On Wed, Nov 27, 2002 at 01:26:02PM -0800, Anthony Green wrote:
Did Sun just release swing, etc under a BSD license??
The FSF legal team is looking into the matter. I'll update everyone
when there is information to report.
___
Classpath mailing
On Tue, Jan 29, 2002 at 08:30:26PM -0700, Tom Tromey wrote:
It turns out that the Gtk+ peers aren't really complete.
They are mostly complete. I never said that they were complete.
For instance, it isn't possible to make an AWT checkbox due to a
missing native method.
This is odd. I
Mark Wielaard [EMAIL PROTECTED] writes:
I was not aware that the license of the awt classes should have been
changed from LGPL to GPL.
The AWT classes are GPL'd. New licensing headers need to be checked
into CVS. If someone could go ahead and do this, that would be a
help. I haven't had
Mark Wielaard [EMAIL PROTECTED] writes:
I am a bit annoyed by this. Why was this never clearly announced!
Things have been murky for some time on this issue. Only recently did
we get clarification on the status of the AWT.
Are you sure the libgcj people know about this?!?!
Yes. They are
Jochen Hoenicke wrote:
As far as I can see, there is no real java.util.zip implementation in
classpath yet, so do you agree if I just commit the content of jazzlib
to the classpath repository, overwriting the three small existing
classes?
That's fine. As you commit files, could you please
Mark Wielaard [EMAIL PROTECTED] writes:
Does anybody know who handles the copyright assignment at the FSF?
I handle assignment paperwork for Classpath in conjunction with
[EMAIL PROTECTED] at the FSF.
John Leuner's paperwork has already been received and processed. I'm
waiting on Jochen's
Etienne M. Gagnon [EMAIL PROTECTED] writes:
But, as I repeatedly said, CNI is ONLY good for gcj.
Kaffe supports CNI.
___
Classpath mailing list
[EMAIL PROTECTED]
http://mail.gnu.org/mailman/listinfo/classpath
John Keiser [EMAIL PROTECTED] writes:
Etienne has brought up some concerns about GNU copyright assignment
being required before any contributions are made to the project, and
I have some concerns as well (albeit different ones).
Copyright assignments are not always required.
Changes that
John Leuner [EMAIL PROTECTED] writes:
I emailed [EMAIL PROTECTED] (IIRC), but never got a response.
Sorry for the lag on this. I'll followup on this matter today.
___
Classpath mailing list
[EMAIL PROTECTED]
From the current copyright.list at GNU:
KAFFE Transvirtual Technologies, Inc. 2000-10-20
Assigns Kaffe RMI Library and future changes submitted to the FSF.
(Tim Wilkinson)
Everything is finally in place to merge Kaffe's RMI library into
Classpath/libgcj.
Feel free to run with
Brian Jones [EMAIL PROTECTED] writes:
Short answer, yes we'd be interested. Long answer is legal issues may
keep it from being useful.
Last I checked (and this was a while ago), Sun had patents which would
prevent the implementation/distribution/use of a free implementation
of JINI.
Fang Wei Jian [EMAIL PROTECTED] writes:
Until now, have classpath already been ported to kaffe, or this
project is in progress?
Classpath has not been ported to Kaffe. This is, however, planned
work. If you're interested in starting work on the port, that would
be great.
to the work that I'll be doing here at TVT.
If any of you live in the bay area, please drop me an email.
--
Paul Fisher * [EMAIL PROTECTED], [EMAIL PROTECTED]
"John Keiser" [EMAIL PROTECTED] writes:
Are we allowed to make the Classpath pages on GNU a bit snazzier
No. (It would be nice if we could have some nice snazzy-looking web
pages, but that's not allowed at www.gnu.org.)
See
Brian Jones [EMAIL PROTECTED] writes:
Anyone against removing that archive and fixing the mailing alias not to
create it?
Please do not remove the archive.
It's standard policy to keep archives for GNU mailing lists in GNU
filespace.
I'm here at USENIX and don't have the ability to look at Jemini.
Could someone please take a look at it and get back to me?
Relevant information regarding Jemini is at:
URL:http://www.upside.com/texis/mvm/story?id=395135c70
and
URL:http://www.websprocket.com/jemini.html
The Upside article says
Brian Jones [EMAIL PROTECTED] writes:
I've updated the web page. I need Paul to disable or fix any cron
job he has updating those since I think he's probably pulling from
the old cvs archive to do it automatically.
Fixed.
