jython-dev — Discussion of developing and enhancing Jython itself(!) (not how to use Jython)

Re: [Jython-dev] Jython license and GPL incompatibility

[<ba...@di...>]

>>>>> "BB" == Ben Burton <be...@ac...> writes:

    BB> The license for Jython, as found on www.jython.org and in the
    BB> CVS, currently contains this GPL-incompatible clause.  CPython
    BB> has recently altered their license to make CPython compatible
    BB> with the GPL (http://www.python.org/2.0.1/).

    BB> Is Jython likely to follow suit?

The Python Software Foundation is trying to figure out a way to effect
the same licensing changes on Jython as were recently accomplished for
Python 2.0.1, 2.1.1, and 2.2.  IOW, GPL compatibility.

It's way to early to tell (which is why I haven't mentioned it here,
or even to Finn yet), but suffice to say that it would be A Good Thing
if we could accomplish this.  It might require Finn and Samuele to
sign some form of disclaimer to the PSF, but even that hasn't
determined yet.

Given the length of time it took to get a GPL compatible license for
(C)Python, I wouldn't expect anything any time soon.  But I'm hoping
that it will eventually be possible, without much effort from Finn
(who should be left to manage the technical side of things and not
worry about this nonsense. ;).  The goal is to eventually place Jython
under the same license as CPython 2.0.1, etc.

-Barry

Re: [Jython-dev] Jython license and GPL incompatibility

[Ben B. <be...@ac...>]

> - ReadlineConsole is neither GPL or LGPL

http://www.bablokb.de/java/readline.html, first paragraph: "It is distributed 
under the LGPL."  Although there should be no problem linking LGPL stuff with 
jython, it's the GPLed libreadline.so that gets linked along with it that 
concerns me.

> So in the distributed form, jython doesn't links with any (L)GPL code.
> If anyone installs bablok's Readline class and libreadline and changes
> the python.console option they are creating a derivative work which they
> do not have the right to distribute.

So what's to stop me taking the GPLed fastlib.so, making a new proprietary 
library superfastlib.so that links to fastlib.so but contains my 
modifications, and then selling superfastlib.so but requiring users to 
download fastlib.so themselves; thus I'm not distributing the GPLed code 
myself but I still get the benefit of having a proprietary derivative work as 
such.

i.e. I believe the claim you make defeats the purpose of GPLing a library.  
Of course I could be very much mistaken.  I am not at all well versed in 
licensing details.  I don't say any of this because I am confident that I'm 
correct; I merely say it for the purpose of opening (what I believe is 
necessary) discussion on the issues.

http://www.fsf.org/copyleft/gpl-faq.html#GPLPluginsInNF

This also suggests to me that the readline plugin to Jython cannot be used, 
i.e. sure, you can distribute support for the plugin but nobody should be 
allowed to use it (and hence there's no reason to distribute support for it).

> I have somehow considered ReadlineConsole to be similar to CPython-2.0's
> readline.c file. In itself readline.c did not require CPython to be GPL
> compatible.

As I understand it, at least the debian python2 release disabled readline 
support because of the GPL-incompatibility issue.

IANAL, in fact I know very little about legal matters; these are just my 
thoughts that I'm throwing in.  If you like I can raise the issue on 
deb...@li... and see what they have to say.

Ben.

-- 

Ben Burton
be...@ac...  |  ba...@de...
http://baasil.humbug.org.au/bab/
Public Key: finger ba...@de...

I don't feel a part of any kind of sisterhood. Again, it's the
most disappointing thing where I get criticized by women more than men
on how I play the piano. They find it offensive. I'm just going, well,
this is how I choose to express myself, so if you're truly a strong, 
independent
woman, then how could you possibly find me being a strong, independent
woman offensive?
	- Tori Amos

Re: [Jython-dev] Jython license and GPL incompatibility

[<bc...@wo...>]

[Ben Burton]

>One thing I should mention though is that there may be issues with readline 
>support (org.python.utils.ReadlineConsole) - Bablok's wrappers are LGPL but 
>they link with GNU libreadline which is GPL.  Thus (according to my 
>understanding which could be wrong), this would require jython - if built 
>with ReadlineConsole - to be GPL-compatible.

