From Artem.Ananiev at Sun.COM  Wed Aug  1 08:51:46 2007
From: Artem.Ananiev at Sun.COM (Artem Ananiev)
Date: Wed, 01 Aug 2007 12:51:46 +0400
Subject: <AWT Dev> Window Decorations do not work on x86_64
In-Reply-To: <46AF3F51.2050908@redhat.com>
References: <46AA381D.3050703@redhat.com> <46AAFDAC.9090405@sun.com>
	<46ADEA37.4060201@redhat.com> <46AE09F2.9070605@redhat.com>
	<46AE497E.9030702@redhat.com> <46AF3F51.2050908@redhat.com>
Message-ID: <46B049A2.1040706@sun.com>

Hi, Francis, all,

the diffs in XPanelPeer are a part of the fix for 6567564 (Canvas's 
graphics config does not change when moved across monitors, Linux). The 
fix also includes some changes in XPanelPeer sub/super classes, and I 
think it's not a right thing to revert the changes from XPanelPeer only.

At the same time, I still can't reproduce the described problem with 
window decorations on my desktop. I have already written to Kyle 
privately about the required system configuration (32/64 bit, window 
manager, window manager version, etc.) but haven't received any reply. 
Until this information is available, we can't proceed with the bug.

Thanks,

Artem

Francis Kung wrote:
> Hi everyone,
> 
>>>>> Could you tell what window manager you're using when you observe 
>>>>> this behavior? Did you try running under Metacity? Is the bug still 
>>>>> reproducible then?
>>>>
>>>> This was tested under Metacity and it is still a bug.
>>> Sun accepted this bug this morning as bug 6586752.
>>>
>>> http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6586752
>>
>>
>> This problem was narrowed down to a regression between b14 and b15. 
>> The patch causing the problem has been uploaded here: 
>> http://icedtea.classpath.org/bugzilla/show_bug.cgi?id=37
>>
>> Though, we are still trying to determine what part of this patch is 
>> the cause.
> 
> I've found the specific area that causes this regression; it was the 
> change in XPanelPeer.java.  I've attached the relevant b14->b15 diff; if 
> you revert that patch then window decorations appear once again.
> 
> Regards,
> Francis
> 


From alexanderschunk at t-online.de  Wed Aug  1 18:37:00 2007
From: alexanderschunk at t-online.de (Alexander Schunk)
Date: 01 Aug 2007 18:37 GMT
Subject: LinAlg API
Message-ID: <1IGJ4K-1MOuhs0@fwd26.aul.t-online.de>

Hi,
i just wanted to announce that i have written a tiny API for linear Algebra that offers basic Vector, Matrix and Complex Number arithmetics.
I have already talked with some people at SUN about the integration and am currently waiting for further information on this process.
I would like to ask for any kind of community support for this API and am also glad for any kind of feedback concerning features and "size" of this API.
I am aware that some developers may need special implementation of Vector or Matrix classes and therefore would like to offer some basic implementations for this.
I am also working on some other math stuff - Fractional Numbers, Statistics - and think this could also be helpful for Java programms though there are specialised APIs already out there.
Best Regards
Alexander
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070801/2e0ae172/attachment.html>

From dan at fabulich.com  Wed Aug  1 18:58:23 2007
From: dan at fabulich.com (Dan Fabulich)
Date: Wed, 1 Aug 2007 11:58:23 -0700 (Pacific Daylight Time)
Subject: encumbrances update
In-Reply-To: <46AFAC07.6060608@sun.com>
References: <46AFAC07.6060608@sun.com>
Message-ID: <Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>

Phil Race wrote:

> It is believed to build and work on all platform combinations : windows,
> linux, solaris, 32 and 64 bit, but testing has focused on the 32 bit
> versions.

"believed to build"?  Has anyone yet tried doing the standard OpenJDK 
build on Windows, following the directions described in the documentation?

There's still no reason to believe these build scripts work until somebody 
(anybody) says: "Yes, I did the build, just like any other non-Sun 
employee would.  The build works for me."

-Dan


From Anthony.Petrov at Sun.COM  Thu Aug  2 07:09:48 2007
From: Anthony.Petrov at Sun.COM (Anthony Petrov)
Date: Thu, 02 Aug 2007 11:09:48 +0400
Subject: encumbrances update
In-Reply-To: <Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>
References: <46AFAC07.6060608@sun.com>
	<Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>
Message-ID: <46B1833C.7050705@sun.com>

Hmm...

On 08/01/2007 10:58 PM Dan Fabulich wrote:
>> It is believed to build and work on all platform combinations : windows,
>> linux, solaris, 32 and 64 bit, but testing has focused on the 32 bit
>> versions.
> "believed to build"?  Has anyone yet tried doing the standard OpenJDK 
> build on Windows, following the directions described in the documentation?
> There's still no reason to believe these build scripts work until 
> somebody (anybody) says: "Yes, I did the build, just like any other 
> non-Sun employee would.  The build works for me."
I'm afraid that until someone tells us exactly: "I've got the following 
problems with builds on Windows: this, this, and this", we can't help 
you at all, can we?..

--
best regards,
Anthony


From dan at fabulich.com  Thu Aug  2 07:19:04 2007
From: dan at fabulich.com (Dan Fabulich)
Date: Thu, 2 Aug 2007 00:19:04 -0700 (Pacific Daylight Time)
Subject: encumbrances update
In-Reply-To: <46B1833C.7050705@sun.com>
References: <46AFAC07.6060608@sun.com>
	<Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>
	<46B1833C.7050705@sun.com>
Message-ID: <Pine.WNT.4.63.0708020012310.2576@tbf>

Anthony Petrov wrote:

> I'm afraid that until someone tells us exactly: "I've got the following 
> problems with builds on Windows: this, this, and this", we can't help you at 
> all, can we?..

Of course you can! :-)  What I'm suggesting is that someone at Sun *test* 
the Windows build on their end, using the same OpenJDK source, plugs, and 
documentation that we're following, to make sure that the build can work 
at least once.

Specifically, we know for sure that all of the builds we've gotten so far 
can't be built on Windows using the provided binary plugs (due to at least 
one missing file); the hope is that b17 will make it possible to build on 
Windows.  What I'm asking is that someone at Sun try it out once, just to 
make sure.

I'm looking for someone to say "this build works for me; it is possible to 
build this code without patching it."  Think of it as doing QA on the 
build process.  If it doesn't work for Sun engineers, it certainly won't 
work for us.

-Dan


From Igor.Nekrestyanov at Sun.COM  Thu Aug  2 09:15:39 2007
From: Igor.Nekrestyanov at Sun.COM (Igor Nekrestyanov)
Date: Thu, 02 Aug 2007 13:15:39 +0400
Subject: encumbrances update
In-Reply-To: <Pine.WNT.4.63.0708020012310.2576@tbf>
References: <46AFAC07.6060608@sun.com>
	<Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>
	<46B1833C.7050705@sun.com> <Pine.WNT.4.63.0708020012310.2576@tbf>
Message-ID: <46B1A0BB.4030405@sun.com>

Hi Dan,

Of course i've tested these changes on all platforms including both 32 
and 64 bit Windows.
It is "believed to build and work" because we had not performed full 
testing of openjdk binaries
and we know that build can be fragile due to different build environments.

My tests are not "ideal" for number of reasons:
  1) It is not exactly same bits as b17 code drop (because someone else 
could integrate after me, etc.)
  2) I was using binary plugs created from my personal workspace.
      I believe they should be the same as those to be published with 
b17 code drop but this is my assumption.
      We work on removing them and may be they changed
  3) after all it was me who updated docs regarding freetype build. 
perhaps i imply some knowledge :)

Change is not small and harmless. So issues are possible despite 
multiple code reviews.
In fact one of severe build issues was found soon after my putback (fix 
for it went into b17).

As of now at least several other people tried to build openjdk with 
freetype-related changes on some platforms and it worked
for them. So hopefully this means this will work for you too :)
If not then i hope we can identify problem and fix it together.

BTW, for windows build you will need to build freetype.dll (i tried only 
dll built with visualc).
I believe that freetype build system does not generate .dll on windows 
and therefore
some manual tweaking of freetype makefiles (and sources) might be 
necessary.
For cross platform build of freetype dll for windows 64 some additional 
changes were required.

-igor

Dan Fabulich wrote:
> Anthony Petrov wrote:
>
>> I'm afraid that until someone tells us exactly: "I've got the 
>> following problems with builds on Windows: this, this, and this", we 
>> can't help you at all, can we?..
>
> Of course you can! :-)  What I'm suggesting is that someone at Sun 
> *test* the Windows build on their end, using the same OpenJDK source, 
> plugs, and documentation that we're following, to make sure that the 
> build can work at least once.
>
> Specifically, we know for sure that all of the builds we've gotten so 
> far can't be built on Windows using the provided binary plugs (due to 
> at least one missing file); the hope is that b17 will make it possible 
> to build on Windows.  What I'm asking is that someone at Sun try it 
> out once, just to make sure.
>
> I'm looking for someone to say "this build works for me; it is 
> possible to build this code without patching it."  Think of it as 
> doing QA on the build process.  If it doesn't work for Sun engineers, 
> it certainly won't work for us.
>
> -Dan
> </div>



From fkung at redhat.com  Thu Aug  2 18:02:33 2007
From: fkung at redhat.com (Francis Kung)
Date: Thu, 02 Aug 2007 14:02:33 -0400
Subject: <AWT Dev> Window Decorations do not work on x86_64
In-Reply-To: <46B0905C.9040805@redhat.com>
References: <46AA381D.3050703@redhat.com> <46AAFDAC.9090405@sun.com>
	<46ADEA37.4060201@redhat.com> <46AE09F2.9070605@redhat.com>
	<46AE497E.9030702@redhat.com> <46AF3F51.2050908@redhat.com>
	<46B049A2.1040706@sun.com> <46B0905C.9040805@redhat.com>
Message-ID: <46B21C39.4030602@redhat.com>

Hi,

It looks like I was a bit premature in blaming a patch for this bug; it 
turns out to be a much more subtle problem with the build system (and 
completely unrelated to XPanelPeer, which had the unfortunate position 
of being in the wrong place at the wrong time =)  ).  My extreme 
apologies for heading in the wrong direction with this.

The real problem here is between lesstif and openmotif.  The 
PropMwmHints struct has a different size, and thus lesstif is 
incompatible with OpenJDK on 64-bit architectures.

This also explains why the Sun developers weren't seeing this (I assume 
they use openmotif), and a stray ALT_MOTIF_DIR setting in my build 
environment, left-over from a b14 build, led me on a wild goose chase 
last week.

Regards,
Francis


From dan at fabulich.com  Thu Aug  2 19:15:56 2007
From: dan at fabulich.com (Dan Fabulich)
Date: Thu, 2 Aug 2007 12:15:56 -0700 (Pacific Daylight Time)
Subject: encumbrances update
In-Reply-To: <46B1A0BB.4030405@sun.com>
References: <46AFAC07.6060608@sun.com>
	<Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>
	<46B1833C.7050705@sun.com> <Pine.WNT.4.63.0708020012310.2576@tbf>
	<46B1A0BB.4030405@sun.com>
Message-ID: <Pine.WNT.4.64.0708021201080.2848@SFENG02.RF.LAN>

Igor Nekrestyanov wrote:

> Of course i've tested these changes on all platforms including both 32 
> and 64 bit Windows. It is "believed to build and work" because we had 
> not performed full testing of openjdk binaries and we know that build 
> can be fragile due to different build environments.

I'm sure you've tested your changes in your build, and maybe that's all I 
can ask for, but I'm pretty sure the same could have been said of all of 
the previous builds, including the code drop we got in May and every build 
since, all of which, we know, don't work for anyone who isn't @sun.com 
(due to at least one missing file).

> My tests are not "ideal" for number of reasons:
> 2) I was using binary plugs created from my personal workspace.
>      I believe they should be the same as those to be published with b17 
> code drop but this is my assumption.

We know for certain that this assumption is FALSE.  No binary plug given 
to the public has ever built successfully on Windows; since your personal 
plugs have been working for you for months, we can conclude that there's 
something critically different between your plugs and the plugs we get.

But who knows?  Maybe we'll get lucky this time.  :-)

> BTW, for windows build you will need to build freetype.dll (i tried only 
> dll built with visualc). I believe that freetype build system does not 
> generate .dll on windows and therefore some manual tweaking of freetype 
> makefiles (and sources) might be necessary.

Sounds like fun! :-)  Will the documentation include a patch?  Or just 
some tips on how to make a .DLL?  What source files needed to change?

-Dan


From Igor.Nekrestyanov at Sun.COM  Thu Aug  2 19:39:01 2007
From: Igor.Nekrestyanov at Sun.COM (Igor Nekrestyanov)
Date: Thu, 02 Aug 2007 23:39:01 +0400
Subject: encumbrances update
In-Reply-To: <Pine.WNT.4.64.0708021201080.2848@SFENG02.RF.LAN>
References: <46AFAC07.6060608@sun.com>
	<Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>
	<46B1833C.7050705@sun.com>
	<Pine.WNT.4.63.0708020012310.2576@tbf> <46B1A0BB.4030405@sun.com>
	<Pine.WNT.4.64.0708021201080.2848@SFENG02.RF.LAN>
Message-ID: <46B232D5.6070809@sun.com>


>> My tests are not "ideal" for number of reasons:
>> 2) I was using binary plugs created from my personal workspace.
>>      I believe they should be the same as those to be published with 
>> b17 code drop but this is my assumption.
>
> We know for certain that this assumption is FALSE.  No binary plug 
> given to the public has ever built successfully on Windows; since your 
> personal plugs have been working for you for months, we can conclude 
> that there's something critically different between your plugs and the 
> plugs we get.
Prior to introduction of the "binary plugs" in the makefile (starting 
from the b16)
it was easy to use our internal product builds to import encumbered bits 
for openjdk build.
So, internal openjdk builds were easier.

However, since b16 we need to export image of binary plugs from the 
product build and
use these exported images for openjdk build.
I think (but i am not 100% sure) that same procedure is used to export 
binary plug bits for people outside of @sun.com.
So now there is fewer chance to have discrepancies (and after all 
t2k.lib is not needed anymore :) ).
> But who knows?  Maybe we'll get lucky this time.  :-)
I can hardly wait to get feedback on build status.
Hopefully this will be success report :)
>> BTW, for windows build you will need to build freetype.dll (i tried 
>> only dll built with visualc). I believe that freetype build system 
>> does not generate .dll on windows and therefore some manual tweaking 
>> of freetype makefiles (and sources) might be necessary.
>
> Sounds like fun! :-)  Will the documentation include a patch?  Or just 
> some tips on how to make a .DLL?  What source files needed to change?
No, there is no patch or detailed instruction in the openjdk docs.

You can follow "official" freetype approach - 
http://freetype.freedesktop.org/wiki/FreeType_DLL.
I believe it worked for me to prepare 32 bit binary.
Perhaps someday they will finally simplify it and we can just run make :)

In addition to described changes you may also want to tweak ftoptions.h 
to enable subpixel rendering.