Tom Tromey [EMAIL PROTECTED] writes:
2. Is there some document describing the indentation style. I find
it oddly inconsistent. For instance, synchronized blocks seem to be
indented differently from everything else. Is this the universal
Classpath style?
Being a GNU project, we follow the
Tom Tromey [EMAIL PROTECTED] writes:
My understanding is that Paul is looking into this. Paul, am I
right? His idea, which isn't listed, is to use wrapper classes to
make it possible to write CNI code which is compiled by the C++
compiler into JNI calls.
That's correct. The wrapper
Brian Jones [EMAIL PROTECTED] writes:
Oh bugger, compared to the amount of Java out there, we might as
well just dup the work and have both versions...
This isn't necessary. The CNI-JNI C++ wrapper classes will work just
fine.
Brian Jones [EMAIL PROTECTED] writes:
How are awt tests to be handled in mauve?
Any event within the AWT can be simulated. The AWT event system is
extremely convoluted and broken, but it does allow for automated
testing programs to be easily written. I'm guessing this is what you
were
"Aaron M. Renn" [EMAIL PROTECTED] writes:
For SSH users, you'll need to have SSH 1.something.
The latest version of OpenSSH should work just fine.
URL:http://www.openssh.com/
SSH will prompt you for a password every time you run it.
You can use a key that has a null passphrase, or use
"John Keiser" [EMAIL PROTECTED] writes:
Could somebody *please* speak up and say what the heck is the deal
with the AWT?
Transvirtual depends on the ability to relicense their AWT under
non-GPL'd terms to companies that are unable to use GPL'd code. If
the FSF were to relicense the AWT under
Patrick Spingys [EMAIL PROTECTED] writes:
But I think there exists no unofficial newer Classpath, too.
The most current code for Classpath can be obtained from CVS.
CVSROOT is :pserver:[EMAIL PROTECTED]:/gd/gnu/anoncvsroot;
Module is classpath. There is no password.
Nic Ferrier [EMAIL PROTECTED] writes:
I am thinking mainly of javax.servlet which I know Paul Siegmann has
done.
Last I heard, Paul was not interested in having javax.servlet be
distributed by GNU.
I am working on a plan to bring JavaMail to the masses and that
would need a similar
Per Bothner [EMAIL PROTECTED] writes:
The first question is why? What is the goal?
The goal is for Classpath to be the "standard" set of Java class
libraries for the majority of free JVMs. Or at least, with a
reasonable amount of effort, a VM should be able to support Classpath.
Therefore,
I've been working on coming up with an "acceptable" solution for
dealing with our current merger problems regarding the use of CNI and
JNI. While it's important that Classpath continue to support JNI, CNI
does have significant benefits, and I'd like to start porting
Classpath to CNI, assuming we
software movement.
Paul Fisher
Anthony Green
Anthony Green [EMAIL PROTECTED] writes:
If we could all agree to switch to this license, then we could start
working on a plan to merge the two projects.
If anyone wants to weigh in on the issue of switching to libgcc or
libstdc++ terms for Classpath, now is the time to do so.
I'd very much
ode needs to be completed. I personally find CNI to be
extremely elegant (about 10 lines of JNI code correspond to one line
of CNI).
--
Paul Fisher
Michael Emmel [EMAIL PROTECTED] writes:
I think there are some serious licensing issues
As long as you aren't mixing Classpath source with Sun source -- that
is, you're only using certain Classpath classes to replace Sun
classes, you should be OK. However, I'm not a lawyer, nor am I
Stuart Ballard [EMAIL PROTECTED] writes:
I was under the impression that the primary difference between the
GPL and the LGPL was that the GPL required software linked with it
to be under the GPL, but the LGPL did not (the oft-mentioned "viral"
property of the GPL).
That's correct. However,
Dick Hollenbeck [EMAIL PROTECTED] writes:
"Java" must be used as an adjective when refering to your
work. "Java Compatible" cannot be claimed unless you pass
the JCK (and apparently have licensed the JCK to have
passed it.)
Use of the name "Java" can be controlled because it is a
Sun
Guan Yang [EMAIL PROTECTED] writes:
1) Is it legally acceptable to implement GNU Classpath on the basis
of javadoc documentation provided by Sun? The reason I ask this is
because this documentation is *generated* from the actual
classes.zip source code.
We do use javadoc documentation,
Bryce McKinlay [EMAIL PROTECTED] writes:
That said, I also agree that libgcj needs to get closer to
classpath, and perhaps it would be nice if classpath could support
gcj 'out of the box'.