Perhaps, but that is not my understanding.

- ReadlineConsole is neither GPL or LGPL
- There is no LGPL or GPL code in the jython distribution.
- Jython only links with the readline (L)GPL code when option 
     python.console=org.python.util.ReadlineConsole
  is enabled.

So in the distributed form, jython doesn't links with any (L)GPL code.
If anyone installs bablok's Readline class and libreadline and changes
the python.console option they are creating a derivative work which they
do not have the right to distribute.

>Sorry, not trying to nitpick here but in packaging jython up for Debian one 
>must take care to ensure that everything is legal. :)

Absolutely.

I have somehow considered ReadlineConsole to be similar to CPython-2.0's
readline.c file. In itself readline.c did not require CPython to be GPL
compatible.

OTOH I'm all ears for input from the authors of the readline software or
from the FSF lawyers. If there is a real problem with including the
ReadlineConsole, I can easily rip it out.

regards,
finn

Re: [Jython-dev] Jython license and GPL incompatibility

[Ben B. <be...@ac...>]

Hi - thanks for your reply.

> >Is Jython likely to follow suit?
>
> I somehow doubt it.

Cool.

One thing I should mention though is that there may be issues with readline 
support (org.python.utils.ReadlineConsole) - Bablok's wrappers are LGPL but 
they link with GNU libreadline which is GPL.  Thus (according to my 
understanding which could be wrong), this would require jython - if built 
with ReadlineConsole - to be GPL-compatible.

Sorry, not trying to nitpick here but in packaging jython up for Debian one 
must take care to ensure that everything is legal. :)

Thanks,
	Ben.

-- 

Ben Burton
be...@ac...  |  ba...@de...
http://baasil.humbug.org.au/bab/
Public Key: finger ba...@de...

The man who sees both sides of a question is a man who sees absolutely
nothing at all.
	- Oscar Wilde

Re: [Jython-dev] Options for distributing a Jython package.

[<ti...@ca...>]

[ munch of stuff about improving docs -- more later. ]

-> On another note...
-> I'm curious about what your deliverable is, so I'm guessing others might
-> be as well. Is it something that should be added to:
-> 
->     http://jython.sourceforge.net/users.html

As you noted before, it IS there -- FamilyRelations, links through

	http://family.caltech.edu/

However, you're all welcome to check out the first pre-alpha release,
now all nicely tar-red up at

	http://chabry.caltech.edu/~t/transfer/FR-prealpha-1.tar.gz

Aside from idle curiousity, I would be interested in having people
check out the distribution situation.  Through the CVS, at

	http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/familyjewels/

you can take a look at the Makefile, which is used only for making
distributions, and that will give you a clue as to the rest.

Essentially, I created a 'startFR.java' file that runs a Python
interpreter, which then runs 'start.py' (the development env
start file) after initializing the Jython interpreter.  This is
the shortest & cleanest, although not the least kludgy, way I have
found of running things.

And, please, let me once more congratulate the Jython developers on
an absolutely wonderful product.  AMAZING.  Writing Java programs is
now a blast...

cheers,
--titus

Re: [Jython-dev] Jython license and GPL incompatibility

[<bc...@wo...>]

[Ben Burton]

>Hi.  I'm currently looking at packaging Jython for debian and I'm wondering 
>what the current status is with the Jython license.  At the moment it seems 
>the Jython license is incompatible with the GPL.

Correct, I think.

>According to the FSF website 
>(http://www.gnu.org/philosophy/license-list.html):
>
>The License of Python 1.6b1 and later versions, through 2.0 and 2.1. 
>This is a free software license but is incompatible with the GNU GPL. The 
>primary incompatibility is that this Python license is governed by the laws 
>of the State of Virginia, in the USA, and the GPL does not permit this.
>
>The license for Jython, as found on www.jython.org and in the CVS, currently 
>contains this GPL-incompatible clause.  CPython has recently altered their 
>license to make CPython compatible with the GPL 
>(http://www.python.org/2.0.1/).
>
>Is Jython likely to follow suit?