-igor



From robilad at kaffe.org  Thu Aug  2 20:10:36 2007
From: robilad at kaffe.org (Dalibor Topic)
Date: Thu, 02 Aug 2007 22:10:36 +0200
Subject: encumbrances update
In-Reply-To: <46B232D5.6070809@sun.com>
References: <46AFAC07.6060608@sun.com>	<Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>	<46B1833C.7050705@sun.com>	<Pine.WNT.4.63.0708020012310.2576@tbf>
	<46B1A0BB.4030405@sun.com>	<Pine.WNT.4.64.0708021201080.2848@SFENG02.RF.LAN>
	<46B232D5.6070809@sun.com>
Message-ID: <46B23A3C.6070401@kaffe.org>

Igor Nekrestyanov wrote:

> You can follow "official" freetype approach -
> http://freetype.freedesktop.org/wiki/FreeType_DLL.
> I believe it worked for me to prepare 32 bit binary.
> Perhaps someday they will finally simplify it and we can just run make :)

>From looking at the easy mingw option, i think that could be folded into
the build system properly by adding -no-undefined to Makefile.am
containing libfreetype_la_LDFLAGS, and running autoconf -f -i.

cheers,
dalibor topic


From robilad at kaffe.org  Thu Aug  2 20:12:55 2007
From: robilad at kaffe.org (Dalibor Topic)
Date: Thu, 02 Aug 2007 22:12:55 +0200
Subject: encumbrances update
In-Reply-To: <46B23A3C.6070401@kaffe.org>
References: <46AFAC07.6060608@sun.com>	<Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>	<46B1833C.7050705@sun.com>	<Pine.WNT.4.63.0708020012310.2576@tbf>	<46B1A0BB.4030405@sun.com>	<Pine.WNT.4.64.0708021201080.2848@SFENG02.RF.LAN>	<46B232D5.6070809@sun.com>
	<46B23A3C.6070401@kaffe.org>
Message-ID: <46B23AC7.9020008@kaffe.org>

Dalibor Topic wrote:
> Igor Nekrestyanov wrote:
> 
>> You can follow "official" freetype approach -
>> http://freetype.freedesktop.org/wiki/FreeType_DLL.
>> I believe it worked for me to prepare 32 bit binary.
>> Perhaps someday they will finally simplify it and we can just run make :)
> 
>>From looking at the easy mingw option, i think that could be folded into
> the build system properly by adding -no-undefined to Makefile.am
> containing libfreetype_la_LDFLAGS, and running autoconf -f -i.

i was a bit too fast, that should have been autoreconf -f -i (for
regular autotools using projects, or bash autogen.sh for those that have
an autogen script instead).

cheers,
dalibor topic


From mark at klomp.org  Fri Aug  3 08:25:04 2007
From: mark at klomp.org (Mark Wielaard)
Date: Fri, 03 Aug 2007 10:25:04 +0200
Subject: JavaSE 7 build 17 is available at the openjdk.java.net website
In-Reply-To: <46B25CC4.6030303@Sun.COM>
References: <46B25CC4.6030303@Sun.COM>
Message-ID: <1186129504.3606.12.camel@dijkstra.wildebeest.org>

Hi Xiomara,

On Thu, 2007-08-02 at 15:37 -0700, Xiomara.Jayasena at Sun.COM wrote:
> The OpenJDK source, Compiler source and Jtreg binary for the promoted 
> JavaSE 7 build b17 is available under the openjdk 
> http://openjdk.java.net website under Source Code (direct link to 
> bundles:  http://download.java.net/openjdk/jdk7 )
> 
> The OpenJDK sources are also available at the subversion repository 
> http://openjdk.dev.java.net/source/browse/openjdk

Thanks. I see the changes listed at:
http://download.java.net/jdk7/changes/jdk7-b17.html
It might be a nice idea to include the ChangeLog with the announcement.
Is there any overview of what is being worked on/pending for b18?

The read-only mercurial mirror on icedtea.classpath.org did auto update
a few hours ago already without any manual interference necessary, so
you can also get the update through hg pull && hg update now if you use
that. (Now back at the old location on http://icedtea.classpath.org/hg/
but also on the iced-tea.org domain now.)
567 files updated, 0 files merged, 6398 files removed, 0 files unresolved

Most of those removed files seem to be the tests under j2se/test/ which
also aren't in the zip or on subversion anymore. Have they to moved
somewhere else?

Cheers,

Mark



From fw at deneb.enyo.de  Fri Aug  3 08:38:42 2007
From: fw at deneb.enyo.de (Florian Weimer)
Date: Fri, 03 Aug 2007 10:38:42 +0200
Subject: JavaSE 7 build 17 is available at the openjdk.java.net website
In-Reply-To: <1186129504.3606.12.camel@dijkstra.wildebeest.org> (Mark
	Wielaard's message of "Fri, 03 Aug 2007 10:25:04 +0200")
References: <46B25CC4.6030303@Sun.COM>
	<1186129504.3606.12.camel@dijkstra.wildebeest.org>
Message-ID: <87d4y5108d.fsf@mid.deneb.enyo.de>

* Mark Wielaard:

> Thanks. I see the changes listed at:
> http://download.java.net/jdk7/changes/jdk7-b17.html

Oh.  This link doesn't work for me.


From mark at klomp.org  Fri Aug  3 08:55:47 2007
From: mark at klomp.org (Mark Wielaard)
Date: Fri, 03 Aug 2007 10:55:47 +0200
Subject: JavaSE 7 build 17 is available at the openjdk.java.net website
In-Reply-To: <87d4y5108d.fsf@mid.deneb.enyo.de>
References: <46B25CC4.6030303@Sun.COM>
	<1186129504.3606.12.camel@dijkstra.wildebeest.org>
	<87d4y5108d.fsf@mid.deneb.enyo.de>
Message-ID: <1186131348.3606.20.camel@dijkstra.wildebeest.org>

On Fri, 2007-08-03 at 10:38 +0200, Florian Weimer wrote:
> > Thanks. I see the changes listed at:
> > http://download.java.net/jdk7/changes/jdk7-b17.html
> 
> Oh.  This link doesn't work for me.

That is strange. I swear it worked for me just a few minutes ago, but
now if gives:

        Not Found
        The requested object does not exist on this server. The link you
        followed is either outdated, inaccurate, or the server has been
        instructed not to let you have it.

Weird. Another good reason to include the actual change log with the
announcement message! :)

Cheers,

Mark



From volker.simonis at gmail.com  Fri Aug  3 13:57:43 2007
From: volker.simonis at gmail.com (Volker Simonis)
Date: Fri, 3 Aug 2007 15:57:43 +0200
Subject: Where have the tests gone???
Message-ID: <d65861880708030657q6019043i385b4609839671e6@mail.gmail.com>

Hi,

I was hoping that while time goes by more and more tests will appeare
(espacially in the hotspot directory)  but now I just realized that
the 'test/' subdirectories from the 'hotspot/' and 'j2se/' directories
dissapeared completely in Build b17. Has this happend intentionally?

What are the futuer plans for the jtreg tests in OpenJDK?

By the way, the build itself succedded without any problems on Suse
Linux 10 on x86. Good job!

Volker


From Jonathan.Gibbons at Sun.COM  Fri Aug  3 14:19:23 2007
From: Jonathan.Gibbons at Sun.COM (Jonathan Gibbons)
Date: Fri, 03 Aug 2007 07:19:23 -0700
Subject: Where have the tests gone???
In-Reply-To: <d65861880708030657q6019043i385b4609839671e6@mail.gmail.com>
References: <d65861880708030657q6019043i385b4609839671e6@mail.gmail.com>
Message-ID: <1ED26BBA-C48E-4CFF-A59C-D71028A6571B@sun.com>

I can't speak for what happened to all the tests, which surely sounds  
like an accident,
but I can say that that a bunch more tool tests will be made  
available soon.

-- Jon G


On Aug 3, 2007, at 6:57 AM, Volker Simonis wrote:

> Hi,
>
> I was hoping that while time goes by more and more tests will appeare
> (espacially in the hotspot directory)  but now I just realized that
> the 'test/' subdirectories from the 'hotspot/' and 'j2se/' directories
> dissapeared completely in Build b17. Has this happend intentionally?
>
> What are the futuer plans for the jtreg tests in OpenJDK?
>
> By the way, the build itself succedded without any problems on Suse
> Linux 10 on x86. Good job!
>
> Volker



From Sandeep.Konchady at Sun.COM  Fri Aug  3 16:34:15 2007
From: Sandeep.Konchady at Sun.COM (Sandeep Konchady)
Date: Fri, 03 Aug 2007 09:34:15 -0700
Subject: JavaSE 7 build 17 is available at the openjdk.java.net website
In-Reply-To: <1186131348.3606.20.camel@dijkstra.wildebeest.org>
References: <46B25CC4.6030303@Sun.COM>
	<1186129504.3606.12.camel@dijkstra.wildebeest.org>
	<87d4y5108d.fsf@mid.deneb.enyo.de>
	<1186131348.3606.20.camel@dijkstra.wildebeest.org>
Message-ID: <46B35907.1030302@Sun.COM>

Not sure what happened, the page seems to be working again.

- Sandeep

Mark Wielaard wrote:
> On Fri, 2007-08-03 at 10:38 +0200, Florian Weimer wrote:
>   
>>> Thanks. I see the changes listed at:
>>> http://download.java.net/jdk7/changes/jdk7-b17.html
>>>       
>> Oh.  This link doesn't work for me.
>>     
>
> That is strange. I swear it worked for me just a few minutes ago, but
> now if gives:
>
>         Not Found
>         The requested object does not exist on this server. The link you
>         followed is either outdated, inaccurate, or the server has been
>         instructed not to let you have it.
>
> Weird. Another good reason to include the actual change log with the
> announcement message! :)
>
> Cheers,
>
> Mark
>
>   

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070803/2e624aa3/attachment.html>

From Xiomara.Jayasena at Sun.COM  Fri Aug  3 17:01:02 2007
From: Xiomara.Jayasena at Sun.COM (Xiomara.Jayasena at Sun.COM)
Date: Fri, 03 Aug 2007 10:01:02 -0700
Subject: JavaSE 7 build 17 is available at the openjdk.java.net website
In-Reply-To: <1186129504.3606.12.camel@dijkstra.wildebeest.org>
References: <46B25CC4.6030303@Sun.COM>
	<1186129504.3606.12.camel@dijkstra.wildebeest.org>
Message-ID: <46B35F4E.6030500@Sun.COM>


Hi Mark,

See my comments embedded...

Mark Wielaard wrote:

>Hi Xiomara,
>
>On Thu, 2007-08-02 at 15:37 -0700, Xiomara.Jayasena at Sun.COM wrote:
>  
>
>>The OpenJDK source, Compiler source and Jtreg binary for the promoted 
>>JavaSE 7 build b17 is available under the openjdk 
>>http://openjdk.java.net website under Source Code (direct link to 
>>bundles:  http://download.java.net/openjdk/jdk7 )
>>
>>The OpenJDK sources are also available at the subversion repository 
>>http://openjdk.dev.java.net/source/browse/openjdk
>>    
>>
>
>Thanks. I see the changes listed at:
>http://download.java.net/jdk7/changes/jdk7-b17.html
>It might be a nice idea to include the ChangeLog with the announcement.
>  
>
I'll see into including it in the future.

>Is there any overview of what is being worked on/pending for b18?
>  
>
need to investigate what can be found

>The read-only mercurial mirror on icedtea.classpath.org did auto update
>a few hours ago already without any manual interference necessary, so
>you can also get the update through hg pull && hg update now if you use
>that. (Now back at the old location on http://icedtea.classpath.org/hg/
>but also on the iced-tea.org domain now.)
>567 files updated, 0 files merged, 6398 files removed, 0 files unresolved
>
>Most of those removed files seem to be the tests under j2se/test/ which
>also aren't in the zip or on subversion anymore. Have they to moved
>somewhere else?
>  
>
Inadvertently removed and will be place back in the next build.

-Xiomara

>Cheers,
>
>Mark
>
>  
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070803/9981eb0f/attachment.html>

From fw at deneb.enyo.de  Fri Aug  3 17:16:52 2007
From: fw at deneb.enyo.de (Florian Weimer)
Date: Fri, 03 Aug 2007 19:16:52 +0200
Subject: JavaSE 7 build 17 is available at the openjdk.java.net website
In-Reply-To: <46B35907.1030302@Sun.COM> (Sandeep Konchady's message of "Fri,
	03 Aug 2007 09:34:15 -0700")
References: <46B25CC4.6030303@Sun.COM>
	<1186129504.3606.12.camel@dijkstra.wildebeest.org>
	<87d4y5108d.fsf@mid.deneb.enyo.de>
	<1186131348.3606.20.camel@dijkstra.wildebeest.org>
	<46B35907.1030302@Sun.COM>
Message-ID: <878x8ssfln.fsf@mid.deneb.enyo.de>

* Sandeep Konchady:

> Not sure what happened, the page seems to be working again.

Not for me, I still receive this error message:

| Not Found
|
| The requested object does not exist on this server. The link you
| followed is either outdated, inaccurate, or the server has been
| instructed not to let you have it.


From Xiomara.Jayasena at Sun.COM  Fri Aug  3 17:19:33 2007
From: Xiomara.Jayasena at Sun.COM (Xiomara.Jayasena at Sun.COM)
Date: Fri, 03 Aug 2007 10:19:33 -0700
Subject: JavaSE 7 build 17 is available at the openjdk.java.net website
In-Reply-To: <878x8ssfln.fsf@mid.deneb.enyo.de>
References: <46B25CC4.6030303@Sun.COM>
	<1186129504.3606.12.camel@dijkstra.wildebeest.org>
	<87d4y5108d.fsf@mid.deneb.enyo.de>
	<1186131348.3606.20.camel@dijkstra.wildebeest.org>
	<46B35907.1030302@Sun.COM> <878x8ssfln.fsf@mid.deneb.enyo.de>
Message-ID: <46B363A5.7060609@Sun.COM>


I have no problem accessing the link below:
http://download.java.net/jdk7/changes/jdk7-b17.html

Is anybody else having problems?

-Xiomara

Florian Weimer wrote:

>* Sandeep Konchady:
>
>  
>
>>Not sure what happened, the page seems to be working again.
>>    
>>
>
>Not for me, I still receive this error message:
>
>| Not Found
>|
>| The requested object does not exist on this server. The link you
>| followed is either outdated, inaccurate, or the server has been
>| instructed not to let you have it.
>  
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070803/258dd446/attachment.html>

From fw at deneb.enyo.de  Fri Aug  3 17:27:06 2007
From: fw at deneb.enyo.de (Florian Weimer)
Date: Fri, 03 Aug 2007 19:27:06 +0200
Subject: JavaSE 7 build 17 is available at the openjdk.java.net website
In-Reply-To: <46B363A5.7060609@Sun.COM> (Xiomara Jayasena's message of "Fri,
	03 Aug 2007 10:19:33 -0700")
References: <46B25CC4.6030303@Sun.COM>
	<1186129504.3606.12.camel@dijkstra.wildebeest.org>
	<87d4y5108d.fsf@mid.deneb.enyo.de>
	<1186131348.3606.20.camel@dijkstra.wildebeest.org>
	<46B35907.1030302@Sun.COM> <878x8ssfln.fsf@mid.deneb.enyo.de>
	<46B363A5.7060609@Sun.COM>
Message-ID: <87tzrgr0k5.fsf@mid.deneb.enyo.de>

* Xiomara Jayasena:

> I have no problem accessing the link below:
> http://download.java.net/jdk7/changes/jdk7-b17.html

During further tests, the page suddenly appeared.  It's not a caching
issue on my side; I had disabled caching.  Very strange.


From Sandeep.Konchady at Sun.COM  Fri Aug  3 17:22:40 2007
From: Sandeep.Konchady at Sun.COM (Sandeep Konchady)
Date: Fri, 03 Aug 2007 10:22:40 -0700
Subject: JavaSE 7 build 17 is available at the openjdk.java.net website
In-Reply-To: <46B363A5.7060609@Sun.COM>
References: <46B25CC4.6030303@Sun.COM>
	<1186129504.3606.12.camel@dijkstra.wildebeest.org>
	<87d4y5108d.fsf@mid.deneb.enyo.de>
	<1186131348.3606.20.camel@dijkstra.wildebeest.org>
	<46B35907.1030302@Sun.COM>
	<878x8ssfln.fsf@mid.deneb.enyo.de> <46B363A5.7060609@Sun.COM>
Message-ID: <46B36460.50704@Sun.COM>

Hi Xiomara,

    Florian is right. Looks like what I saw initially was a cached 
version of the page. Please clear your cache and reload the page, you 
will get "Not Found" error page.

Thanks,
Sandeep

Xiomara.Jayasena at Sun.COM wrote:
>
> I have no problem accessing the link below:
> http://download.java.net/jdk7/changes/jdk7-b17.html
>
> Is anybody else having problems?
>
> -Xiomara
>
> Florian Weimer wrote:
>> * Sandeep Konchady:
>>
>>   
>>> Not sure what happened, the page seems to be working again.
>>>     
>>
>> Not for me, I still receive this error message:
>>
>> | Not Found
>> |
>> | The requested object does not exist on this server. The link you
>> | followed is either outdated, inaccurate, or the server has been
>> | instructed not to let you have it.
>>   
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070803/1648fee0/attachment.html>

From fkung at redhat.com  Fri Aug  3 17:28:01 2007
From: fkung at redhat.com (Francis Kung)
Date: Fri, 03 Aug 2007 13:28:01 -0400
Subject: JavaSE 7 build 17 is available at the openjdk.java.net website
In-Reply-To: <46B363A5.7060609@Sun.COM>
References: <46B25CC4.6030303@Sun.COM>	<1186129504.3606.12.camel@dijkstra.wildebeest.org>	<87d4y5108d.fsf@mid.deneb.enyo.de>	<1186131348.3606.20.camel@dijkstra.wildebeest.org>	<46B35907.1030302@Sun.COM>
	<878x8ssfln.fsf@mid.deneb.enyo.de> <46B363A5.7060609@Sun.COM>
Message-ID: <46B365A1.3020209@redhat.com>

The page seems to be going up and down sporadically - it was working a 
few hours ago, then returning "Not Found" a few minutes ago, and seems 
to have *just* come back up again.