This is a priority.
on the details of how/why that happens, I'd love to know.
Here's a picture of the pre-6.1 Shadow Man in an applet:
URL:http://people.redhat.com/pnfisher/classpath-gdk-pixbuf.png
--
Paul Fisher
Paul Fisher [EMAIL PROTECTED] writes:
gdk-pixbuf has been successfully wrapped, and it now functions as an
ImageProducer. We can load any image format that gdk-pixbuf
supports -- bmp, gif, jpeg, png, pnm, ras, tiff, and xpm.
Oh, and since gdk-pixbuf only supports single frame images, we
tree later today that explains how to get the
peers up and running.
--
Paul Fisher * [EMAIL PROTECTED]
The GTK+ peers are now building with the rest of Classpath, in the
"proper" build environment. However, there are a few pieces of
Classpath which cannot be compiled.
Could those responsible for the following files please fix these
errors?
I'd like to change the build process, such that if any
Brian Jones [EMAIL PROTECTED] writes:
- GTK 1.2
Errr.. "GTK+ 1.2.x"
* java.math needs some work.
java.math has _one_ missing function.
--
Paul Fisher
.
--
Paul Fisher
messages over there (yes, he
has an archive of _everything_).
(Brian was supposed to have moved all his archives from his old work
machine at Nortel (which was http accessible) to GNU months ago, but
for some reason, he's been terribly slow about it. :)
--
Paul Fisher * [EMAIL PROTECTED]
Unfortunately, the end portion of this week hasn't been so kind to me,
and I haven't been able to finish up everything that's necessary for a
release.
New target date is sometime this week...
--
Paul Fisher * [EMAIL PROTECTED]
redhat.com/pnfisher/metalworks.png
--
Paul Fisher * [EMAIL PROTECTED]
URL:http://people.redhat.com/pnfisher/simple-metal-correct.png
URL:http://people.redhat.com/pnfisher/simple-motif-correct.png
Yes, these are screenshots of Swing 1.1.1 being run under the GTK+
AWT.
Brian Jones [EMAIL PROTECTED] writes:
java.awt.image Paul Fisher [EMAIL PROTECTED]
I think some folks had volunteered to do the image decoders
and I'm wondering who and what the status is. :)
There were a couple of people that volunteered, but I haven't heard
from any
URL:http://people.redhat.com/pnfisher/fonts.png
beautiful.
--
Paul Fisher * [EMAIL PROTECTED]
.
--
Paul Fisher * [EMAIL PROTECTED]
Thomas Down [EMAIL PROTECTED] writes:
I haven't had a serious look at libart, but I'd rather doubt that,
for normal applications, it would be sensible to implement java2d on
top of any other high-level graphics API.
libart is meant to mirror the java2D API. So, in the end (that is,
when
Does somebody have time to look into this issue?
It does seem as though the code for __ieee754_rem_pio2 is not in the
tree, but it's been ages since I've worked on the math stuff. My
guess is that we might be linking against the libm cos/sin/tan
functions and that's why we haven't had problems
John Keiser [EMAIL PROTECTED] writes:
But while we don't need to decide now, I think we *do* need to
decide before release.
Release? What's that? :)
--
Paul Fisher * [EMAIL PROTECTED]
Aaron M. Renn [EMAIL PROTECTED] writes:
I just wanted to let you all know that in the case of the menu
classes, I modified our peer interface so that it is different from
the JDK's.
Why did you decide to make this change?
The GTK+ AWT peers currently function as drop-in replacement for
round info on JavaSpaces) to [EMAIL PROTECTED] Please CC me on the
message.
--
Paul Fisher * [EMAIL PROTECTED]
tr objects. It slightly
hampers code readability, but this is the only way we can match the
speed of Sun's Character implementation.
--
Paul Fisher * [EMAIL PROTECTED]
Exams are over. The semester is over. Time to code.
--
Paul Fisher * [EMAIL PROTECTED]
Four different stories covering the topic:
URL:http://www.zdnet.co.uk/news/1999/16/ns-7955.html
URL:http://www.news.com/News/Item/0,4,35889,00.html?st.ne.fd.gif.d
URL:http://www.infoworld.com/cgi-bin/displayStory.pl?990429.icjava.htm
operating system's copy of rsh. If you're using krb4, you'll need to
set your CVS_RSH environment variable to krsh.
Anonymous pserver access on anoncvs.gnu.org will not be turned off.
--
Paul Fisher * [EMAIL PROTECTED]
"Jochen Hoenicke" [EMAIL PROTECTED] writes:
1) For JDK-1.1.7 all Methods in BitSet except size() are synchronized.