I somehow doubt it.

Not only must CNRI be prepared to release their JPython-1.1 code with a
changed license, but they must also be able to actually create a new release.
It is not my impression that any of the fine people working at CNRI have the
tools or experience to create a JPython-1.1.1 release.

Even if is 10 times easier for jython to change license that it was for
python, I still think it will be 100 times too much work for me to tackle.

regards,
finn

[Jython-dev] Jython license and GPL incompatibility

[Ben B. <be...@ac...>]

Hi.  I'm currently looking at packaging Jython for debian and I'm wondering 
what the current status is with the Jython license.  At the moment it seems 
the Jython license is incompatible with the GPL.

According to the FSF website 
(http://www.gnu.org/philosophy/license-list.html):

The License of Python 1.6b1 and later versions, through 2.0 and 2.1. 
This is a free software license but is incompatible with the GNU GPL. The 
primary incompatibility is that this Python license is governed by the laws 
of the State of Virginia, in the USA, and the GPL does not permit this.

The license for Jython, as found on www.jython.org and in the CVS, currently 
contains this GPL-incompatible clause.  CPython has recently altered their 
license to make CPython compatible with the GPL 
(http://www.python.org/2.0.1/).

Is Jython likely to follow suit?

Thanks,
	Ben.

-- 

Ben Burton
be...@ac...  |  ba...@de...
http://baasil.humbug.org.au/bab/
Public Key: finger ba...@de...

I don't have a computer. I stay away from the internet mainly because
I don't want to know what color underwear I was wearing during sound-check.
I just don't want to know.
	- Tori Amos

Re: [Jython-dev] casting

[Kevin B. <kb...@ca...>]

???
public class A
{
    public Object get()
    {
        return "Hello";
    }

    public void doit( String s )
    {
        System.out.println( s );
    }
}



$ jython
Jython 2.1a1 on java1.3.1 (JIT: null)
Type "copyright", "credits" or "license" for more information.
>>> import A
>>> a = A()
>>> o = a.get()
>>> o
'Hello'
>>> type( o )
<jclass org.python.core.PyString at 4729773>
>>> a.doit( o )
Hello
>>> type( "" )
<jclass org.python.core.PyString at 4729773>
>>>


jython converts Java String objects into python strings, and transparently
converts them back into Java strings if you call a method that requires
them.

You may have problems if you put them in a jarray, a Collection, etc., but
just passing Strings should work fine.

Or you may have problems if you want to do String method ops on the
returned value:

>>> o.length()
Traceback (innermost last):
  File "<console>", line 1, in ?
AttributeError: 'string' object has no attribute 'length'

So you can construct a new Java String from the Python string object:

>>> from java.lang import *
>>> o2 = String( o )
>>> o2.length()
5

Does that help?

kb
Roman Milner wrote:

> >>>>> "BZ" == Brian Zhou <bri...@ya...> writes:
>
>     BZ> In dynamic languages like Python/Smalltalk/Scheme, you usually
>     BZ> don't need cast. Because "the methods a particular instance
>     BZ> can perform" becomes the type of the instance, it's not as
>     BZ> important which class that instance actually belongs. At
>     BZ> instantiation time, "all the methods the instance can perform"
>     BZ> is determined. At runtime, you either can call the method or
>     BZ> you get an AttributeError.
>
> What if you need to pass an object to a java method that requires it
> to be cast to a different java type?
>
> What if one java method gives me an Object back, but it is actually a
> String. I have another java method that requires a String argument. Is
> there a way to do that cast in jython?
>
> I ran in to this when trying to use the Java xml-rpc library from
> jython.
>
> ^Roman
>
> _______________________________________________
> Jython-dev mailing list
> Jyt...@li...
> http://lists.sourceforge.net/lists/listinfo/jython-dev

Re: [Jython-dev] casting

[Brian Z. <bri...@ya...>]

That's exactly like the example I gave.

