time, which means that we're creating a lot of numeric objects. Some
of these may be nearly free, immediately-collectable. Some may be
On Oct 6, 2:42 am, John Rose john.r...@sun.com wrote:
3) Can only the coroutine itself yield (shallow), or can any
subroutine invoked by the coroutine yield on its behalf (deep)?
To do samefringe (and IMO be useful) you need deep coroutines. And:
If coroutine C1 is going to call
On Sat, Sep 26, 2009 at 1:41 AM, Charles Oliver Nutter
head...@headius.com wrote:
Ok, here's a classic challenge for those of us on the JVM: coroutines.
In order to know what you need, you need to say what *kind* of
coroutines these are. The Lua people have three binary features that
a given
On Tue, Jun 30, 2009 at 5:30 AM, Attila Szegediszege...@gmail.com wrote:
I'd be delighted to see someone smarter than me
overcome these problems for having an *efficient* Haskell on the JVM
implemented *elegantly*. That'd really rock.
That might be too much to ask, given that GHC is an
On Tue, Jun 30, 2009 at 7:28 AM, Patrick Wrightpdoubl...@gmail.com wrote:
See also PDFs linked at bottom of WP entry, in particular CAL for
Haskell Programmers.
What this boils down to, AFAICT, is that 100% CAL code is pure, but
CAL that calls Java stuff is not.
Regards
Patrick
Cowan wrote:
On Sat, May 9, 2009 at 11:06 AM, Robert Fischer
robert.fisc...@smokejumperit.com wrote:
John Cowan wrote:
Depending on your schema of call types, you'll want to choose between
grouping call types on interface types vs. one call type per interface
type.
Since the arguments and returns
On Wed, May 6, 2009 at 11:38 AM, John Wilson tugwil...@gmail.com wrote:
2009/5/6 John Cowan johnwco...@gmail.com:
In my language, I need objects which represent Java static methods.
How about:
public abstract class Function3 {
public static final Function3 foo = new Function3
On Thu, Apr 23, 2009 at 12:52 PM, Jon Harrop j...@ffconsultancy.com wrote:
http://neopythonic.blogspot.com/2009/04/tail-recursion-elimination.html
Not worth reading. Guido has absolutely no idea what he is talking about and
is probably the last person on Earth you should listen to in this
On Fri, Apr 10, 2009 at 5:41 AM, Attila Szegedi szege...@gmail.com wrote:
I'm replying more-or-less randomly to this message among all messages
debating whether TCO is a JVM implementation detail or not.
Well, thanks for the implicit compliment.
My stance is that it is, in the same sense as
On Thu, Apr 2, 2009 at 10:36 AM, kirk kirk.pepperd...@gmail.com wrote:
why do they have to be exposed? Isn't tail recursion and implementation
detail? And an optimization at that?
No.
Being able to rely on tail recursion is not an implementation detail;
Scheme programmers routinely write
On Fri, Mar 20, 2009 at 9:35 AM, Frank Wierzbicki fwierzbi...@gmail.com wrote:
The first time Jython is started, it walks through all of the jars and
class directories on the filesystem that it knows about (including,
for example classes.jar, the jars in the ext dir, etc) and creates a
On Tue, Mar 10, 2009 at 10:40 AM, Jochen Theodorou blackd...@gmx.org wrote:
Is there any material about plans to add ephemerons to the JDK? Using
google I wasn't able to ind anything. Only that thread is takling about
it seems.
It seems to me that it should be possible to simulate ephemerons
On Tue, Feb 24, 2009 at 9:03 AM, Robert Fischer
robert.fisc...@smokejumperit.com wrote:
What optimizing efforts are worthwhile on generated bytecode given the
optimizing capabilities of
the JIT? I'm having trouble finding an explicit or definitive list of
optimizations, but trying to
be
On Tue, Feb 24, 2009 at 6:31 PM, Matt Fowles matt.fow...@gmail.com wrote:
Given that you are generating code, do you generate java directly or
generate an AST?
From an AST, I suppose is the right answer. But it's a Lisp-family
language, so the surface syntax is already a tree -- I just keep
On Mon, Jan 12, 2009 at 12:58 AM, Ted Neward ted.new...@gmail.com wrote:
Is that out of concern that Jeroen will object, or because it might cause
confusion between the two projects?
The latter. If you google for ikJVM, you get about 208 hits, but
right at the top Google asks Do you mean
I'm implementing a language that has dynamic types, multiple
inheritance with ordered superclasses, and multimethods. I'm writing
a whole-program compiler. The class structure and available methods
are fixed at compile time, and I'm trying to figure out the best way
to generate dispatchers
On Tue, Apr 29, 2008 at 5:36 PM, Jochen Theodorou [EMAIL PROTECTED] wrote:
I wanted to collect a bit data to how you avoid boxing in your language
implementations.
My language provides bignums and flonums, which I simply represent as
BigIntegers and Doubles. I pay the boxing penalty, but
On Sat, Apr 26, 2008 at 9:33 AM, Patrick Wright [EMAIL PROTECTED] wrote:
That said, I'm a big fan of VM Vendors,
esp. that one film about angels hanging out in Berlin.
/me agrees about the movie, but defenestrates P.W. anyway, on
principle and to preserve the purity of the Internet's
On Tue, Apr 22, 2008 at 12:21 AM, hlovatt [EMAIL PROTECTED] wrote:
That allowed keywords to be escaped by surrounding them with an
underscore, e.g. the Java keyword class could be escaped via _class_.
And _class_ would be escaped as __class__, and so on?
--
GMail doesn't have rotating
Could someone consider filing a bug against Java asking that
Method#invoke return either Boolean.TRUE or Boolean.FALSE when calling
a boolean method, rather than boxing the return value into a freshly
allocated java.lang.Boolean? That would make life much simpler for
those of us who need to
On Jan 17, 2008 1:06 PM, Charles Oliver Nutter [EMAIL PROTECTED] wrote:
* Then we used a hand-written indexed method handle like you describe.
Again, it worked (albeit a bit slower than individual methods), but it
was too much effort to implement by hand and wouldn't work for generated
code.
On Dec 29, 2007 9:05 AM, Ola Bini [EMAIL PROTECTED] wrote:
I'm looking at implementing a language that would be best represented
using CPS. Now, I can't really find any references to full CPS
implementations on the JVM anywhere. I'm looking at using the technique
Kawa is planning for
On Dec 22, 2007 6:38 AM, Charles Oliver Nutter [EMAIL PROTECTED] wrote:
I don't think it's directly relevant, unless John C is using exceptions
for flow control.
The facility is used under the covers for flow control, but is also
exposed directly to the user. I'm planning, based on what you
On 11/7/07, Jochen Theodorou [EMAIL PROTECTED] wrote:
well there is at last the problem that you will need root rights to
start the server. Root rights are to be avoided as much as possible.
that's a general rule. If it is not possible, then there must be a very
good reason for this or no
F# was a research project. I doubt there would be enough research-fu
in just doing it again for the JVM, which is after all very similar.
On 11/1/07, Matt Hellige [EMAIL PROTECTED] wrote:
On Nov 1, 10:01 pm, Jon Harrop [EMAIL PROTECTED] wrote:
Could a CAML derivative language for the JVM
Just so.
function plustwo (i : integer) : integer
begin
plustwo := i + 2
end
On 10/29/07, Jochen Theodorou [EMAIL PROTECTED] wrote:
John Cowan schrieb:
On 10/29/07, Jochen Theodorou [EMAIL PROTECTED] wrote:
I did read his text as that Tennent was against using multiple exit
points
53 matches
Mail list logo