2) For JDK-1.2 no Method is synchronized.
Is it possible that they're synchronizing on something other than the
this instance, possibly a field?
--
Paul Fishe
with our 0.0 release?
--
Paul Fisher * [EMAIL PROTECTED]
Christoph Toshok [EMAIL PROTECTED] writes:
BTW: I had two compile problems with the latest Japhar. First, it
is looking for a file called nativeglue.h, which doesn't exist and
doesn't get generated for me. I just touched it and all was well.
Also, japhard is bombing bigger than life on
the applet's source code and release that as
well. Like all good GNU software, it does everything that the JDK's
Animator applet does, plus much much more.
(OK, OK, I know this isn't really contributing anything to the
discussion, but I'm all for meme propagation...)
meme?
--
Paul Fisher * [EMAIL
.tar.gz.
If you have a patch, you can send it to the list.
--
Paul Fisher * [EMAIL PROTECTED]
ANTAFE.EDU
classpath.org NS NS1.GNU.ORG
root@io:/home/rao/gnus/texi# host -t a www.classpath.org
www.classpath.org A 209.81.8.252
--
Paul Fisher * [EMAIL PROTECTED]
for JPEG and GIF. For those of you that might be
interested in writing a decoder, our XBM decoder is a good example of
how to structure one.
--
Paul Fisher * [EMAIL PROTECTED]
) or rely on libart_lgpl in GNOME.
--
Paul Fisher * [EMAIL PROTECTED]
hat the MX record will be going
away.
--
Paul Fisher * [EMAIL PROTECTED]
y portable implementation, how do you ever
display anything to the screen? You have to interface with the
hardware at some level, and that's going to be non-portable.
--
Paul Fisher * [EMAIL PROTECTED]
``Sun demurs on making Java open source''
URL:http://www.zdnet.com/pcweek/stories/news/0,4153,1013904,00.html
*ahem*. Back to coding...
--
Paul Fisher * [EMAIL PROTECTED]
"John Keiser" [EMAIL PROTECTED] writes:
Perhaps it's time to hang the docs up on the web site? I'd be happy
to generate them and tar them up if someone can place them up there.
I'd prefer not to put up API docs on the web until there's a free
replacement for JavaDoc.
--
P
with Emacs' cc-mode reindent, should be able to do
an automagic conversion of the source code to GNU style.
I do plan to move the source code to GNU formatting conventions;
however, naming conventions will keep with Sun's style.
--
Paul Fisher * [EMAIL PROTECTED]
were added in JDK 1.1. To retain
compatibility with Java 1.0 programs, default implementations of
byteValue() and shortValue() were added.
Please change your commenting style to conform to GNU standards. The
double stars on each line are terribly ugly.
--
Paul Fisher * [EMAIL PROTECTED]
s declared in the throws
clause are ignored.
--
Paul Fisher * [EMAIL PROTECTED]
know ahead of time if its possible to throw a
NullPointerException because of an argument. If it is, we declare
that we throw it, if not, then we don't.
--
Paul Fisher * [EMAIL PROTECTED]
thors make a contradictory statement in that
they say it's distributed under the GPL, but if you use it
commercially, you're required to pay them.
I'll contact them about the discrepancy.
--
Paul Fisher * [EMAIL PROTECTED]
John Keiser [EMAIL PROTECTED] writes:
I've noticed that a lot of the copyright comments are written thusly:
Per GNU coding standards, multi-line comments should look like this:
/* this is a multi-line
comment */
So the same should apply to our file headers.
--
Paul Fisher * [EMAIL
not currently. I am
interested in setting up Bonsai for the GNU CVS packages that allow
anoncvs access, so I'll put that on the TODO list.
--
Paul Fisher * [EMAIL PROTECTED]
be an option for you. We also need something along the lines of the
GNU C Library Reference Manual for Classpath. I've always planned to
do work on such a document, but only after we get v1.0 out the door.
--
Paul Fisher * [EMAIL PROTECTED]
ell. Make sure you have Java
turned on, or else he won't tap his pitchfork.
Assuming everyone is happy with the new pages, I'll announce them on
the ``What's New'' portion of www.gnu.org and link to them from the
main GNU software page.
--
Paul Fisher * [EMAIL PROTECTED]
.
System.identityHashCode() should be fast.
--
Paul Fisher * [EMAIL PROTECTED]
(unless you want Motif support).