You get the Object back (you know it's actually a String, right?), then you
can just pass the instance directly to the xml-rpc method expecting a
String. No cast needed. Because you call the method from Jython, there won't
be any compilation error. Runtime, if it's really a String, no problem; if
not, it will raise AttributeError for any String method get invoked.

-Brian

----- Original Message -----
From: "Roman Milner" <ro...@sp...>
To: "Brian Zhou" <bri...@ya...>
Cc: <jyt...@li...>
Sent: Friday, June 29, 2001 5:21 PM
Subject: Re: [Jython-dev] casting


> >>>>> "BZ" == Brian Zhou <bri...@ya...> writes:
>
>     BZ> In dynamic languages like Python/Smalltalk/Scheme, you usually
>     BZ> don't need cast. Because "the methods a particular instance
>     BZ> can perform" becomes the type of the instance, it's not as
>     BZ> important which class that instance actually belongs. At
>     BZ> instantiation time, "all the methods the instance can perform"
>     BZ> is determined. At runtime, you either can call the method or
>     BZ> you get an AttributeError.
>
> What if you need to pass an object to a java method that requires it
> to be cast to a different java type?
>
> What if one java method gives me an Object back, but it is actually a
> String. I have another java method that requires a String argument. Is
> there a way to do that cast in jython?
>
> I ran in to this when trying to use the Java xml-rpc library from
> jython.
>
> ^Roman
>

Re: [Jython-dev] casting

[Roman M. <ro...@sp...>]

>>>>> "BZ" == Brian Zhou <bri...@ya...> writes:

    BZ> In dynamic languages like Python/Smalltalk/Scheme, you usually
    BZ> don't need cast. Because "the methods a particular instance
    BZ> can perform" becomes the type of the instance, it's not as
    BZ> important which class that instance actually belongs. At
    BZ> instantiation time, "all the methods the instance can perform"
    BZ> is determined. At runtime, you either can call the method or
    BZ> you get an AttributeError.

What if you need to pass an object to a java method that requires it
to be cast to a different java type?

What if one java method gives me an Object back, but it is actually a
String. I have another java method that requires a String argument. Is
there a way to do that cast in jython?

I ran in to this when trying to use the Java xml-rpc library from
jython.

^Roman

Re: [Jython-dev] casting

[Brian Z. <bri...@ya...>]

In dynamic languages like Python/Smalltalk/Scheme, you usually don't need
cast. Because "the methods a particular instance can perform" becomes the
type of the instance, it's not as important which class that instance
actually belongs. At instantiation time, "all the methods the instance can
perform" is determined. At runtime, you either can call the method or you
get an AttributeError.

For example

>>> import java
>>> v = java.util.Vector()
>>> v.add(java.util.Properties()) #  Properties extends Hashtable
1
>>> v.add(java.util.Hashtable())
1
>>> p = v.elementAt(0)
>>> h = v.elementAt(1)

>>> h.contains("abc")
0
>>> p.contains("abc")
0
>>>
>>> h.getProperty("abc")
Traceback (innermost last):
  File "<console>", line 1, in ?
AttributeError: getProperty
>>> p.getProperty("abc")    # works ok, just returns None

In the above example, the JVM+jython knows you can can call contains() on
both p and h, but you can only call getProperty() on p.

You can see that you don't need casting when assigning the returned Object
from v.elementAt() to a variable. Java must keep some class information
available at runtime for polymorphism to work, otherwise even with cast you
wouldn't know whether you can perform a certain method call or not.

BTW, this topic probably should be in jython-users list instead..