Cheers,
Francis


Xiomara.Jayasena at Sun.COM wrote:
> 
> I have no problem accessing the link below:
> http://download.java.net/jdk7/changes/jdk7-b17.html
> 
> Is anybody else having problems?
> 
> -Xiomara
> 
> Florian Weimer wrote:
>> * Sandeep Konchady:
>>
>>   
>>> Not sure what happened, the page seems to be working again.
>>>     
>>
>> Not for me, I still receive this error message:
>>
>> | Not Found
>> |
>> | The requested object does not exist on this server. The link you
>> | followed is either outdated, inaccurate, or the server has been
>> | instructed not to let you have it.
>>   
> 


From David.Herron at Sun.COM  Fri Aug  3 17:42:04 2007
From: David.Herron at Sun.COM (David Herron)
Date: Fri, 03 Aug 2007 10:42:04 -0700
Subject: JavaSE 7 build 17 is available at the openjdk.java.net website
In-Reply-To: <46B365A1.3020209@redhat.com>
References: <46B25CC4.6030303@Sun.COM>
	<1186129504.3606.12.camel@dijkstra.wildebeest.org>
	<87d4y5108d.fsf@mid.deneb.enyo.de>
	<1186131348.3606.20.camel@dijkstra.wildebeest.org>
	<46B35907.1030302@Sun.COM>
	<878x8ssfln.fsf@mid.deneb.enyo.de> <46B363A5.7060609@Sun.COM>
	<46B365A1.3020209@redhat.com>
Message-ID: <46B368EC.20400@sun.com>

Weird, yesterday I visited that page .. then shutdown my laptop, just 
rebooted, and I, too, get Not Found.

I believe there are multiple servers involved with providing 
download.java.net.  These servers were among the ones affected last week 
by that power outage which knocked out several sites.  I've sent a note 
to the alias where we discuss that server, to get some advice.

- David Herron



Francis Kung wrote:
> The page seems to be going up and down sporadically - it was working a 
> few hours ago, then returning "Not Found" a few minutes ago, and seems 
> to have *just* come back up again.
>
> Cheers,
> Francis
>
>
> Xiomara.Jayasena at Sun.COM wrote:
>>
>> I have no problem accessing the link below:
>> http://download.java.net/jdk7/changes/jdk7-b17.html
>>
>> Is anybody else having problems?
>>
>> -Xiomara
>>
>> Florian Weimer wrote:
>>> * Sandeep Konchady:
>>>
>>>  
>>>> Not sure what happened, the page seems to be working again.
>>>>     
>>>
>>> Not for me, I still receive this error message:
>>>
>>> | Not Found
>>> |
>>> | The requested object does not exist on this server. The link you
>>> | followed is either outdated, inaccurate, or the server has been
>>> | instructed not to let you have it.
>>>   
>>



From Xiomara.Jayasena at Sun.COM  Fri Aug  3 17:41:24 2007
From: Xiomara.Jayasena at Sun.COM (Xiomara.Jayasena at Sun.COM)
Date: Fri, 03 Aug 2007 10:41:24 -0700
Subject: JavaSE 7 build 17 is available at the openjdk.java.net website
In-Reply-To: <46B365A1.3020209@redhat.com>
References: <46B25CC4.6030303@Sun.COM>
	<1186129504.3606.12.camel@dijkstra.wildebeest.org>
	<87d4y5108d.fsf@mid.deneb.enyo.de>
	<1186131348.3606.20.camel@dijkstra.wildebeest.org>
	<46B35907.1030302@Sun.COM>
	<878x8ssfln.fsf@mid.deneb.enyo.de> <46B363A5.7060609@Sun.COM>
	<46B365A1.3020209@redhat.com>
Message-ID: <46B368C4.4040808@Sun.COM>


All,

Thanks to those of you who responded.
This issue is currently being investigated -- stay tuned.

-Xiomara

Francis Kung wrote:

> The page seems to be going up and down sporadically - it was working a 
> few hours ago, then returning "Not Found" a few minutes ago, and seems 
> to have *just* come back up again.
>
> Cheers,
> Francis
>
>
> Xiomara.Jayasena at Sun.COM wrote:
>
>>
>> I have no problem accessing the link below:
>> http://download.java.net/jdk7/changes/jdk7-b17.html
>>
>> Is anybody else having problems?
>>
>> -Xiomara
>>
>> Florian Weimer wrote:
>>
>>> * Sandeep Konchady:
>>>
>>>  
>>>
>>>> Not sure what happened, the page seems to be working again.
>>>>     
>>>
>>>
>>> Not for me, I still receive this error message:
>>>
>>> | Not Found
>>> |
>>> | The requested object does not exist on this server. The link you
>>> | followed is either outdated, inaccurate, or the server has been
>>> | instructed not to let you have it.
>>>   
>>
>>



From Xiomara.Jayasena at Sun.COM  Fri Aug  3 17:45:45 2007
From: Xiomara.Jayasena at Sun.COM (Xiomara.Jayasena at Sun.COM)
Date: Fri, 03 Aug 2007 10:45:45 -0700
Subject: JavaSE 7 build 17 is available at the openjdk.java.net website
In-Reply-To: <46B368EC.20400@sun.com>
References: <46B25CC4.6030303@Sun.COM>
	<1186129504.3606.12.camel@dijkstra.wildebeest.org>
	<87d4y5108d.fsf@mid.deneb.enyo.de>
	<1186131348.3606.20.camel@dijkstra.wildebeest.org>
	<46B35907.1030302@Sun.COM>
	<878x8ssfln.fsf@mid.deneb.enyo.de> <46B363A5.7060609@Sun.COM>
	<46B365A1.3020209@redhat.com> <46B368EC.20400@sun.com>
Message-ID: <46B369C9.6090800@Sun.COM>


David Herron wrote:

> Weird, yesterday I visited that page .. then shutdown my laptop, just 
> rebooted, and I, too, get Not Found.
>
> I believe there are multiple servers involved with providing 
> download.java.net.  These servers were among the ones affected last 
> week by that power outage which knocked out several sites.  I've sent 
> a note to the alias where we discuss that server, to get some advice.

Thank you David!  I have too :-)

-Xiomara

>
> - David Herron
>
>
>
> Francis Kung wrote:
>
>> The page seems to be going up and down sporadically - it was working 
>> a few hours ago, then returning "Not Found" a few minutes ago, and 
>> seems to have *just* come back up again.
>>
>> Cheers,
>> Francis
>>
>>
>> Xiomara.Jayasena at Sun.COM wrote:
>>
>>>
>>> I have no problem accessing the link below:
>>> http://download.java.net/jdk7/changes/jdk7-b17.html
>>>
>>> Is anybody else having problems?
>>>
>>> -Xiomara
>>>
>>> Florian Weimer wrote:
>>>
>>>> * Sandeep Konchady:
>>>>
>>>>  
>>>>
>>>>> Not sure what happened, the page seems to be working again.
>>>>>     
>>>>
>>>>
>>>> Not for me, I still receive this error message:
>>>>
>>>> | Not Found
>>>> |
>>>> | The requested object does not exist on this server. The link you
>>>> | followed is either outdated, inaccurate, or the server has been
>>>> | instructed not to let you have it.
>>>>   
>>>
>>>
>



From alexanderschunk at t-online.de  Fri Aug  3 20:12:00 2007
From: alexanderschunk at t-online.de (Alexander Schunk)
Date: 03 Aug 2007 20:12 GMT
Subject: discuss Digest, Vol 4, Issue 5
In-Reply-To: <mailman.1.1186167600.19505.discuss@openjdk.java.net>
References: <mailman.1.1186167600.19505.discuss@openjdk.java.net>
Message-ID: <1IH3VA-1LsBt20@fwd32.aul.t-online.de>

Hi,
i a have recently subscribed to this list and am working on an API for linear Algebra operations which may be part of the OpenJDK SDK. 
This API currently provides methods for 2D and 3D Vector operations, Matrix operations an Complex Number operations.
I have also written a simple Fraction Number class and some other basic algorithms - faculty, binomial etc. which may be used for statistics.
I am also working on the Java SDK for the Windows plattform however i do not build very new release so the current discussion on Windows Vista builds is interesting.
Regards
Alexander
 <mailto:discuss-request at openjdk.java.net> schrieb:
> Send discuss mailing list submissions to
>     mailto:discuss at openjdk.java.net
> 
> To subscribe or unsubscribe via the World Wide Web, visit
>     http://mail.openjdk.java.net/mailman/listinfo/discuss
> or, via email, send a message with subject or body 'help' to
>     mailto:discuss-request at openjdk.java.net
> 
> You can reach the person managing the list at
>     mailto:discuss-owner at openjdk.java.net
> 
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of discuss digest..."
> 
> 
> Today's Topics:
> 
>    1. Re: JavaSE 7 build 17 is available at the openjdk.java.net
>       website (mailto:Xiomara.Jayasena at Sun.COM)
> 
> 
> ----------------------------------------------------------------------
> 
> Message: 1
> Date: Fri, 03 Aug 2007 10:45:45 -0700
> From: mailto:Xiomara.Jayasena at Sun.COM
> Subject: Re: JavaSE 7 build 17 is available at the openjdk.java.net
>     website
> To: David Herron <mailto:David.Herron at Sun.COM>
> Cc: mailto:discuss at openjdk.java.net, mailto:Sandeep.Konchady at Sun.COM
> Message-ID: <mailto:46B369C9.6090800 at Sun.COM>
> Content-Type: text/plain; format=flowed; charset=ISO-8859-1
> 
> 
> David Herron wrote:
> 
> > Weird, yesterday I visited that page .. then shutdown my laptop, just 
> > rebooted, and I, too, get Not Found.
> >
> > I believe there are multiple servers involved with providing 
> > download.java.net.  These servers were among the ones affected last 
> > week by that power outage which knocked out several sites.  I've sent 
> > a note to the alias where we discuss that server, to get some advice.
> 
> Thank you David!  I have too :-)
> 
> -Xiomara
> 
> >
> > - David Herron
> >
> >
> >
> > Francis Kung wrote:
> >
> >> The page seems to be going up and down sporadically - it was working 
> >> a few hours ago, then returning "Not Found" a few minutes ago, and 
> >> seems to have *just* come back up again.
> >>
> >> Cheers,
> >> Francis
> >>
> >>
> >> mailto:Xiomara.Jayasena at Sun.COM wrote:
> >>
> >>>
> >>> I have no problem accessing the link below:
> >>> http://download.java.net/jdk7/changes/jdk7-b17.html
> >>>
> >>> Is anybody else having problems?
> >>>
> >>> -Xiomara
> >>>
> >>> Florian Weimer wrote:
> >>>
> >>>> * Sandeep Konchady:
> >>>>
> >>>>  
> >>>>
> >>>>> Not sure what happened, the page seems to be working again.
> >>>>>      
> >>>>
> >>>>
> >>>> Not for me, I still receive this error message:
> >>>>
> >>>> | Not Found
> >>>> |
> >>>> | The requested object does not exist on this server. The link you
> >>>> | followed is either outdated, inaccurate, or the server has been
> >>>> | instructed not to let you have it.
> >>>>   
> >>>
> >>>
> >
> 
> 
> 
> End of discuss Digest, Vol 4, Issue 5
> *************************************
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070803/fd70b054/attachment.html>

From Phil.Race at Sun.COM  Sat Aug  4 03:59:58 2007
From: Phil.Race at Sun.COM (Phil Race)
Date: Fri, 03 Aug 2007 20:59:58 -0700
Subject: encumbrances update
In-Reply-To: <096901c7d635$71959e30$802ca8c0@XPWork>
References: <096901c7d635$71959e30$802ca8c0@XPWork>
Message-ID: <46B3F9BE.9@sun.com>

freetypecheck.c(42) : warning C4003: not enough actual parameters for macro
'QUO
TEME'

Did it really print a newline in the middle of that name ?

your make/tools/freetypecheck/freetypecheck.c  may be corrupt.

the macro is  defined there  and  its hard to  imagine how it could  be 
affected by anything exernal.

-phil.

Ted Neward wrote:
> OK... Using a private drop of Ivan's freetype build, with an "SVN update" of
> the source base, I get this:
>
> CYGWIN:Ted at XPJAVA:/cygdrive/c/Prg/OpenJDK/openjdk/control/make
> $ make sanity
> make[1]: Entering directory `/cygdrive/c/Prg/OpenJDK/openjdk/j2se/make'
> make[2]: Entering directory
> `/cygdrive/c/Prg/OpenJDK/openjdk/j2se/make/tools/fre
> etypecheck'
> Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 13.10.3077 for 80x86
> Copyright (C) Microsoft Corporation 1984-2002. All rights reserved.
>
> freetypecheck.c
> freetypecheck.c(42) : warning C4003: not enough actual parameters for macro
> 'QUO
> TEME'
> freetypecheck.c(42) : error C2059: syntax error : ')'
> freetypecheck.c(45) : warning C4003: not enough actual parameters for macro
> 'QUO
> TEME'
> freetypecheck.c(45) : error C2059: syntax error : ')'
> freetypecheck.c(49) : error C2143: syntax error : missing ')' before '&'
> freetypecheck.c(49) : error C2143: syntax error : missing '{' before '&'
> freetypecheck.c(49) : error C2059: syntax error : '&'
> freetypecheck.c(49) : error C2059: syntax error : ')'
> freetypecheck.c(50) : error C2059: syntax error : ','
> freetypecheck.c(50) : error C2143: syntax error : missing ')' before '&'
> freetypecheck.c(50) : error C2143: syntax error : missing '{' before '&'
> freetypecheck.c(50) : error C2059: syntax error : '&'
> freetypecheck.c(50) : error C2059: syntax error : ')'
> freetypecheck.c(51) : error C2059: syntax error : ','
> freetypecheck.c(51) : error C2143: syntax error : missing ')' before
> 'string'
> freetypecheck.c(51) : error C2143: syntax error : missing '{' before
> 'string'
> freetypecheck.c(51) : error C2059: syntax error : '<Unknown>'
> freetypecheck.c(51) : error C2059: syntax error : ')'
> freetypecheck.c(53) : error C2143: syntax error : missing ')' before
> 'string'
> freetypecheck.c(53) : error C2143: syntax error : missing '{' before
> 'string'
> freetypecheck.c(53) : error C2059: syntax error : '<Unknown>'
> freetypecheck.c(53) : error C2059: syntax error : ')'
> freetypecheck.c(54) : error C2059: syntax error : 'if'
> freetypecheck.c(54) : warning C4003: not enough actual parameters for macro
> 'QUO
> TEME'
> freetypecheck.c(58) : error C2059: syntax error : 'return'
> freetypecheck.c(59) : error C2059: syntax error : '}'
> make[2]: ***
> [c:/Prg/OpenJDK/openjdk/control/build/WINDOW~1/tmp/freetype_version
> check] Error 2
> make[2]: Leaving directory
> `/cygdrive/c/Prg/OpenJDK/openjdk/j2se/make/tools/free
> typecheck'
> make[1]: [sane-freetype] Error 2 (ignored)
> /bin/sh:
> c:/Prg/OpenJDK/openjdk/control/build/WINDOW~1/tmp/freetype_versioncheck
> .exe: No such file or directory
> make[1]: Leaving directory `/cygdrive/c/Prg/OpenJDK/openjdk/j2se/make'
>
> Somebody have an idea what's going on here? (Obviously, freetypecheck.c
> isn't building, but I'm not sure why, beyond the obvious "there's a macro
> without enough parameters" as declared on line 42....)
>
> Ted Neward
> Java, .NET, XML Services
> Consulting, Teaching, Speaking, Writing
> http://www.tedneward.com
>  
>
>   
>> -----Original Message-----
>> From: build-dev-bounces at openjdk.java.net [mailto:build-dev-
>> bounces at openjdk.java.net] On Behalf Of Igor Nekrestyanov
>> Sent: Thursday, August 02, 2007 12:39 PM
>> To: Dan Fabulich
>> Cc: discuss at openjdk.java.net; Anthony Petrov; Phil Race; build-
>> dev at openjdk.java.net
>> Subject: Re: encumbrances update
>>
>>
>>     
>>>> My tests are not "ideal" for number of reasons:
>>>> 2) I was using binary plugs created from my personal workspace.
>>>>      I believe they should be the same as those to be published with
>>>> b17 code drop but this is my assumption.
>>>>         
>>> We know for certain that this assumption is FALSE.  No binary plug
>>> given to the public has ever built successfully on Windows; since your
>>> personal plugs have been working for you for months, we can conclude
>>> that there's something critically different between your plugs and the
>>> plugs we get.
>>>       
>> Prior to introduction of the "binary plugs" in the makefile (starting
>> from the b16)
>> it was easy to use our internal product builds to import encumbered bits
>> for openjdk build.
>> So, internal openjdk builds were easier.
>>
>> However, since b16 we need to export image of binary plugs from the
>> product build and
>> use these exported images for openjdk build.
>> I think (but i am not 100% sure) that same procedure is used to export
>> binary plug bits for people outside of @sun.com.
>> So now there is fewer chance to have discrepancies (and after all
>> t2k.lib is not needed anymore :) ).
>>     
>>> But who knows?  Maybe we'll get lucky this time.  :-)
>>>       
>> I can hardly wait to get feedback on build status.
>> Hopefully this will be success report :)
>>     
>>>> BTW, for windows build you will need to build freetype.dll (i tried
>>>> only dll built with visualc). I believe that freetype build system
>>>> does not generate .dll on windows and therefore some manual tweaking
>>>> of freetype makefiles (and sources) might be necessary.
>>>>         
>>> Sounds like fun! :-)  Will the documentation include a patch?  Or just
>>> some tips on how to make a .DLL?  What source files needed to change?
>>>       
>> No, there is no patch or detailed instruction in the openjdk docs.
>>
>> You can follow "official" freetype approach -
>> http://freetype.freedesktop.org/wiki/FreeType_DLL.
>> I believe it worked for me to prepare 32 bit binary.
>> Perhaps someday they will finally simplify it and we can just run make :)
>>
>> In addition to described changes you may also want to tweak ftoptions.h
>> to enable subpixel rendering.
>>
>> -igor
>>
>> No virus found in this incoming message.
>> Checked by AVG Free Edition.
>> Version: 7.5.476 / Virus Database: 269.11.2/933 - Release Date: 8/2/2007
>> 2:22 PM
>>
>>     
>
> No virus found in this outgoing message.
> Checked by AVG Free Edition. 
> Version: 7.5.476 / Virus Database: 269.11.2/933 - Release Date: 8/2/2007
> 2:22 PM
>  
>
>   