If you really want to do native/peer 1.2 AWT stuff, your time might be
better spent hacking on Classpath's AWT. It looks as though our 1.2
AWT might end up relying on GnomeCanvas, since it already does most of
the fancy features required for Java 2D support.
--
Paul
"Aaron M. Renn" [EMAIL PROTECTED] writes:
I think at this point we are all targeting Java 1.2. Certainly 1.1
compliance is a must, but I am personally doing as much 1.2 as I
can.
Our AWT is targeting Java 1.1. 1.2 support will come after 1.1 is
working.
--
Paul Fisher * [EMAIL PROTECTED]
://www.cs.berkeley.edu/~mdw/proj/ninja/ninjarmi.html for use
with GNU Classpath. I haven't heard much from him lately.
On the downside, the modified NinjaRMI will not be part of GNU
Classpath -- although we'll be distributing the latest version of his
work along with the classpath tar balls.
--
Paul Fisher * [EMAIL
, we must have the same values.
(I spent quite a bit of time scratching my head over really weird
results with ScrollPane's scrollbar visibility settings last night.)
--
Paul Fisher
of the javax packages
should be sufficient.
Of course, if any of the authors would like for their packages to fall
under the FSF umbrella, they'd be able to create a setup much like
what Classpath has (cvs/web/mailing lists).
--
Paul Fisher * [EMAIL PROTECTED]
at the moment
-- hopefully I can get back to the peers this weekend.
--
Paul Fisher * [EMAIL PROTECTED]
java CharacterTest
Reading unicode database...done
DONE: 728/3578944 failures/tests
real0m41.337s
user0m0.260s
sys 0m0.190s
--
Paul Fisher * [EMAIL PROTECTED]
, we'll start at
0.9.00?)
No.
Or major.minor.feature.bugfix = 0.0.0.0
Something along this line is fine. Most GNU programs use a versioning
system of `foo-M.N' or `foo-M.N.O'.
--
Paul Fisher * [EMAIL PROTECTED]
, you were going to cache these files right?
Yes. I supposed you'd like for me to do that now? :) Until I get the
changes committed, just disable compilation/use of Character.
--
Paul Fisher * [EMAIL PROTECTED]
, you can pretend that our code isn't in the main CVS
branch.
Once we come out of our pre-alpha stage, we'll fix things up to work
with the rest of the distribution.
--
Paul Fisher * [EMAIL PROTECTED]
to JNI 1.2 code (ie. start using 1.2 VMs for development).
--
Paul Fisher * [EMAIL PROTECTED]
to work well in a GNU
system. Such a setup means working on top of X11 -- not under it.
--
Paul Fisher * [EMAIL PROTECTED]
be considered
normal - look at the errors reported and try to find any which should
not be reported.
I'll get back to you shortly.
--
Paul Fisher * [EMAIL PROTECTED]
classloading works, but could the Unicode
data be compartmentalized so that the 21k is loaded iff any Unicode
attributes are accessed?
Yes. All the loading code would be put into a static class block
which would be executed the first time that Character is referenced.
--
Paul Fisher * [EMAIL
"John Keiser" [EMAIL PROTECTED] writes:
I was unaware that the creation of Java clones was forbidden in the
past
Take everything you hear in IT mags with a very large grain of salt.
It has always been permissible to create Java clones.
--
Paul Fisher * [EMAIL PROTECTED]
"Aaron M. Renn" [EMAIL PROTECTED] writes:
See http://java.sun.com/features/1998/12/jdk.ship.html
For those that just want the documentation, it's at:
ftp://ftp.java.sun.com/pub/jdk/1.2/13466fcs/jdk12-doc.tar.gz
--
Paul Fisher * [EMAIL PROTECTED]
any AWT 1.1 call is made, the applications wants the
1.1 event model.
However, the 1.1 event model should be implemented first -- as it's
the more important of the two.
--
Paul Fisher * [EMAIL PROTECTED]
exist -- both natively, and using
cygwin32.
--
Paul Fisher * [EMAIL PROTECTED]
"John Keiser" [EMAIL PROTECTED] writes:
No, protected classes *should* be in there.
Non-public classes should not be shown. The JLS and corresponding
documentation comments (javadoc) specs only cover public classes.
We have no need to see the APIs of Sun's internal classes.
and Sun's.
--
Paul Fisher * [EMAIL PROTECTED]
the Kaffe people believe that using javap to grab signature
headers would break their clean-room environment.
--
Paul Fisher * [EMAIL PROTECTED]
1 - 100 of 151 matches
Mail list logo