Regards,

-Brian

----- Original Message -----
From: "Stuart Swerdloff" <st...@me...>
To: <jyt...@li...>
Sent: Friday, June 29, 2001 2:31 PM
Subject: Re: [Jython-dev] casting


> It isn't clear to me how object 'y' would be of Class myJavaClass2...
> Without any additional information, I would expect Jython to assign y to
> be
> a reference to x, of Class myJavaClass1.  Unless y had been previously
> instantiated
> or referenced to (assigned to?) an object of Class myJavaClass2.
>
> I have had some difficulty in dealing with this (upcasting) in Java, and
> resorted to
> a shallow copy based on reflection...
>
> Stuart
>
> Kevin Butler wrote:
>
> > marco wrote:
> > >
> > > hy there,
> > >
> > > how can i cast a java-instance to another type.
> > > ex.
> > > x = myJavaClass1()
> > > y = (myJavaClass2) x
> >
> > In Jython, that looks like:
> >
> > x = myJavaClass1()
> > y = x
> >
> > :-)
> >
> > No casts needed.
> >
> > kb
> >
> > _______________________________________________
> > Jython-dev mailing list
> > Jyt...@li...
> > http://lists.sourceforge.net/lists/listinfo/jython-dev
>
>
> _______________________________________________
> Jython-dev mailing list
> Jyt...@li...
> http://lists.sourceforge.net/lists/listinfo/jython-dev
>

Re: [Jython-dev] casting

[Kevin B. <kb...@ca...>]

Woops!

I completely misread your post:

Marco wrote:
> how can i cast a java-instance to another type.
> ex.
> x = myJavaClass1()
> y = (myJavaClass2) x
> 
> where:
> myJavaClass2 extends myJavaClass1

So you have:

public class A
{
}

public class B extends A
{
}

and you want to do:

{
  A a = new A();
  B b = (B) a;
}

Right?

You can't do that -- you will get a ClassCastException, because you're trying to use an object as if it were something it _knows_ it is not.

You can only cast to what the instance _really_ _is_.

The usual pattern is:

{
  A a = new B();
  B b = (B) b;
}

Which works as I described:  So I've been using an instance of B as if it were an instance of A, but after the cast, I want to use it as what it really is.

The idiom, if you _really_ have an "A" and need to get a "B" is to do do a shallow copy/a delegation pattern something like this:

public B( A other ) // "copy constructor"
{
  // either a shallow copy, as you said, or retain reference to 'other' instance to delegate to
}

public static B makeB( A other )
{
  if ( other instanceof B )
  {
 	return (B) other;
  }
  else
  {
	return new B( other );
  }
}

This is not something you want to do often, and generally indicates you have a problem in the design.

kb

Re: [Jython-dev] casting

[Stuart S. <st...@me...>]

It isn't clear to me how object 'y' would be of Class myJavaClass2...
Without any additional information, I would expect Jython to assign y to
be
a reference to x, of Class myJavaClass1.  Unless y had been previously
instantiated
or referenced to (assigned to?) an object of Class myJavaClass2.

I have had some difficulty in dealing with this (upcasting) in Java, and
resorted to
a shallow copy based on reflection...

Stuart

Kevin Butler wrote:

> marco wrote:
> >
> > hy there,
> >
> > how can i cast a java-instance to another type.
> > ex.
> > x = myJavaClass1()
> > y = (myJavaClass2) x
>
> In Jython, that looks like:
>
> x = myJavaClass1()
> y = x
>
> :-)
>
> No casts needed.
>
> kb
>
> _______________________________________________
> Jython-dev mailing list
> Jyt...@li...
> http://lists.sourceforge.net/lists/listinfo/jython-dev

Re: [Jython-dev] casting

[Kevin B. <kb...@ca...>]

marco wrote:
> 
> hy there,
> 
> how can i cast a java-instance to another type.
> ex.
> x = myJavaClass1()
> y = (myJavaClass2) x

In Jython, that looks like:

x = myJavaClass1()
y = x

:-)

No casts needed.