From Phil.Race at Sun.COM  Sat Aug  4 04:02:19 2007
From: Phil.Race at Sun.COM (Phil Race)
Date: Fri, 03 Aug 2007 21:02:19 -0700
Subject: encumbrances update
In-Reply-To: <096a01c7d635$d226d700$802ca8c0@XPWork>
References: <096a01c7d635$d226d700$802ca8c0@XPWork>
Message-ID: <46B3FA4B.9020603@sun.com>

Ted Neward wrote:
> One other question: does the new build still need an
> ALT_CLOSED_JDK_IMPORT_PATH that points to a valid JDK 7 instance?
>   

There was no change in the overall build architecture in b17
In that regard whatever Kelly did for b16 is  still the case in b17,
which AFAIK means it can be just the sparse binary plugs *OR* a full JDK ..
 
-phil.
> Ted Neward
> Java, .NET, XML Services
> Consulting, Teaching, Speaking, Writing
> http://www.tedneward.com
>  
>
>   
>> -----Original Message-----
>> From: build-dev-bounces at openjdk.java.net [mailto:build-dev-
>> bounces at openjdk.java.net] On Behalf Of Igor Nekrestyanov
>> Sent: Thursday, August 02, 2007 12:39 PM
>> To: Dan Fabulich
>> Cc: discuss at openjdk.java.net; Anthony Petrov; Phil Race; build-
>> dev at openjdk.java.net
>> Subject: Re: encumbrances update
>>
>>
>>     
>>>> My tests are not "ideal" for number of reasons:
>>>> 2) I was using binary plugs created from my personal workspace.
>>>>      I believe they should be the same as those to be published with
>>>> b17 code drop but this is my assumption.
>>>>         
>>> We know for certain that this assumption is FALSE.  No binary plug
>>> given to the public has ever built successfully on Windows; since your
>>> personal plugs have been working for you for months, we can conclude
>>> that there's something critically different between your plugs and the
>>> plugs we get.
>>>       
>> Prior to introduction of the "binary plugs" in the makefile (starting
>> from the b16)
>> it was easy to use our internal product builds to import encumbered bits
>> for openjdk build.
>> So, internal openjdk builds were easier.
>>
>> However, since b16 we need to export image of binary plugs from the
>> product build and
>> use these exported images for openjdk build.
>> I think (but i am not 100% sure) that same procedure is used to export
>> binary plug bits for people outside of @sun.com.
>> So now there is fewer chance to have discrepancies (and after all
>> t2k.lib is not needed anymore :) ).
>>     
>>> But who knows?  Maybe we'll get lucky this time.  :-)
>>>       
>> I can hardly wait to get feedback on build status.
>> Hopefully this will be success report :)
>>     
>>>> BTW, for windows build you will need to build freetype.dll (i tried
>>>> only dll built with visualc). I believe that freetype build system
>>>> does not generate .dll on windows and therefore some manual tweaking
>>>> of freetype makefiles (and sources) might be necessary.
>>>>         
>>> Sounds like fun! :-)  Will the documentation include a patch?  Or just
>>> some tips on how to make a .DLL?  What source files needed to change?
>>>       
>> No, there is no patch or detailed instruction in the openjdk docs.
>>
>> You can follow "official" freetype approach -
>> http://freetype.freedesktop.org/wiki/FreeType_DLL.
>> I believe it worked for me to prepare 32 bit binary.
>> Perhaps someday they will finally simplify it and we can just run make :)
>>
>> In addition to described changes you may also want to tweak ftoptions.h
>> to enable subpixel rendering.
>>
>> -igor
>>
>> No virus found in this incoming message.
>> Checked by AVG Free Edition.
>> Version: 7.5.476 / Virus Database: 269.11.2/933 - Release Date: 8/2/2007
>> 2:22 PM
>>
>>     
>
> No virus found in this outgoing message.
> Checked by AVG Free Edition. 
> Version: 7.5.476 / Virus Database: 269.11.2/933 - Release Date: 8/2/2007
> 2:22 PM
>  
>
>   



From Igor.Nekrestyanov at Sun.COM  Sat Aug  4 09:25:01 2007
From: Igor.Nekrestyanov at Sun.COM (Igor Nekrestyanov)
Date: Sat, 04 Aug 2007 13:25:01 +0400
Subject: encumbrances update
In-Reply-To: <099a01c7d675$670ea840$802ca8c0@XPWork>
References: <099a01c7d675$670ea840$802ca8c0@XPWork>
Message-ID: <46B445ED.6040600@sun.com>

 This is actually bug.
 make/common/Defs.gmk includes Platform.gmk before it sets OPENJDK variable.
 As a result REQUIRED_FREETYPE_VERSION is never set.
(BTW, this is because we always set OPENJDK=true manually because our 
internal builds have access to closed_src.
 So, yet one problem that sneak in because of difference in the 
environments)

 This seems only affect sanity check but we still have to fix it.
And this should be reproducible on any platform, not just windows.
 
Meanwhile please set environment variable OPENJDK=true prior to starting 
build
or pass it in the command line as you did.

I do not know what happens in the second case when build fails for you.
I'll try clean ws from openjdk.dev.java.net and try to reproduce it.

-igor

Ted Neward wrote:
> The newline was line-wrapping from my console window. I checked the C code,
> it's fine. The problem is that the REQUIRED_FREETYPE_VERSION macro wasn't
> being set inside the sanity makefile checks unless I explicitly put
> "OPENJDK=1" (which I think should be "OPENJDK=true") on the build command
> line. "make sanity OPENJDK=true" works, trying a full build now.
>
> Ted Neward
> Java, .NET, XML Services
> Consulting, Teaching, Speaking, Writing
> http://www.tedneward.com
>  
>
>   
>> -----Original Message-----
>> From: Phil.Race at Sun.COM [mailto:Phil.Race at Sun.COM]
>> Sent: Friday, August 03, 2007 9:00 PM
>> To: Ted Neward
>> Cc: 'Igor Nekrestyanov'; 'Dan Fabulich'; discuss at openjdk.java.net;
>> 'Anthony Petrov'; build-dev at openjdk.java.net
>> Subject: Re: encumbrances update
>>
>> freetypecheck.c(42) : warning C4003: not enough actual parameters for
>> macro
>> 'QUO
>> TEME'
>>
>> Did it really print a newline in the middle of that name ?
>>
>> your make/tools/freetypecheck/freetypecheck.c  may be corrupt.
>>
>> the macro is  defined there  and  its hard to  imagine how it could  be
>> affected by anything exernal.
>>
>> -phil.
>>
>> Ted Neward wrote:
>>     
>>> OK... Using a private drop of Ivan's freetype build, with an "SVN
>>>       
>> update" of
>>     
>>> the source base, I get this:
>>>
>>> CYGWIN:Ted at XPJAVA:/cygdrive/c/Prg/OpenJDK/openjdk/control/make
>>> $ make sanity
>>> make[1]: Entering directory `/cygdrive/c/Prg/OpenJDK/openjdk/j2se/make'
>>> make[2]: Entering directory
>>> `/cygdrive/c/Prg/OpenJDK/openjdk/j2se/make/tools/fre
>>> etypecheck'
>>> Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 13.10.3077 for
>>>       
>> 80x86
>>     
>>> Copyright (C) Microsoft Corporation 1984-2002. All rights reserved.
>>>
>>> freetypecheck.c
>>> freetypecheck.c(42) : warning C4003: not enough actual parameters for
>>>       
>> macro
>>     
>>> 'QUO
>>> TEME'
>>> freetypecheck.c(42) : error C2059: syntax error : ')'
>>> freetypecheck.c(45) : warning C4003: not enough actual parameters for
>>>       
>> macro
>>     
>>> 'QUO
>>> TEME'
>>> freetypecheck.c(45) : error C2059: syntax error : ')'
>>> freetypecheck.c(49) : error C2143: syntax error : missing ')' before '&'
>>> freetypecheck.c(49) : error C2143: syntax error : missing '{' before '&'
>>> freetypecheck.c(49) : error C2059: syntax error : '&'
>>> freetypecheck.c(49) : error C2059: syntax error : ')'
>>> freetypecheck.c(50) : error C2059: syntax error : ','
>>> freetypecheck.c(50) : error C2143: syntax error : missing ')' before '&'
>>> freetypecheck.c(50) : error C2143: syntax error : missing '{' before '&'
>>> freetypecheck.c(50) : error C2059: syntax error : '&'
>>> freetypecheck.c(50) : error C2059: syntax error : ')'
>>> freetypecheck.c(51) : error C2059: syntax error : ','
>>> freetypecheck.c(51) : error C2143: syntax error : missing ')' before
>>> 'string'
>>> freetypecheck.c(51) : error C2143: syntax error : missing '{' before
>>> 'string'
>>> freetypecheck.c(51) : error C2059: syntax error : '<Unknown>'
>>> freetypecheck.c(51) : error C2059: syntax error : ')'
>>> freetypecheck.c(53) : error C2143: syntax error : missing ')' before
>>> 'string'
>>> freetypecheck.c(53) : error C2143: syntax error : missing '{' before
>>> 'string'
>>> freetypecheck.c(53) : error C2059: syntax error : '<Unknown>'
>>> freetypecheck.c(53) : error C2059: syntax error : ')'
>>> freetypecheck.c(54) : error C2059: syntax error : 'if'
>>> freetypecheck.c(54) : warning C4003: not enough actual parameters for
>>>       
>> macro
>>     
>>> 'QUO
>>> TEME'
>>> freetypecheck.c(58) : error C2059: syntax error : 'return'
>>> freetypecheck.c(59) : error C2059: syntax error : '}'
>>> make[2]: ***
>>> [c:/Prg/OpenJDK/openjdk/control/build/WINDOW~1/tmp/freetype_version
>>> check] Error 2
>>> make[2]: Leaving directory
>>> `/cygdrive/c/Prg/OpenJDK/openjdk/j2se/make/tools/free
>>> typecheck'
>>> make[1]: [sane-freetype] Error 2 (ignored)
>>> /bin/sh:
>>> c:/Prg/OpenJDK/openjdk/control/build/WINDOW~1/tmp/freetype_versioncheck
>>> .exe: No such file or directory
>>> make[1]: Leaving directory `/cygdrive/c/Prg/OpenJDK/openjdk/j2se/make'
>>>
>>> Somebody have an idea what's going on here? (Obviously, freetypecheck.c
>>> isn't building, but I'm not sure why, beyond the obvious "there's a
>>>       
>> macro
>>     
>>> without enough parameters" as declared on line 42....)
>>>
>>> Ted Neward
>>> Java, .NET, XML Services
>>> Consulting, Teaching, Speaking, Writing
>>> http://www.tedneward.com
>>>
>>>
>>>
>>>       
>>>> -----Original Message-----
>>>> From: build-dev-bounces at openjdk.java.net [mailto:build-dev-
>>>> bounces at openjdk.java.net] On Behalf Of Igor Nekrestyanov
>>>> Sent: Thursday, August 02, 2007 12:39 PM
>>>> To: Dan Fabulich
>>>> Cc: discuss at openjdk.java.net; Anthony Petrov; Phil Race; build-
>>>> dev at openjdk.java.net
>>>> Subject: Re: encumbrances update
>>>>
>>>>
>>>>
>>>>         
>>>>>> My tests are not "ideal" for number of reasons:
>>>>>> 2) I was using binary plugs created from my personal workspace.
>>>>>>      I believe they should be the same as those to be published with
>>>>>> b17 code drop but this is my assumption.
>>>>>>
>>>>>>             
>>>>> We know for certain that this assumption is FALSE.  No binary plug
>>>>> given to the public has ever built successfully on Windows; since your
>>>>> personal plugs have been working for you for months, we can conclude
>>>>> that there's something critically different between your plugs and the
>>>>> plugs we get.
>>>>>
>>>>>           
>>>> Prior to introduction of the "binary plugs" in the makefile (starting
>>>> from the b16)
>>>> it was easy to use our internal product builds to import encumbered
>>>>         
>> bits
>>     
>>>> for openjdk build.
>>>> So, internal openjdk builds were easier.
>>>>
>>>> However, since b16 we need to export image of binary plugs from the
>>>> product build and
>>>> use these exported images for openjdk build.
>>>> I think (but i am not 100% sure) that same procedure is used to export
>>>> binary plug bits for people outside of @sun.com.
>>>> So now there is fewer chance to have discrepancies (and after all
>>>> t2k.lib is not needed anymore :) ).
>>>>
>>>>         
>>>>> But who knows?  Maybe we'll get lucky this time.  :-)
>>>>>
>>>>>           
>>>> I can hardly wait to get feedback on build status.
>>>> Hopefully this will be success report :)
>>>>
>>>>         
>>>>>> BTW, for windows build you will need to build freetype.dll (i tried
>>>>>> only dll built with visualc). I believe that freetype build system
>>>>>> does not generate .dll on windows and therefore some manual tweaking
>>>>>> of freetype makefiles (and sources) might be necessary.
>>>>>>
>>>>>>             
>>>>> Sounds like fun! :-)  Will the documentation include a patch?  Or just
>>>>> some tips on how to make a .DLL?  What source files needed to change?
>>>>>
>>>>>           
>>>> No, there is no patch or detailed instruction in the openjdk docs.
>>>>
>>>> You can follow "official" freetype approach -
>>>> http://freetype.freedesktop.org/wiki/FreeType_DLL.
>>>> I believe it worked for me to prepare 32 bit binary.
>>>> Perhaps someday they will finally simplify it and we can just run make
>>>>         
>> :)
>>     
>>>> In addition to described changes you may also want to tweak ftoptions.h
>>>> to enable subpixel rendering.
>>>>
>>>> -igor
>>>>
>>>> No virus found in this incoming message.
>>>> Checked by AVG Free Edition.
>>>> Version: 7.5.476 / Virus Database: 269.11.2/933 - Release Date:
>>>>         
>> 8/2/2007
>>     
>>>> 2:22 PM
>>>>
>>>>
>>>>         
>>> No virus found in this outgoing message.
>>> Checked by AVG Free Edition.
>>> Version: 7.5.476 / Virus Database: 269.11.2/933 - Release Date: 8/2/2007
>>> 2:22 PM
>>>
>>>
>>>
>>>       
>> No virus found in this incoming message.
>> Checked by AVG Free Edition.
>> Version: 7.5.476 / Virus Database: 269.11.2/933 - Release Date: 8/2/2007
>> 2:22 PM
>>
>>     
>
> No virus found in this outgoing message.
> Checked by AVG Free Edition. 
> Version: 7.5.476 / Virus Database: 269.11.2/933 - Release Date: 8/2/2007
> 2:22 PM
>  
>
>   



From Kelly.Ohair at Sun.COM  Thu Aug  9 15:16:41 2007
From: Kelly.Ohair at Sun.COM (Kelly O'Hair)
Date: Thu, 09 Aug 2007 08:16:41 -0700
Subject: Where have the tests gone???
In-Reply-To: <1ED26BBA-C48E-4CFF-A59C-D71028A6571B@sun.com>
References: <d65861880708030657q6019043i385b4609839671e6@mail.gmail.com>
	<1ED26BBA-C48E-4CFF-A59C-D71028A6571B@sun.com>
Message-ID: <46BB2FD9.7010908@sun.com>

This was a mistake. In the process of optimizing our automated build system,
the test directory was skipped for the build process, but that unfortunately
removed it from the source bundles. This should be corrected soon.