kb

[Jython-dev] casting

[marco <ma...@em...>]

hy there,

how can i cast a java-instance to another type.
ex.
x = myJavaClass1()
y = (myJavaClass2) x

where:
myJavaClass2 extends myJavaClass1

x
ma

--
*******************************************************
Marco Machmer

Phone : +49-2638-94 666 4
Fax   : +49-2638-94 666 7
email : mailto:ma...@em...
www   : http://www.emha.de
        http://www.Marco-Machmer.de
Um ein tadelloses Mitglied einer Schafherde
sein zu koennen, muss man vor allem ein Schaf sein.
Albert Einstein

[Jython-dev] How do compile several .py file in jythonc

[syKim <re...@ne...>]

SGVsbG8uLg0KDQpJJ20gdHJ5aW5nIHNvbWUgcHJvZ3JhbSB3aXRoIHNldmVyYWwgcHkgZmlsZSAo
bWFpbiBtb2R1bGUgJiBzdWIgbW9kdWxlcykNCg0KaXQgaXMgYXBwbGV0IHVzaW5nIHN3aW5nLi4o
aW1wb3J0IGphdmEgbGlicmFyeSkNCg0KSSBjb21waWxlZCBpdCB0aGlzIG1ldGhvZCA6DQoNCmp5
dGhvbmMgLWMgLWQgLWogTWFpbi5qYXIgTWFpbi5weQ0KDQp3aGVuIEkgbW9kaWZ5IGEgc3ViIG1v
ZHVsZSwgSSBjb21waWxlIHRoZSBtYWluIG1vZHVsZS4NCg0KYnV0LCBpdCBpcyBjb21waWxlZCBl
dmVyeSBzdWIgbW9kdWxlIGZpbGUgdG9nZXRoZXIuDQoNCkl0IHRha2VzIHNvIGxvbmcgdGltZSBp
biBteSBjb21wdXRlci4uDQoNCmNhbiBhbnkgbWV0aG9kIGNvbXBpbGUgb25seSBhIHN1YiBtb2R1
bGU/DQoNCnBsZWFzZSBzb21lIGFkdmljZSA6KQ0KDQog

[Jython-dev] multi-level overriding fix: jythonc part

[Samuele P. <pe...@in...>]

I have posted to sf also the jythonc part of a possible fix
for 

#222819 Multi-level Java method overriding fails.

It should fix also a new discovered (by me) jythonc bug "Wrong 'extends'".

regards, Samuele Pedroni.

[Jython-dev] [ jython-Patches-431224 ] possible fix for #222819

[<no...@so...>]

Patches item #431224, was opened at 2001-06-07 13:27
You can respond by visiting: 
http://sourceforge.net/tracker/?func=detail&atid=312867&aid=431224&group_id=12867

Category: None
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Samuele Pedroni (pedronis)
>Assigned to: Finn Bock (bckfnn)
Summary: possible fix for #222819

Initial Comment:
This is a patch that seems to solve

#222819 Multi-level Java method overriding fails 

while keeping

#231507 stack overflow for simple class
#406193 stack overflow when delegating

closed and

http://www.geocrawler.com/lists/3/SourceForge/7018/0/56
85848

working.

Note: only interpreter part (not jythonc).

PS: It's a hack, I know.

----------------------------------------------------------------------

>Comment By: Samuele Pedroni (pedronis)
Date: 2001-06-27 07:29

Message:
Logged In: YES 
user_id=61408

Now, jythonc part too.

Fix "Wrong 'extends'" too.

----------------------------------------------------------------------

Comment By: Finn Bock (bckfnn)
Date: 2001-06-08 09:52

Message:
Logged In: YES 
user_id=4201

Looks good. Please commit it but keep #222819 open until we 
have a fix for jythonc as well.

----------------------------------------------------------------------

You can respond by visiting: 
http://sourceforge.net/tracker/?func=detail&atid=312867&aid=431224&group_id=12867

Re: [Jython-dev] i want to call jython script in java

[Jim A. <ji...@tr...>]

Rajesh Vilasrao Bhujbal wrote:
> 
> Hi,
> In my project i want to call jython script file (.py) in java
> programme. I dont' want to compile jython script file, because i want
> to cange some things in script file. I want  avoid recompiling of java
> file.
> 
> So can you help me in this matter. Please send me code to call .py
> file through java.
> 
> Rajesh
> 

Snippet below from Bruce Eckel's 'Thinking in Patterns' with Java,
Chapter 9.

Download the book here:

  http://www.mindview.net/Books/TIPatterns/

%<-----------

 PythonInterpreter interp =  new PythonInterpreter();
 System.out.println( "Loading GreenHouse Language");
 interp.execfile("GreenHouseLanguage.py");
 System.out.println("Loading GreenHouse Script");
 interp.execfile("Schedule.ghs");
 System.out.println("Executing GreenHouse Script");
 interp.exec("run()");

%<-----------

-- 
                                 __o
Jim Adrig                      _ \<,_
ji...@tr...          ' `/ ' `
                  ___________ `-'  `-'


"There are many ways of moving forward,
 but only one way of standing still."
                                 - Franklin D. Roosevelt

Re: [Jython-dev] not able to download jython

[<bc...@wo...>]

[Rajesh Vilasrao Bhujbal]

>HI,
>i  am not able to download jython from your site. Download page contains
>one big jython "class" file due to which i am not able to download.
>Please provide jython's zip file in download page.

The jython-20.class file is already zipped. Re-zipping the file will not
make it much smaller.

regards,
finn

Re: [Jython-dev] Webpage Applet Demos

[<bc...@wo...>]

On Mon, 25 Jun 2001 12:29:06 -0500, you wrote:

> From jython.sourceforge.net/applets
>
>Can't seem to get any of them to work...
>and I seem to recall they've worked before.

It seems like the appletdemo.[jar,cab] was compiled while running java2.
That results in a reference to java.awt.event.HierarchyEvent which fails
in both browsers. I have recompile the demos and uploaded the new
appletdemo.* files and now it works for me (W2K SP2 & IE 5.0).

regards,
finn

[Jython-dev] Webpage Applet Demos

[Mats W. <ma...@la...>]

 From jython.sourceforge.net/applets

Can't seem to get any of them to work...
and I seem to recall they've worked before.
Running W2K (SP2).  From IE 5.0, here's
the Java console output:

java.lang.InstantiationException: HelloWorld
	at com/ms/applet/BrowserAppletFrame.newInstance
	at com/ms/applet/AppletPanel.processSentEvent
	at com/ms/applet/AppletPanel.processSentEvent
	at com/ms/applet/AppletPanel.run
	at java/lang/Thread.run

(this is, of course, for the HelloWorld applet).
The Jython Loading display does come up.


 From Netscape 4.77 (what are these signature file
and manifest file errors?)

# Error: Invalid digital signature file (-7885)
#	jar file: C:\DOCUME~1\mats\LOCALS~1\Temp\jzipSB4A.TMP
#	path:     C:\DOCUME~1\mats\LOCALS~1\Temp\jzipSB4A.TMP
# Error: internal error - error in manifest file (-1)
#	jar file: C:\DOCUME~1\mats\LOCALS~1\Temp\jzipSB4A.TMP
#	path:     /applets/
# Applet exception: exception: java.lang.NullPointerException

java.lang.NullPointerException
   at org.python.core.PySystemState.add_package(Compiled Code)
   at org.python.core.PySystemState.add_package(Compiled Code)
   at org.python.core.Py.initProperties(Compiled Code)
   at org.python.core.Py.initProxy(Compiled Code)
* at HelloWorld.__initProxy__(Compiled Code)
   at HelloWorld.<init>(Compiled Code)
   at netscape.applet.DerivedAppletFrame$LoadAppletEvent.dispatch(Compiled 
Code)
   at java.awt.EventDispatchThread$EventPump.dispatchEvents(Compiled Code)
   at java.awt.EventDispatchThread.run(Compiled Code)
   at 
netscape.applet.DerivedAppletFrame$AppletEventDispatchThread.run(Compiled Code)

[Jython-dev] Questions for jython.