We probably need some kind of audit mechanism in place...
Once we have the Mercurial repositories available, this won't be an issue.

-kto

Jonathan Gibbons wrote:
> I can't speak for what happened to all the tests, which surely sounds 
> like an accident,
> but I can say that that a bunch more tool tests will be made available 
> soon.
> 
> -- Jon G
> 
> 
> On Aug 3, 2007, at 6:57 AM, Volker Simonis wrote:
> 
>> Hi,
>>
>> I was hoping that while time goes by more and more tests will appeare
>> (espacially in the hotspot directory)  but now I just realized that
>> the 'test/' subdirectories from the 'hotspot/' and 'j2se/' directories
>> dissapeared completely in Build b17. Has this happend intentionally?
>>
>> What are the futuer plans for the jtreg tests in OpenJDK?
>>
>> By the way, the build itself succedded without any problems on Suse
>> Linux 10 on x86. Good job!
>>
>> Volker
> 


From Kelly.Ohair at Sun.COM  Thu Aug  9 17:22:17 2007
From: Kelly.Ohair at Sun.COM (Kelly O'Hair)
Date: Thu, 09 Aug 2007 10:22:17 -0700
Subject: encumbrances update
In-Reply-To: <Pine.WNT.4.64.0708021201080.2848@SFENG02.RF.LAN>
References: <46AFAC07.6060608@sun.com>
	<Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>
	<46B1833C.7050705@sun.com>
	<Pine.WNT.4.63.0708020012310.2576@tbf> <46B1A0BB.4030405@sun.com>
	<Pine.WNT.4.64.0708021201080.2848@SFENG02.RF.LAN>
Message-ID: <46BB4D49.9080900@sun.com>

I thought it was clear from the beginning, perhaps I wasn't making this
as well known as I should. If so, my aplogies.

It was never expected that the initial OpenJDK source drops would be buildable
on Windows. We focused on Linux and OpenSolaris and purposely left out
Windows due to the t2k issue, and knowing that the team working on this
was very close to removing it as a dependence, we felt this was justfified.
So that fact that it has never built on Windows was well known, I thought.

While I was trying to make t2k.lib available in the binary plugs, another team
was busy trying to remove it completely. So you can't say we haven't been trying
to deal with this issue.

It is pretty impossible to guarantee builds will be successful in all
possible configurations, but once we get past the basic issues, we will
do regular test builds of OpenJDK on Windows. But speaking from years of
experience, Windows is a difficult platform to get consistent and reliable.
Just because we can build on Windows isn't worth as much as saying it for
Linux or OpenSolaris.

-kto

Dan Fabulich wrote:
> Igor Nekrestyanov wrote:
> 
>> Of course i've tested these changes on all platforms including both 32 
>> and 64 bit Windows. It is "believed to build and work" because we had 
>> not performed full testing of openjdk binaries and we know that build 
>> can be fragile due to different build environments.
> 
> I'm sure you've tested your changes in your build, and maybe that's all 
> I can ask for, but I'm pretty sure the same could have been said of all 
> of the previous builds, including the code drop we got in May and every 
> build since, all of which, we know, don't work for anyone who isn't 
> @sun.com (due to at least one missing file).
> 
>> My tests are not "ideal" for number of reasons:
>> 2) I was using binary plugs created from my personal workspace.
>>      I believe they should be the same as those to be published with 
>> b17 code drop but this is my assumption.
> 
> We know for certain that this assumption is FALSE.  No binary plug given 
> to the public has ever built successfully on Windows; since your 
> personal plugs have been working for you for months, we can conclude 
> that there's something critically different between your plugs and the 
> plugs we get.
> 
> But who knows?  Maybe we'll get lucky this time.  :-)
> 
>> BTW, for windows build you will need to build freetype.dll (i tried 
>> only dll built with visualc). I believe that freetype build system 
>> does not generate .dll on windows and therefore some manual tweaking 
>> of freetype makefiles (and sources) might be necessary.
> 
> Sounds like fun! :-)  Will the documentation include a patch?  Or just 
> some tips on how to make a .DLL?  What source files needed to change?
> 
> -Dan


From alexanderschunk at t-online.de  Thu Aug  9 18:28:00 2007
From: alexanderschunk at t-online.de (Alexander Schunk)
Date: 09 Aug 2007 18:28 GMT
Subject: IRC chat about the OpenJDK TCK @ 15h UTC
In-Reply-To: <46BA9499.1030706@sun.com>
References: <46BA9499.1030706@sun.com>
Message-ID: <1IJCkY-1M6UQi0@fwd28.aul.t-online.de>

Hi,
what do i need to provide for as url when using the stand alone
jIRCII client?
Regards
Alex
 "Tom Marble" <mailto:Tom.Marble at Sun.COM> schrieb:
> All:
> 
> Sun is announcing today the availability of a new license that
> will fulfill our promise to the Free software world to make it
> possible to test OpenJDK-based implementations for compatibility.
> OpenJDK implementations that pass the Java SE 6 TCK will be eligible
> for the license for the "Java Compatible" word mark and logo.
> 
> With this step, Free AND compatible implementations become possible.
> We're hoping this will make it a lot easier for you to know if your
> contributions break compatibility or not, and for Free software
> implementations to thus become part of the Java platform's
> "Write Once, Run Anywhere" promise.
> 
> You can review the text of the license [1] and read the
> updated FAQ [2] about it.
> 
> Join us for a community chat to talk about the OpenJDK TCK license,
> answer  your questions, and think about the implications and
> next steps on the #openjdk channel [3] at
> 15:00 UTC (= 8:00 am PDT = 11:00 am EDT).
> 
> Regards,
> 
> --Tom
> 
> [1] http://openjdk.java.net/legal/openjdk-tck-license.pdf
> [2] http://www.sun.com/software/opensource/java/faq.jsp#k
> [3] http://openjdk.java.net/irc/   #openjdk
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070809/527f50f5/attachment.html>

From alexanderschunk at t-online.de  Thu Aug  9 19:24:00 2007
From: alexanderschunk at t-online.de (Alexander Schunk)
Date: 09 Aug 2007 19:24 GMT
Subject: IRC chat about the OpenJDK TCK @ 15h UTC
In-Reply-To: <46BA9499.1030706@sun.com>
References: <46BA9499.1030706@sun.com>
Message-ID: <1IJDcd-0sy89o0@fwd34.aul.t-online.de>

Hi,
from a developers point of view i think this approach makes sense because it ensures Java code stability and also comformity since anyone contributing to this project has to comply with the standars used by SUN.
This avoids "loose" integration of unstable software and should make integrateion process also more transparent for the public and easier for the SUN engineers. 
>From a Users point of view i also appreciate it becuase i know that software that is contributed to the JDK does match the standards used by SUN and no other standards that might be out there.
This license should be used on all OpenJDK projects so all parts of OpenJKD comply to this license too.
Regards
Alex
"Tom Marble" <mailto:Tom.Marble at Sun.COM> schrieb:
> All:
> 
> Sun is announcing today the availability of a new license that
> will fulfill our promise to the Free software world to make it
> possible to test OpenJDK-based implementations for compatibility.
> OpenJDK implementations that pass the Java SE 6 TCK will be eligible
> for the license for the "Java Compatible" word mark and logo.
> 
> With this step, Free AND compatible implementations become possible.
> We're hoping this will make it a lot easier for you to know if your
> contributions break compatibility or not, and for Free software
> implementations to thus become part of the Java platform's
> "Write Once, Run Anywhere" promise.
> 
> You can review the text of the license [1] and read the
> updated FAQ [2] about it.
> 
> Join us for a community chat to talk about the OpenJDK TCK license,
> answer  your questions, and think about the implications and
> next steps on the #openjdk channel [3] at
> 15:00 UTC (= 8:00 am PDT = 11:00 am EDT).
> 
> Regards,
> 
> --Tom
> 
> [1] http://openjdk.java.net/legal/openjdk-tck-license.pdf
> [2] http://www.sun.com/software/opensource/java/faq.jsp#k
> [3] http://openjdk.java.net/irc/   #openjdk
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070809/be2b14f7/attachment.html>

From Tom.Marble at Sun.COM  Thu Aug  9 20:40:55 2007
From: Tom.Marble at Sun.COM (Tom Marble)
Date: Thu, 09 Aug 2007 15:40:55 -0500
Subject: IRC chat about the OpenJDK TCK @ 15h UTC
In-Reply-To: <1IJDcd-0sy89o0@fwd34.aul.t-online.de>
References: <46BA9499.1030706@sun.com> <1IJDcd-0sy89o0@fwd34.aul.t-online.de>
Message-ID: <46BB7BD7.9020001@sun.com>

Alexander Schunk wrote:
> what do i need to provide for as url when using the stand alone
> jIRCII client?
It used to be that `butane had a nice cgi script that would allow
one to construct such a URL to launch Java Web Start.  Alas, apparently
the website had somewhat of a crash:
http://jircii.hick.org/

I considered, on one point, hosting that script, but opted instead
for the JNLP it created:
http://openjdk.java.net/irc/openjdk.jnlp

Of course there are many, many other IRC clients to choose from.

> From a Users point of view i also appreciate it becuase i know that
> software that is contributed to the JDK does match the standards used by
> SUN and no other standards that might be out there.

Recall that the standard is set by the Java Community Process, which
includes many other voices than just Sun.

Regards,

--Tom


From dan at fabulich.com  Thu Aug  9 20:42:33 2007
From: dan at fabulich.com (Dan Fabulich)
Date: Thu, 9 Aug 2007 13:42:33 -0700 (Pacific Daylight Time)
Subject: encumbrances update
In-Reply-To: <46BB4D49.9080900@sun.com>
References: <46AFAC07.6060608@sun.com>
	<Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>
	<46B1833C.7050705@sun.com> <Pine.WNT.4.63.0708020012310.2576@tbf>
	<46B1A0BB.4030405@sun.com>
	<Pine.WNT.4.64.0708021201080.2848@SFENG02.RF.LAN>
	<46BB4D49.9080900@sun.com>
Message-ID: <Pine.WNT.4.64.0708091338200.4600@SFENG02.RF.LAN>

Kelly O'Hair wrote:

> It was never expected that the initial OpenJDK source drops would be 
> buildable on Windows.
[...]
> So that fact that it has never built on Windows was well known, I thought.

Actually, that comes as a (small) surprise to me.  When I asked about 
Windows build problems at Mark's OpenJDK talk at JavaOne, I asked him if 
he believed it should work; he said yes.  I then followed up with a 
question like: "Would you believe that it doesn't work?" he again replied 
yes.  I showed him the failure, to which I believe he replied: "Bummer." 
:-)

Certainly there's no sign on the website or in the README that Windows is 
*expected* not to work; instead, what you find are build instructions, 
with warnings that Windows is a difficult environment requiring a gentle 
touch, etc.

Especially in light of earlier e-mails from you back in June suggesting 
that build 15 would probably work, I frankly believed all this time that 
you guys just had no free Windows testing resources, and so nobody had 
ever tried a realistic clean OpenJDK build on Windows using the same plugs 
that we get.

> While I was trying to make t2k.lib available in the binary plugs, 
> another team was busy trying to remove it completely. So you can't say 
> we haven't been trying to deal with this issue.

I would never have said that.  I sincerely appreciate all of the work that 
you and other engineers at Sun have done on this, and look forward to the 
glorious day when one of us can get this thing to build! :-)

-Dan


From Kelly.Ohair at Sun.COM  Thu Aug  9 22:05:57 2007
From: Kelly.Ohair at Sun.COM (Kelly O'Hair)
Date: Thu, 09 Aug 2007 15:05:57 -0700
Subject: encumbrances update
In-Reply-To: <Pine.WNT.4.64.0708091338200.4600@SFENG02.RF.LAN>
References: <46AFAC07.6060608@sun.com>
	<Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>
	<46B1833C.7050705@sun.com>
	<Pine.WNT.4.63.0708020012310.2576@tbf> <46B1A0BB.4030405@sun.com>
	<Pine.WNT.4.64.0708021201080.2848@SFENG02.RF.LAN>
	<46BB4D49.9080900@sun.com>
	<Pine.WNT.4.64.0708091338200.4600@SFENG02.RF.LAN>
Message-ID: <46BB8FC5.5050308@sun.com>



Dan Fabulich wrote:
> Kelly O'Hair wrote:
> 
>> It was never expected that the initial OpenJDK source drops would be 
>> buildable on Windows.
> [...]
>> So that fact that it has never built on Windows was well known, I 
>> thought.
> 
> Actually, that comes as a (small) surprise to me.  When I asked about 
> Windows build problems at Mark's OpenJDK talk at JavaOne, I asked him if 
> he believed it should work; he said yes.  I then followed up with a 
> question like: "Would you believe that it doesn't work?" he again 
> replied yes.  I showed him the failure, to which I believe he replied: 
> "Bummer." :-)
> 

Many of us had hopes it would work in the end, I was hopeful til the end too.
Ted obviously figured out how to create a t2k.lib from a t2k.dll, so
he got pretty close...  Not that we expected anyone to try that hard. ;^)

> Certainly there's no sign on the website or in the README that Windows 
> is *expected* not to work; instead, what you find are build 
> instructions, with warnings that Windows is a difficult environment 
> requiring a gentle touch, etc.

I suspect that many of us at the lower levels were surprised that the
windows binary plugs were even made available.
But as to the README, I was responsible for that, and I was thinking more
'long term' when I kept the Windows information in it, some people didn't
want it there until we delivered windows buildable bundles. I was probably
wrong to include it, just confused things.

> 
> Especially in light of earlier e-mails from you back in June suggesting 
> that build 15 would probably work, I frankly believed all this time that 
> you guys just had no free Windows testing resources, and so nobody had 
> ever tried a realistic clean OpenJDK build on Windows using the same 
> plugs that we get.
> 

A fairly valid analysis. We had high hopes, but our delivery was poor.
You are getting pretty raw sources when it comes to windows right now.

>> While I was trying to make t2k.lib available in the binary plugs, 
>> another team was busy trying to remove it completely. So you can't say 
>> we haven't been trying to deal with this issue.
> 
> I would never have said that.  I sincerely appreciate all of the work 
> that you and other engineers at Sun have done on this, and look forward 
> to the glorious day when one of us can get this thing to build! :-)

I look forward to it too, but in the meantime I need to get going on the
Mercurial repositories.

---

We have openjdk builds, and product jdk builds, and part of our internal issue here
is that it isn't easy to make sure both are building, so I'll be looking at
how we can get some/many of the developers inside Sun using a make target that does
both builds, chained together so to speak. Stay tuned on this...

As always, it will take some time for all this to get better, and hopefully
it's more steps forward than backward.

-kto

> 
> -Dan


From dan at fabulich.com  Fri Aug 10 00:32:46 2007
From: dan at fabulich.com (Dan Fabulich)
Date: Thu, 9 Aug 2007 17:32:46 -0700 (Pacific Daylight Time)
Subject: ted's build on windows succeeded!
In-Reply-To: <015701c7dae5$740b44d0$5c21ce70$@com>
References: <46AFAC07.6060608@sun.com>
	<Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>
	<46B1833C.7050705@sun.com> <Pine.WNT.4.63.0708020012310.2576@tbf>
	<46B1A0BB.4030405@sun.com>
	<Pine.WNT.4.64.0708021201080.2848@SFENG02.RF.LAN>
	<46BB4D49.9080900@sun.com> <015701c7dae5$740b44d0$5c21ce70$@com>
Message-ID: <Pine.WNT.4.64.0708091731470.4600@SFENG02.RF.LAN>

Ted Neward wrote:

> (5) It's kinda all moot--I did a fresh re-fetch from SVN for the b17 drop,
> used Ivan's build of the freetype DLL, and it all builds, both debug and
> fastdebug.
>
> So yes, Dan, somebody outside of Sun managed to make it work on Windows. :-)

Congratulations!  Now I'm going to try it again. ;-)

-Dan


From Kelly.Ohair at Sun.COM  Fri Aug 10 01:02:49 2007
From: Kelly.Ohair at Sun.COM (Kelly O'Hair)
Date: Thu, 09 Aug 2007 18:02:49 -0700
Subject: encumbrances update
In-Reply-To: <015701c7dae5$740b44d0$5c21ce70$@com>
References: <46AFAC07.6060608@sun.com>
	<Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>
	<46B1833C.7050705@sun.com>
	<Pine.WNT.4.63.0708020012310.2576@tbf> <46B1A0BB.4030405@sun.com>
	<Pine.WNT.4.64.0708021201080.2848@SFENG02.RF.LAN>
	<46BB4D49.9080900@sun.com> <015701c7dae5$740b44d0$5c21ce70$@com>
Message-ID: <46BBB939.9000305@sun.com>



Ted Neward wrote:
> Well... several thoughts come to mind:
> 
> (1) The fact that OpenJDK was not aiming to be buildable on Windows from the
> beginning was definitely NOT clear, at least not to me.
> (2) The fact that it would be viewed as acceptable to release the source in
> an unbuildable form on any of the "supported" platforms--Windows, Linux,
> Solaris--surprises me. I accepted the idea that it was unbuildable as I
> thought it was a temporary break, to be fixed "real soon now", not an
> acceptable state of affairs.
> (3) The idea that Windows is not as important as Linux or Solaris is a
> dangerous idea, IMHO. It essentially suggests that Sun is throwing away a
> whole legion of developers who work on Windows and would want to contribute
> patches and suggestions for improving the Java-Windows experience. IMHO, the
> success of .NET due in no small part to the fairly substantive gap between
> Java and Windows.
> (4) I don't want to suggest in any way that I'm unhappy with the progress
> made by you (Kelly) or any of the other build team developers--having
> wrestled with other build systems in the past, I'm stunned at the size of
> building the JVM, and awed at the fact that it works at all, much less on
> systems that aren't tightly locked down in terms of tools and filesystem
> layout. (I worked at Intuit for a while, on Quicken 5, and there you got to
> choose what drive letter the code would be installed on--several guys had
> external SCSI drives they took with them to work from home--but beyond that,
> everything had to look "just like this".) I said it publicly on my blog, I
> said it to Simon Phipps during an interview with him at OSCon (which is
> going up on iTunes), I'll say it again here: you guys are doing a DAMN
> impressive job.
> (5) It's kinda all moot--I did a fresh re-fetch from SVN for the b17 drop,
> used Ivan's build of the freetype DLL, and it all builds, both debug and
> fastdebug.

Hurrah!

My sincere apologies for items 1-3 above. I'll give myself 20 lashes. :^(

-kto

> 
> So yes, Dan, somebody outside of Sun managed to make it work on Windows. :-)
> 
> Ted Neward
> Java, .NET, XML Services
> Consulting, Teaching, Speaking, Writing
> http://www.tedneward.com
>  
> 
>> -----Original Message-----
>> From: build-dev-bounces at openjdk.java.net [mailto:build-dev-
>> bounces at openjdk.java.net] On Behalf Of Kelly O'Hair
>> Sent: Thursday, August 09, 2007 10:22 AM
>> To: Dan Fabulich
>> Cc: discuss at openjdk.java.net; Anthony Petrov; Phil Race; Igor
>> Nekrestyanov; build-dev at openjdk.java.net
>> Subject: Re: encumbrances update
>>
>> I thought it was clear from the beginning, perhaps I wasn't making this
>> as well known as I should. If so, my aplogies.
>>
>> It was never expected that the initial OpenJDK source drops would be
>> buildable
>> on Windows. We focused on Linux and OpenSolaris and purposely left out
>> Windows due to the t2k issue, and knowing that the team working on this
>> was very close to removing it as a dependence, we felt this was
>> justfified.
>> So that fact that it has never built on Windows was well known, I
>> thought.
>>
>> While I was trying to make t2k.lib available in the binary plugs,
>> another team
>> was busy trying to remove it completely. So you can't say we haven't
>> been trying
>> to deal with this issue.
>>
>> It is pretty impossible to guarantee builds will be successful in all
>> possible configurations, but once we get past the basic issues, we will
>> do regular test builds of OpenJDK on Windows. But speaking from years
>> of
>> experience, Windows is a difficult platform to get consistent and
>> reliable.
>> Just because we can build on Windows isn't worth as much as saying it
>> for
>> Linux or OpenSolaris.
>>
>> -kto
>>
>> Dan Fabulich wrote:
>>> Igor Nekrestyanov wrote:
>>>
>>>> Of course i've tested these changes on all platforms including both
>> 32
>>>> and 64 bit Windows. It is "believed to build and work" because we
>> had
>>>> not performed full testing of openjdk binaries and we know that
>> build
>>>> can be fragile due to different build environments.
>>> I'm sure you've tested your changes in your build, and maybe that's
>> all
>>> I can ask for, but I'm pretty sure the same could have been said of
>> all
>>> of the previous builds, including the code drop we got in May and
>> every
>>> build since, all of which, we know, don't work for anyone who isn't
>>> @sun.com (due to at least one missing file).
>>>
>>>> My tests are not "ideal" for number of reasons:
>>>> 2) I was using binary plugs created from my personal workspace.
>>>>      I believe they should be the same as those to be published with
>>>> b17 code drop but this is my assumption.
>>> We know for certain that this assumption is FALSE.  No binary plug
>> given
>>> to the public has ever built successfully on Windows; since your
>>> personal plugs have been working for you for months, we can conclude
>>> that there's something critically different between your plugs and
>> the
>>> plugs we get.
>>>
>>> But who knows?  Maybe we'll get lucky this time.  :-)
>>>
>>>> BTW, for windows build you will need to build freetype.dll (i tried
>>>> only dll built with visualc). I believe that freetype build system
>>>> does not generate .dll on windows and therefore some manual tweaking
>>>> of freetype makefiles (and sources) might be necessary.
>>> Sounds like fun! :-)  Will the documentation include a patch?  Or
>> just
>>> some tips on how to make a .DLL?  What source files needed to change?
>>>
>>> -Dan
>> No virus found in this incoming message.
>> Checked by AVG Free Edition.
>> Version: 7.5.476 / Virus Database: 269.11.11/944 - Release Date:
>> 8/9/2007 2:44 PM
>>
> 
> No virus found in this outgoing message.
> Checked by AVG Free Edition. 
> Version: 7.5.476 / Virus Database: 269.11.11/944 - Release Date: 8/9/2007
> 2:44 PM
>  
> 


From dan at fabulich.com  Mon Aug 13 00:55:41 2007
From: dan at fabulich.com (Dan Fabulich)
Date: Sun, 12 Aug 2007 17:55:41 -0700 (Pacific Daylight Time)
Subject: dan's build on windows succeeded too!
In-Reply-To: <Pine.WNT.4.64.0708091731470.4600@SFENG02.RF.LAN>
References: <46AFAC07.6060608@sun.com>
	<Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>
	<46B1833C.7050705@sun.com> <Pine.WNT.4.63.0708020012310.2576@tbf>
	<46B1A0BB.4030405@sun.com>
	<Pine.WNT.4.64.0708021201080.2848@SFENG02.RF.LAN>
	<46BB4D49.9080900@sun.com> <015701c7dae5$740b44d0$5c21ce70$@com>
	<Pine.WNT.4.64.0708091731470.4600@SFENG02.RF.LAN>
Message-ID: <Pine.WNT.4.63.0708121732470.3900@tbf>


I'm pleased to announce that I've also gotten build 17 to succeed on 
Windows.

Notably, I used a copied-and-pasted version of Igor's buildenv.bat file, 
which I then tweaked with my own paths.  Here's the tweaking I needed to 
do:

1) Path ordering confused localegen: Using Igor's env vars, I reproduced 
the '\r' problem in LocaleMetaDataInfo.java that Igor reported earlier. 
Unfortunately, at least in my case, changing the awk command line did not 
work around the problem.

Digging just a bit deeper, I found that localegen.sh was naively invoking 
"sort" right off the path, a version of which comes with Windows as 
c:\windows\system32\sort.exe.  Since Igor's PATH called for using 
c:\windows\system32 prior to cygwin\bin, Windows' sort.exe was used, which 
inserted Windows-ish line endings.  Changing the PATH to use cygwin\bin 
first fixed the problem.

It would be wise to add a sanity check to make sure that "sort" is not 
Windows' sort.exe.

2) User name contained spaces: I also found that the scripts like to 
insert the current user's name into various version strings; this doesn't 
work when one's user name has a space in it.  (My default user name on 
Windows XP is "Dan Fabulich.")  To remedy this, I set my USERNAME 
environment variable to just be "dan", and made a symlink from /home/dan 
to /home/Dan\ Fabulich.

It would be wise to add a sanity check to make sure that $USERNAME 
contains no spaces.

3) Unicows.lib: Igor linked to a place where you can download unicows, and 
his buildenv.bat indicated that unicows.dll and unicows.lib would be in 
the same location.
http://www.microsoft.com/downloads/details.aspx?FamilyId=73BA7BD7-ED06-4F0D-80A4-2A7EEAEE17E2&displaylang=en

I found that when I downloaded unicows.exe and extracted it, I got 
unicows.dll and .pdb, but no .lib.  As far as I know, unicows.lib is NOT 
available on the web, but is only available as part of Visual Studio.

In summary, here's the buildenv.bat that worked for me:

******************
title openjdk
set PATH=c:\cygwin\bin;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem
set LD_LIBRARY_PATH=
set CLASSPATH=
set JAVA_HOME=

set ALT_BOOTDIR=C:/devtools/jdk1.6.0_00
set ALT_OUTPUTDIR=c:/devtools/openjdk-build
set ALT_BINARY_PLUGS_PATH=C:/devtools/openjdk-binary-plugs-b17

rem not mentioned in the README-builds.html? (place hotspot VM comes from) 
rem TODO: try to point to openjdk vm
set ALT_JDK_IMPORT_PATH=C:/devtools/jdk1.6.0_00

set ALT_DXSDK_PATH=C:/PROGRA~1/MICROS~1.0SD

set VC=C:/PROGRA~1/MICROS~2.NET
set ALT_COMPILER_PATH=%VC%/VC7/Bin
set ALT_MSDEVTOOLS_PATH=%VC%/Common7/Tools/Bin
set ALT_MSVCR71_DLL_PATH=%VC%/SDK/v1.1/Bin

rem using Igor's build of Freetype
set ALT_FREETYPE_LIB_PATH=C:/devtools/freetype/freetype-i586/lib
set ALT_FREETYPE_HEADERS_PATH=C:/devtools/freetype/freetype-i586/include

rem NB: not documented? will be removed in b18?
rem http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6332635
rem unicows.dll is available on the web. 
rem E.g. http://tinyurl.com/qynq
set ALT_UNICOWS_LIB_PATH=%VC%/Vc7/PLATFO~1/Lib
set ALT_UNICOWS_DLL_PATH=c:/devtools/unicows

rem bug in the makefiles: should be set prior to freetypecheck. 
rem TODO: file it
set OPENJDK=true

set HOME=/home/dan
set USERNAME=dan

call %vc%\Common7\Tools\vsvars32.bat

bash
****************

Running this script and executing the build from within bash, I did "cd 
control/make" and ran "make sanity".  It succeeded.  Then I ran "make"; a 
while later, I got a working JDK.

-Dan


From Igor.Nekrestyanov at Sun.COM  Mon Aug 13 06:40:26 2007
From: Igor.Nekrestyanov at Sun.COM (Igor Nekrestyanov)
Date: Mon, 13 Aug 2007 10:40:26 +0400
Subject: dan's build on windows succeeded too!
In-Reply-To: <Pine.WNT.4.63.0708121732470.3900@tbf>
References: <46AFAC07.6060608@sun.com>
	<Pine.WNT.4.64.0708011154040.2848@SFENG02.RF.LAN>
	<46B1833C.7050705@sun.com>
	<Pine.WNT.4.63.0708020012310.2576@tbf> <46B1A0BB.4030405@sun.com>
	<Pine.WNT.4.64.0708021201080.2848@SFENG02.RF.LAN>
	<46BB4D49.9080900@sun.com> <015701c7dae5$740b44d0$5c21ce70$@com>
	<Pine.WNT.4.64.0708091731470.4600@SFENG02.RF.LAN>
	<Pine.WNT.4.63.0708121732470.3900@tbf>
Message-ID: <46BFFCDA.6030309@sun.com>

Congratulations!

Thanks for providing additional details. We will use it to improve build.
Do you remember particular details on build failure if user name has 
spaces (which scripts failed and what was error message)?

BTW, it seems that dependency on unicows.dll/lib will be gone in b18.
Also, setting ALT_JDK_IMPORT_PATH is only necessary for partial builds.
For full builds (i.e. make from control/make) it is not required.

-igor

Dan Fabulich wrote:
>
> I'm pleased to announce that I've also gotten build 17 to succeed on 
> Windows.
>
> Notably, I used a copied-and-pasted version of Igor's buildenv.bat 
> file, which I then tweaked with my own paths.  Here's the tweaking I 
> needed to do:
>
> 1) Path ordering confused localegen: Using Igor's env vars, I 
> reproduced the '\r' problem in LocaleMetaDataInfo.java that Igor 
> reported earlier. Unfortunately, at least in my case, changing the awk 
> command line did not work around the problem.
>
> Digging just a bit deeper, I found that localegen.sh was naively 
> invoking "sort" right off the path, a version of which comes with 
> Windows as c:\windows\system32\sort.exe.  Since Igor's PATH called for 
> using c:\windows\system32 prior to cygwin\bin, Windows' sort.exe was 
> used, which inserted Windows-ish line endings.  Changing the PATH to 
> use cygwin\bin first fixed the problem.
>
> It would be wise to add a sanity check to make sure that "sort" is not 
> Windows' sort.exe.
>
> 2) User name contained spaces: I also found that the scripts like to 
> insert the current user's name into various version strings; this 
> doesn't work when one's user name has a space in it.  (My default user 
> name on Windows XP is "Dan Fabulich.")  To remedy this, I set my 
> USERNAME environment variable to just be "dan", and made a symlink 
> from /home/dan to /home/Dan\ Fabulich.
>
> It would be wise to add a sanity check to make sure that $USERNAME 
> contains no spaces.
>
> 3) Unicows.lib: Igor linked to a place where you can download unicows, 
> and his buildenv.bat indicated that unicows.dll and unicows.lib would 
> be in the same location.
> http://www.microsoft.com/downloads/details.aspx?FamilyId=73BA7BD7-ED06-4F0D-80A4-2A7EEAEE17E2&displaylang=en 
>
>
> I found that when I downloaded unicows.exe and extracted it, I got 
> unicows.dll and .pdb, but no .lib.  As far as I know, unicows.lib is 
> NOT available on the web, but is only available as part of Visual Studio.
>
> In summary, here's the buildenv.bat that worked for me:
>
> ******************
> title openjdk
> set 
> PATH=c:\cygwin\bin;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem 
>
> set LD_LIBRARY_PATH=
> set CLASSPATH=
> set JAVA_HOME=
>
> set ALT_BOOTDIR=C:/devtools/jdk1.6.0_00
> set ALT_OUTPUTDIR=c:/devtools/openjdk-build
> set ALT_BINARY_PLUGS_PATH=C:/devtools/openjdk-binary-plugs-b17
>
> rem not mentioned in the README-builds.html? (place hotspot VM comes 
> from) rem TODO: try to point to openjdk vm
> set ALT_JDK_IMPORT_PATH=C:/devtools/jdk1.6.0_00
>
> set ALT_DXSDK_PATH=C:/PROGRA~1/MICROS~1.0SD
>
> set VC=C:/PROGRA~1/MICROS~2.NET
> set ALT_COMPILER_PATH=%VC%/VC7/Bin
> set ALT_MSDEVTOOLS_PATH=%VC%/Common7/Tools/Bin
> set ALT_MSVCR71_DLL_PATH=%VC%/SDK/v1.1/Bin
>
> rem using Igor's build of Freetype
> set ALT_FREETYPE_LIB_PATH=C:/devtools/freetype/freetype-i586/lib
> set ALT_FREETYPE_HEADERS_PATH=C:/devtools/freetype/freetype-i586/include
>
> rem NB: not documented? will be removed in b18?
> rem http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6332635
> rem unicows.dll is available on the web. rem E.g. http://tinyurl.com/qynq
> set ALT_UNICOWS_LIB_PATH=%VC%/Vc7/PLATFO~1/Lib
> set ALT_UNICOWS_DLL_PATH=c:/devtools/unicows
>
> rem bug in the makefiles: should be set prior to freetypecheck. rem 
> TODO: file it
> set OPENJDK=true
>
> set HOME=/home/dan
> set USERNAME=dan
>
> call %vc%\Common7\Tools\vsvars32.bat
>
> bash
> ****************
>
> Running this script and executing the build from within bash, I did 
> "cd control/make" and ran "make sanity".  It succeeded.  Then I ran 
> "make"; a while later, I got a working JDK.
>
> -Dan



From aph at redhat.com  Mon Aug 13 14:35:35 2007
From: aph at redhat.com (Andrew Haley)
Date: Mon, 13 Aug 2007 15:35:35 +0100
Subject: OpenJDK TCK announcement
Message-ID: <18112.27703.551230.30450@zebedee.pink>

We at Red Hat are delighted to see that Sun will make good their
promise to make the Java TCK available to OpenJDK implementations.
This certainly takes us a step closer to fully free Java.