[Nabil O. <Nab...@cl...>]

Hi there, 

I want to use a java python translator but before I would like 
get informed about few issues. I would be grateful to have answer
for these questions ASAP.


        - Does jyhton provide a source only bridge or can it 
          use .jar and .class hierarchies ?

         - What is the runtime overhead associated passing
           between Java and Python ? Is some overhead present
           even when data is not passed between both of them ?

         - Can we have native methods written in Python for 
           Java classes ?

         - How is Java called from Python ?

         - What's the memory footprint for the system ?

         - Can python debugger access Java instances and
           classes?

         - Are there any limits restricting the use of 
           jython ? How about platforms ?

         - What are the limits imposed by the semantic
           differences of Java and Python ?


Thank you.

- Nabil:)
                   _\\|//_
                    (o-o)
     +-*%+-*%+-*%ooO (_) Ooo+-*%+-*%+-*%
	  Nabil Obeid.
	  Cloakware Corp.
	  311-260 Hearst Way,
	  Kanata, Ontario, Canada,
	  Earth, Solar System, K2L 3H1.
                 .oooO
                 (   )   Oooo.
     +-*%+-*%+-*% \ (+-*%(   )+-*%+-*%+-*%
                   \_)    ) /
                         (_/

[Jython-dev] Bug? Jython misinterprets the letter =?iso-8859-1?q?=D6?= (\xd6 in Phyton) within strings as EOF

[<Chr...@t-...>]

Hi,
Jython 2.1a1 misinterprets the letter Ö (big O-Umlaut or Unicode u00d6 or \xd6 
in Phyton) within strings as EOF.
My JDK ist 1.3.1.
OS is Windows 98SE
Local ist German, with characterset  850

This is what Python 2.1 shows if Ö is the only letter in the string
>>> "Ö"
'\xd6'

This ist what Jython 2.1a1 shows for the same case
>>> "Ö"
Traceback (innermost last):
  (no code object) at line 0
  File "<console>", line 2
SyntaxError: Lexical error at line 2, column 0.  Encountered: <EOF> after : ""
>>>

Here now Python when Ö ist inclosed in a word
>>> "KÖLN"
'K\xd6LN'

This is Jython 2.1a1 for the same case:
>>> "KÖLN"
Traceback (innermost last):
  (no code object) at line 0
  File "<console>", line 1
        "KÖLN"
           ^
SyntaxError: invalid syntax

I would appreciate a hint on how to bypass or correct this problem if it is not 
a bug. I can not use Jython for my purposes, while my Java is still to limited 
to fix the problem on my own and send you a patch.

Regards
Christoph Becker

Re: [Jython-dev] Is Jython 2.0 compatible with JDK 1.4 (Merlin)?

[Timo-Pekka <py...@ik...>]

Ype Kingma wrote:

> >I'm researching Forte For Java 3.0 compatibility with JDK 1.4 (currently beta).
> >I was wondering if you have any information about the compatibility of Jython
> >2.0 with JDK 1.4?  Any known issue or limitation?
> 
> A few weeks ago there was a post from someone who had Jython
> running under java 1.4 _faster_ than CPython on the same machine.
> You could have a look in the list archives.

I did the test and to be fair, pystone which I used, ran under _server_ vm
and it took around 15 minutes to complete. With much fewer iterations CPython
won with about 1000 pystones per second. Hotspot server vm is very good, but
it does not suit for short running applications and need time to optimize the
code to the fullest.

JDK 1.4 isn't very stable with Jython :-( Although pystone ran perfectly, I
couldn't get pybench running more than about half of the tests without
segfaults and if I remember correctly, there were even more problems with the
client vm than server vm.

Pybench is quite interesting benchmark, because it shows clearly that Jython
and CPython has their own strengths and weaknesses, and differences can be
quite big for both directions. For example Jython's exception handling seems
to be something like 30 (IIRC) times faster than CPython's, but on other
areas CPython wins with large margins.

But, there is some work to be done before the JDK 1.4/Jython 2.x combination
is really usable, but it looks very promising.

-- 
Pege