However, I have a couple of reservations.  Just to get those out of
the way first:

1.  Of course I'm disappointed that the TCK isn't going to be
available to all GPL'ed Java implementations, including those not
based on OpenJDK.  I had hoped to get the Java Compatible stamp of
approval for GCJ.

2.  More seriously, I'm concerned about how the confidentiality
clauses will work out in practice.  Hitherto we have shipped
testsuites with the packages that we're testing, so that everyone who
rebuilds a package can rerun the tests.  Also, we have never had the
convention in the free software world of holding our discussions on
private e-mail lists, and I doubt that it will work very well.  The
confidentiality clauses will make it difficult (or impossible) to
integrate TCK testing into the Fedora release process.

Nevertheless, as far as I'm aware there is nothing to stop us within
Red Hat from running the TCK on a Fedora OpenJDK package, and I expect
we will.  Other developers will doubtless do so as well.

I really hope that we will be able to run the TCK on the OpenJDK soon.
In fact I would have liked to start testing immediately!  Never mind:
IcedTea is a technology preview that's available today, and from our
own tests we know that it works very well.  For the time being we can
live with IcedTea not being officially Java[TM].

We would welcome early access to the Java 1.6 TCK so that we could do
some testing on what we have at the moment, even if we weren't able to
use it to claim Java compatibility.  Is a compromise possible here?

It's worth going over a little bit of history at this point: if the
entirety of OpenJDK had been free software we would almost certainly
have shipped it in Fedora (and, later, Red Hat Enterprise Linux) as it
was.  However, parts of it were not (and still are not) free software,
so we had to create the IcedTea package to fill the gaps.  As we've
said from the start, the purpose of IcedTea is to provide the
infrastructure for constructing a completely free implementation while
Sun continues to free more of the OpenJDK.

It has been somewhat frustrating that we haven't been able to work
more closely with Sun on ironing out these problems, but there are
still some legal issues to sort out, and opening up Sun's
well-established processes is doubtless a huge sink of time.

However I must point out that even given these problems we in the
free software community are in a far better position today than we
were with GCJ (and other free VMs) and GNU Classpath: with IcedTea
based on the OpenJDK code base we are much closer to Java
compatibility.

I'm very excited by the prospect of a 100% free and 100% compatible
Java, and I'd like to thank Sun for that.  But still, there is work to
be done.

Andrew.


From volker.simonis at gmail.com  Wed Aug 15 20:07:32 2007
From: volker.simonis at gmail.com (Volker Simonis)
Date: Wed, 15 Aug 2007 22:07:32 +0200
Subject: Search facility for the list archives?
Message-ID: <d65861880708151307u57f873c5wdffc552306e47772@mail.gmail.com>

Are there any plans for a search facility that can search all the
mailing lists from java.net. I meanwhile counted 33 different lists
and from the names it's not always clear where to look/post for a
specific topic.

I think providing a good search tool would greatly increase the
usefullness of the lists alltogether and would reduce the burdon of
double/cross posting.

I also remember that I've already posted to discuss at openjdk on the 1st
of March, however the current archives start from May only. Where have
the old postings gone too?

Regards,
Volker


From David.Herron at Sun.COM  Wed Aug 15 21:04:58 2007
From: David.Herron at Sun.COM (David Herron)
Date: Wed, 15 Aug 2007 14:04:58 -0700
Subject: Search facility for the list archives?
In-Reply-To: <d65861880708151307u57f873c5wdffc552306e47772@mail.gmail.com>
References: <d65861880708151307u57f873c5wdffc552306e47772@mail.gmail.com>
Message-ID: <46C36A7A.9010303@sun.com>

Volker Simonis wrote:
> Are there any plans for a search facility that can search all the
> mailing lists from java.net. I meanwhile counted 33 different lists
> and from the names it's not always clear where to look/post for a
> specific topic.
>
> I think providing a good search tool would greatly increase the
> usefullness of the lists alltogether and would reduce the burdon of
> double/cross posting.
>
> I also remember that I've already posted to discuss at openjdk on the 1st
> of March, however the current archives start from May only. Where have
> the old postings gone too?
>
> Regards,
> Volker
>   
Dunno about the old postings.. they were on differently implemented 
infrastructure.

It seems to me that going to google and entering:

    site:mail.openjdk.java.net search string

would be a very good way to handle this.  But there's a robot.txt saying 
otherwise...

     wget http://mail.openjdk.java.net/robots.txt

Yields:

    User-agent: *
    Disallow: /

There must be a reason for this to which I'm not privy.  To me it makes 
absolute perfect sense to index this stuff because it's the search 
engines that draw the most traffic to a web site.

- David Herron

 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070815/3d49d84d/attachment.html>

From volker.simonis at gmail.com  Wed Aug 15 22:41:40 2007
From: volker.simonis at gmail.com (Volker Simonis)
Date: Thu, 16 Aug 2007 00:41:40 +0200
Subject: Search facility for the list archives?
In-Reply-To: <46C36A7A.9010303@sun.com>
References: <d65861880708151307u57f873c5wdffc552306e47772@mail.gmail.com>
	<46C36A7A.9010303@sun.com>
Message-ID: <d65861880708151541iabdaf6axfcd77b5d1a1ae907@mail.gmail.com>

On 8/15/07, David Herron <David.Herron at sun.com> wrote:
>
>  Volker Simonis wrote:
>  Are there any plans for a search facility that can search all the
> mailing lists from java.net. I meanwhile counted 33 different lists
> and from the names it's not always clear where to look/post for a
> specific topic.
>
> I think providing a good search tool would greatly increase the
> usefullness of the lists alltogether and would reduce the burdon of
> double/cross posting.
>
> I also remember that I've already posted to discuss at openjdk on the 1st
> of March, however the current archives start from May only. Where have
> the old postings gone too?
>
> Regards,
> Volker
>
>  Dunno about the old postings.. they were on differently implemented
> infrastructure.
>
>  It seems to me that going to google and entering:
>
>      site:mail.openjdk.java.net search string
>
>  would be a very good way to handle this.  But there's a robot.txt saying
> otherwise...
>
>  wget http://mail.openjdk.java.net/robots.txt
>  Yields:
>
> User-agent: *
>  Disallow: /
>  There must be a reason for this to which I'm not privy.  To me it makes
> absolute perfect sense to index this stuff because it's the search engines
> that draw the most traffic to a web site.
>
>  - David Herron
>

Hopefully, it is not because the responsible admins are still not sure
if the current infrastructure is the final one :)


From volker.simonis at gmail.com  Wed Aug 15 23:04:05 2007
From: volker.simonis at gmail.com (Volker Simonis)
Date: Thu, 16 Aug 2007 01:04:05 +0200
Subject: Will OpenJDK never contain other GPLed code?
Message-ID: <d65861880708151604l7d580dfere819d98158bb310b@mail.gmail.com>

During the discussion on a different thread on this list regarding the
disassembler library for the HotSpot I realized, that although the
OpenJDK is released under the GPL, it is currently impossible to
contribute or integrate other source code, that is also licensed under
the GPL, into OpenJDK, because it would be probably impossible to get
a Sun Contributor Agreement (SCA) for such code. And SUN will probably
refuse to integrate code into the JDK that can't be integrated into
the official SUN J2SE distribution which is not GPL.

In the OpenJDK FAQ (see
http://www.sun.com/software/opensource/java/faq.jsp#n3) SUN writes:
"..when we chose the GPL as the basis for Sun's open-source Java SE
implementations, we made it easy to combine the open-source JDK code
base with other GPL-licensed code bases such as GNU/Linux, GNU
Classpath, Kaffe, GNOME, and others".

That's true, however the other way round, combine GPL-licensed
software with OpenJDK becomes impossible because of the need of a SCA.
Therefore it becomes impossible for OpenJDK to profit from GPL
software. Notice that I'm speaking about the "official" OpenJDK here,
not any branches thereof (like for example IcedTea), that don't have
these problems.

Is this intended? Any comments?

Volker


From David.Herron at Sun.COM  Wed Aug 15 23:25:37 2007
From: David.Herron at Sun.COM (David Herron)
Date: Wed, 15 Aug 2007 16:25:37 -0700
Subject: Will OpenJDK never contain other GPLed code?
In-Reply-To: <d65861880708151604l7d580dfere819d98158bb310b@mail.gmail.com>
References: <d65861880708151604l7d580dfere819d98158bb310b@mail.gmail.com>
Message-ID: <46C38B71.9020503@sun.com>

Volker Simonis wrote:
> During the discussion on a different thread on this list regarding the
> disassembler library for the HotSpot I realized, that although the
> OpenJDK is released under the GPL, it is currently impossible to
> contribute or integrate other source code, that is also licensed under
> the GPL, into OpenJDK, because it would be probably impossible to get
> a Sun Contributor Agreement (SCA) for such code. And SUN will probably
> refuse to integrate code into the JDK that can't be integrated into
> the official SUN J2SE distribution which is not GPL.
>
> In the OpenJDK FAQ (see
> http://www.sun.com/software/opensource/java/faq.jsp#n3) SUN writes:
> "..when we chose the GPL as the basis for Sun's open-source Java SE
> implementations, we made it easy to combine the open-source JDK code
> base with other GPL-licensed code bases such as GNU/Linux, GNU
> Classpath, Kaffe, GNOME, and others".
>
> That's true, however the other way round, combine GPL-licensed
> software with OpenJDK becomes impossible because of the need of a SCA.
> Therefore it becomes impossible for OpenJDK to profit from GPL
> software. Notice that I'm speaking about the "official" OpenJDK here,
> not any branches thereof (like for example IcedTea), that don't have
> these problems.
>
> Is this intended? Any comments?
>
> Volker
>   

Does the SCA FAQ answer any questions?
    http://www.sun.com/software/opensource/contributor_agreement.jsp

In particular the SCA has the contributor assign joint rights so that 
both Sun and the contributor own the contributed code. 

- David Herron




From Peter.Kessler at Sun.COM  Wed Aug 15 23:44:48 2007
From: Peter.Kessler at Sun.COM (Peter B. Kessler)
Date: Wed, 15 Aug 2007 16:44:48 -0700
Subject: Search facility for the list archives?
In-Reply-To: <d65861880708151307u57f873c5wdffc552306e47772@mail.gmail.com>
References: <d65861880708151307u57f873c5wdffc552306e47772@mail.gmail.com>
Message-ID: <46C38FF0.4090309@Sun.COM>

Volker Simonis wrote:

> Are there any plans for a search facility that can search all the
> mailing lists from java.net. I meanwhile counted 33 different lists
> and from the names it's not always clear where to look/post for a
> specific topic.
> 
> I think providing a good search tool would greatly increase the
> usefullness of the lists alltogether and would reduce the burdon of
> double/cross posting.
> 
> I also remember that I've already posted to discuss at openjdk on the 1st
> of March, however the current archives start from May only. Where have
> the old postings gone too?
> 
> Regards,
> Volker


I thought we were depending on

     http://gmane.org/

for searching.  Though, I don't see a way to search a group of
lists, e.g., at

     http://gmane.org/find.php?list=openjdk

but maybe I just don't know the syntax.  I also find that while
I can't find your message by searching for "search facility" on

     http://search.gmane.org/?query=%22search+facility%22&group=gmane.comp.java.openjdk.general

I can find your message by asking for all the topics from that
list and then choosing

     http://comments.gmane.org/gmane.comp.java.openjdk.general/383

Searching for "volker" turns up messages from 2007-03-01 through
2007-08-09, so maybe they are just behind in their indexing.

			... peter


From volker.simonis at gmail.com  Thu Aug 16 00:12:11 2007
From: volker.simonis at gmail.com (Volker Simonis)
Date: Thu, 16 Aug 2007 02:12:11 +0200
Subject: Search facility for the list archives?
In-Reply-To: <46C38FF0.4090309@Sun.COM>
References: <d65861880708151307u57f873c5wdffc552306e47772@mail.gmail.com>
	<46C38FF0.4090309@Sun.COM>
Message-ID: <d65861880708151712p4604622jd6f04e45c1430cff@mail.gmail.com>

It would be probably a good idea to put a link to
http://gmane.org/find.php?list=openjdk on the main mailing list page
at http://mail.openjdk.java.net/mailman/listinfo and probably also a
corresponding link on the page of each list (e.g.
http://mail.openjdk.java.net/mailman/listinfo/discuss) for people
(like me) that didn't knew gmane.

Maybe a small input form with a "Search at gmane" button below the "To
see the collection of prior postings to the list, visit the discuss
Archives"-line would be even more usefull?

Volker

On 8/16/07, Peter B. Kessler <Peter.Kessler at sun.com> wrote:
> Volker Simonis wrote:
>
> > Are there any plans for a search facility that can search all the
> > mailing lists from java.net. I meanwhile counted 33 different lists
> > and from the names it's not always clear where to look/post for a
> > specific topic.
> >
> > I think providing a good search tool would greatly increase the
> > usefullness of the lists alltogether and would reduce the burdon of
> > double/cross posting.
> >
> > I also remember that I've already posted to discuss at openjdk on the 1st
> > of March, however the current archives start from May only. Where have
> > the old postings gone too?
> >
> > Regards,
> > Volker
>
>
> I thought we were depending on
>
>      http://gmane.org/
>
> for searching.  Though, I don't see a way to search a group of
> lists, e.g., at
>
>      http://gmane.org/find.php?list=openjdk
>
> but maybe I just don't know the syntax.  I also find that while
> I can't find your message by searching for "search facility" on
>
>      http://search.gmane.org/?query=%22search+facility%22&group=gmane.comp.java.openjdk.general
>
> I can find your message by asking for all the topics from that
> list and then choosing
>
>      http://comments.gmane.org/gmane.comp.java.openjdk.general/383
>
> Searching for "volker" turns up messages from 2007-03-01 through
> 2007-08-09, so maybe they are just behind in their indexing.
>
>                         ... peter
>


From David.Herron at Sun.COM  Thu Aug 16 18:22:40 2007
From: David.Herron at Sun.COM (David Herron)
Date: Thu, 16 Aug 2007 11:22:40 -0700
Subject: Will OpenJDK never contain other GPLed code?
In-Reply-To: <d65861880708161037g1edc7a0fq4de8915e9f8c777a@mail.gmail.com>
References: <d65861880708151604l7d580dfere819d98158bb310b@mail.gmail.com>
	<46C38B71.9020503@sun.com>
	<d65861880708151701u4eb2b285sb4dea375caea0e40@mail.gmail.com>
	<46C46F84.7070907@sun.com>
	<d65861880708160933u721016aexff5fd271c4cbe6c1@mail.gmail.com>
	<d65861880708161037g1edc7a0fq4de8915e9f8c777a@mail.gmail.com>
Message-ID: <46C495F0.60504@sun.com>

Volker Simonis wrote:
> On 8/16/07, Volker Simonis <volker.simonis at gmail.com> wrote:
>   
>> On 8/16/07, David Herron <David.Herron at sun.com> wrote:
>>     
>>>
>>>  Volker Simonis wrote:
>>>  On 8/16/07, David Herron <David.Herron at sun.com> wrote:
>>>
>>>
>>>  Volker Simonis wrote:
>>>
>>>
>>>  During the discussion on a different thread on this list regarding the
>>> disassembler library for the HotSpot I realized, that although the
>>> OpenJDK is released under the GPL, it is currently impossible to
>>> contribute or integrate other source code, that is also licensed under
>>> the GPL, into OpenJDK, because it would be probably impossible to get
>>> a Sun Contributor Agreement (SCA) for such code. And SUN will probably
>>> refuse to integrate code into the JDK that can't be integrated into
>>> the official SUN J2SE distribution which is not GPL.
>>>
>>> In the OpenJDK FAQ (see
>>> http://www.sun.com/software/opensource/java/faq.jsp#n3) SUN
>>> writes:
>>> "..when we chose the GPL as the basis for Sun's open-source Java SE
>>> implementations, we made it easy to combine the open-source JDK code
>>> base with other GPL-licensed code bases such as GNU/Linux, GNU
>>> Classpath, Kaffe, GNOME, and others".
>>>
>>> That's true, however the other way round, combine GPL-licensed
>>> software with OpenJDK becomes impossible because of the need of a SCA.
>>> Therefore it becomes impossible for OpenJDK to profit from GPL
>>> software. Notice that I'm speaking about the "official" OpenJDK here,
>>> not any branches thereof (like for example IcedTea), that don't have
>>> these problems.
>>>
>>> Is this intended? Any comments?
>>>
>>> Volker
>>>
>>>
>>>  Does the SCA FAQ answer any questions?
>>> http://www.sun.com/software/opensource/contributor_agreement.jsp
>>>
>>> In particular the SCA has the contributor assign joint rights so that
>>> both Sun and the contributor own the contributed code.
>>>
>>> - David Herron
>>>
>>>
>>>  Yes, but this is exactly the problem. Let's say there's some very
>>> cool, GPL-licensed software that OpenJDK could benefit from (in the
>>> case I'm refering to this was binutils). You can't put it into
>>> OpenJDK, because SUN uses the OpenJDK sources to build its non-GPL JDK
>>> and you can't get SCAs from all original authors of the software in
>>> question.
>>>
>>> So OpenJDK, although licensed under GPL, is effectly still bound by
>>> the SUN JDK, which is not GPL licensed (at least with respect to
>>> already existing GPLed code)!
>>>
>>> Volker
>>>
>>>
>>>  Maybe this isn't spelled out carefully enough .. but the scenario you're
>>> describing is why the contributor agreement has the contributor agree to
>>> joint ownership of the code.  I've been told by the people who worked with
>>> the lawyers, that the joint ownership allows us to redistribute the code
>>> under different licenses.  We are following a dual license model, as you
>>> note, distributing code under both GPL and non-free licenses.
>>>
>>>  Before code can be contributed to any Sun project a contributor agreement
>>> is required from the contributor.  The scenario you're describing cannot
>>> happen.  Or, put another way, for some project with several owners, you're
>>> right that it would be very hard to track down all those owners and get them
>>> all to sign an SCA.  Hence it's a good thing that the SCA requires dual
>>> ownership of contributed code.  Without dual ownership, 10 years the Sun
>>> projects would be having a wide range of owners making it difficult to make
>>> any legal changes...
>>>
>>>  I found this page to have some similar considerations
>>>  http://www.gnu.org/licenses/why-assign.html
>>>
>>>  - David
>>>
>>>       
>> David,
>>
>> I understand what you want to say and I think I also understand SUNs
>> dual licence policy. After all its not a bad policy and I didn't
>> wanted to criticise it. I just wanted to point out the conclusion that
>> this dual-licence policy makes it impossible to contribute
>> 'single-licenced' GPL code to OpenJDK and consequently, OpenJDK will
>> always remain a SUN project (what's not  bad - again, no criticism
>> here). This point only has not been obvious to me...
>>
>> I think as of today, most of the available GPL code is single licenced
>> only under the GPL (making things kind of easy) but in recent times,
>> more and more companies begin to release their software under dual
>> licences (one open, not even necessarily GPL and a closed, commercial
>> licence). Combining code from such projects will only be possible in
>> the "open" part, and this will irrevocably result in project forks.
>> Hopefully, this won't happen to OpenJDK soon!
>>
>> Volker
>>
>>     
>
> David,
>
> I just realized that the last three posts on this topic havn't been
> cc'ed to the mailing list. At least from my part, this happend
> unintentionally. Would you mind if I will repost my last message that
> conttains all the other posts inlined to the list?
>
> As an addendum to our discussion, I also found the article "Does dual
> licensing threaten free software?" by Glyn Moody quite usefull and
> interesting (see http://linuxjournal.com/node/1000069)
>
> Regards,
> Volker
>   

Hi Volker, I do agree we should have had this discussion on the list.

Okay, hmm, it remains to be seen how successful we will be using the SCA 
to entice contributions.

But I don't understand the "single licensed" code.  I do think it 
depends on the individual contributor as to whether they will see fit to 
agree to the SCA.  The GPL doesn't, so far as I can tell, prohibit an 
individual project from dual licensing code.  If it did then Sun 
wouldn't be able to as we are doing with the OpenJDK because we're using 
both GPL and non-free licenses for the same code.

I'm sure there are other organizations using the same GPL/non-free dual 
license model.  er... MySQL for instance is available that way.

On my earlier reply to you I spent awhile looking at the MySQL website 
to see their contribution policy.  But they don't seem very open to 
outside contributions in the first place.. this page is where they list 
outside contributors, and it says at the front the MySQL AB owns all the 
copyrights, and then the type of contributions they list are usually not 
about the code to the database but ancilliary things.

    http://dev.mysql.com/doc/refman/5.0/en/contributors.html

That makes MySQL not a good organization to look to for a model of 
gaining contributions while dual licensing the code.


- David Herron

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070816/252dec96/attachment.html>

From volker.simonis at gmail.com  Thu Aug 16 18:24:01 2007
From: volker.simonis at gmail.com (Volker Simonis)
Date: Thu, 16 Aug 2007 20:24:01 +0200
Subject: Will OpenJDK never contain other GPLed code?
In-Reply-To: <46C495F0.60504@sun.com>
References: <d65861880708151604l7d580dfere819d98158bb310b@mail.gmail.com>
	<46C38B71.9020503@sun.com>
	<d65861880708151701u4eb2b285sb4dea375caea0e40@mail.gmail.com>
	<46C46F84.7070907@sun.com>
	<d65861880708160933u721016aexff5fd271c4cbe6c1@mail.gmail.com>
	<d65861880708161037g1edc7a0fq4de8915e9f8c777a@mail.gmail.com>
	<46C495F0.60504@sun.com>
Message-ID: <d65861880708161124n7d6195c9i822ffef85d11905e@mail.gmail.com>

On 8/16/07, David Herron <David.Herron at sun.com> wrote:
>
>  Volker Simonis wrote:
>  On 8/16/07, Volker Simonis <volker.simonis at gmail.com> wrote:
>
>
>  On 8/16/07, David Herron <David.Herron at sun.com> wrote:
>
>
>
>  Volker Simonis wrote:
>  On 8/16/07, David Herron <David.Herron at sun.com> wrote:
>
>
>  Volker Simonis wrote:
>
>
>  During the discussion on a different thread on this list regarding the
> disassembler library for the HotSpot I realized, that although the
> OpenJDK is released under the GPL, it is currently impossible to
> contribute or integrate other source code, that is also licensed under
> the GPL, into OpenJDK, because it would be probably impossible to get
> a Sun Contributor Agreement (SCA) for such code. And SUN will probably
> refuse to integrate code into the JDK that can't be integrated into
> the official SUN J2SE distribution which is not GPL.
>
> In the OpenJDK FAQ (see
> http://www.sun.com/software/opensource/java/faq.jsp#n3) SUN
> writes:
> "..when we chose the GPL as the basis for Sun's open-source Java SE
> implementations, we made it easy to combine the open-source JDK code
> base with other GPL-licensed code bases such as GNU/Linux, GNU
> Classpath, Kaffe, GNOME, and others".
>
> That's true, however the other way round, combine GPL-licensed
> software with OpenJDK becomes impossible because of the need of a SCA.
> Therefore it becomes impossible for OpenJDK to profit from GPL
> software. Notice that I'm speaking about the "official" OpenJDK here,
> not any branches thereof (like for example IcedTea), that don't have
> these problems.
>
> Is this intended? Any comments?
>
> Volker
>
>
>  Does the SCA FAQ answer any questions?
> http://www.sun.com/software/opensource/contributor_agreement.jsp
>
> In particular the SCA has the contributor assign joint rights so that
> both Sun and the contributor own the contributed code.
>
> - David Herron
>
>
>  Yes, but this is exactly the problem. Let's say there's some very
> cool, GPL-licensed software that OpenJDK could benefit from (in the
> case I'm refering to this was binutils). You can't put it into
> OpenJDK, because SUN uses the OpenJDK sources to build its non-GPL JDK
> and you can't get SCAs from all original authors of the software in
> question.
>
> So OpenJDK, although licensed under GPL, is effectly still bound by
> the SUN JDK, which is not GPL licensed (at least with respect to
> already existing GPLed code)!
>
> Volker
>
>
>  Maybe this isn't spelled out carefully enough .. but the scenario you're
> describing is why the contributor agreement has the contributor agree to
> joint ownership of the code. I've been told by the people who worked with
> the lawyers, that the joint ownership allows us to redistribute the code
> under different licenses. We are following a dual license model, as you
> note, distributing code under both GPL and non-free licenses.
>
>  Before code can be contributed to any Sun project a contributor agreement
> is required from the contributor. The scenario you're describing cannot
> happen. Or, put another way, for some project with several owners, you're
> right that it would be very hard to track down all those owners and get them
> all to sign an SCA. Hence it's a good thing that the SCA requires dual
> ownership of contributed code. Without dual ownership, 10 years the Sun
> projects would be having a wide range of owners making it difficult to make
> any legal changes...
>
>  I found this page to have some similar considerations
>  http://www.gnu.org/licenses/why-assign.html
>
>  - David
>
>
>  David,
>
> I understand what you want to say and I think I also understand SUNs
> dual licence policy. After all its not a bad policy and I didn't
> wanted to criticise it. I just wanted to point out the conclusion that
> this dual-licence policy makes it impossible to contribute
> 'single-licenced' GPL code to OpenJDK and consequently, OpenJDK will
> always remain a SUN project (what's not bad - again, no criticism
> here). This point only has not been obvious to me...
>
> I think as of today, most of the available GPL code is single licenced
> only under the GPL (making things kind of easy) but in recent times,
> more and more companies begin to release their software under dual
> licences (one open, not even necessarily GPL and a closed, commercial
> licence). Combining code from such projects will only be possible in
> the "open" part, and this will irrevocably result in project forks.
> Hopefully, this won't happen to OpenJDK soon!
>
> Volker
>
>
>  David,
>
> I just realized that the last three posts on this topic havn't been
> cc'ed to the mailing list. At least from my part, this happend
> unintentionally. Would you mind if I will repost my last message that
> conttains all the other posts inlined to the list?
>
> As an addendum to our discussion, I also found the article "Does dual
> licensing threaten free software?" by Glyn Moody quite usefull and
> interesting (see http://linuxjournal.com/node/1000069)
>
> Regards,
> Volker
>
>
>  Hi Volker, I do agree we should have had this discussion on the list.
>
>  Okay, hmm, it remains to be seen how successful we will be using the SCA to
> entice contributions.
>
>  But I don't understand the "single licensed" code.  I do think it depends
> on the individual contributor as to whether they will see fit to agree to
> the SCA.  The GPL doesn't, so far as I can tell, prohibit an individual
> project from dual licensing code.  If it did then Sun wouldn't be able to as
> we are doing with the OpenJDK because we're using both GPL and non-free
> licenses for the same code.
>
>  I'm sure there are other organizations using the same GPL/non-free dual
> license model.  er... MySQL for instance is available that way.
>
>  On my earlier reply to you I spent awhile looking at the MySQL website to
> see their contribution policy.  But they don't seem very open to outside
> contributions in the first place.. this page is where they list outside
> contributors, and it says at the front the MySQL AB owns all the copyrights,
> and then the type of contributions they list are usually not about the code
> to the database but ancilliary things.
>
>
> http://dev.mysql.com/doc/refman/5.0/en/contributors.html
>
>  That makes MySQL not a good organization to look to for a model of gaining
> contributions while dual licensing the code.
>
>
>  - David Herron
>

In theory you're right: "the GPL doesn't, so far as I can tell,
prohibit an individual project from dual licensing code". But in
reality, it will be impossible to get the permissions (i.e. the SCA)
of tens or even hundered of contributors of a GPL licensed project
(even if they would all be willing to sign it). Probably you won't
even be able to find them all. So in reality, you can only use it
under the GPL licence.

And yes, MySQL is a BAD example for a contribution policy, because in
fact they don't accept outside contributions. They also distribute
MaxDB under a dual licence model (GPL + commercial licence), but it's
impossible to contribute to MaxDB, because it is actively developed by
SAP and only by SAP.

Hopefully SUN will make it better with OpenJDK and establish a vital
community of contributors.

Volker


From David.Herron at Sun.COM  Fri Aug 17 00:52:50 2007
From: David.Herron at Sun.COM (David Herron)
Date: Thu, 16 Aug 2007 17:52:50 -0700
Subject: Will OpenJDK never contain other GPLed code?
In-Reply-To: <d65861880708161124n7d6195c9i822ffef85d11905e@mail.gmail.com>
References: <d65861880708151604l7d580dfere819d98158bb310b@mail.gmail.com>
	<46C38B71.9020503@sun.com>
	<d65861880708151701u4eb2b285sb4dea375caea0e40@mail.gmail.com>
	<46C46F84.7070907@sun.com>
	<d65861880708160933u721016aexff5fd271c4cbe6c1@mail.gmail.com>
	<d65861880708161037g1edc7a0fq4de8915e9f8c777a@mail.gmail.com>
	<46C495F0.60504@sun.com>
	<d65861880708161124n7d6195c9i822ffef85d11905e@mail.gmail.com>
Message-ID: <46C4F162.6070300@sun.com>

Volker Simonis wrote:
>> As an addendum to our discussion, I also found the article "Does dual
>> licensing threaten free software?" by Glyn Moody quite usefull and
>> interesting (see http://linuxjournal.com/node/1000069)
>>     


This article has some interesting thoughts in it, as do the comments.

Like:

    /The FSF could conceivably release close-source versions of the GNU
    tools if they wanted to; nobody would be able to stop them. Maybe
    the argument should be "copyright assignment gives power away to
    someone else", which is true irrespective of any pre-existing
    dual-licenses.
    /

The SCA by contrast gives dual ownership, making it more attractive than 
simply assigning copyright and giving your code away to someone else.  
Under the SCA both parties have power.

And:

    /You can fork the GNU GPL code, certainly; but you can't "fork" the
    commercial model. That is, you can't really set up another company
    to do what the copyright holders are doing. So, in a sense, you have
    lost the power to offer users a choice, generally one of the key
    features of the free software world, of which these businesses are
    now a part.
    /

I see that was written a year ago, before the OpenJDK license strategy 
was announced, and during the time the team here was still studying what 
to do.  However if I understand this statement correctly, we are close 
to doing what this guy says can't be done in that other entities besides 
Sun are offering commercialized Java implementations most of which are 
derived from the same code Sun now has under the OpenJDK project.  They 
are doing this by getting an appropriate license from Sun allowing that 
use of the code.

And:

    / It is only when a company owns all the copyright in the code that
    it can employ dual licensing. /

In one of your other replies you noted how hard it would be for a 
GPL-licensed project to donate their code to Sun because of the mixed 
ownership.  Another way of looking at this is when a given project is 
careful about copyright ownership, the project in question can be more 
flexible about how the code is used and licensed.  For example Linux may 
never be able to move from GPLv2 because the project never had copyright 
assignment to a central party. 

- David Herron





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070816/85a86983/attachment.html>

From Tom.Marble at Sun.COM  Fri Aug 17 21:35:41 2007
From: Tom.Marble at Sun.COM (Tom Marble)
Date: Fri, 17 Aug 2007 16:35:41 -0500
Subject: LinAlg API
In-Reply-To: <1IGJ4K-1MOuhs0@fwd26.aul.t-online.de>
References: <1IGJ4K-1MOuhs0@fwd26.aul.t-online.de>
Message-ID: <46C614AD.8000904@sun.com>

Alexander Schunk wrote:
> i just wanted to announce that i have written a tiny API for linear
> Algebra that offers basic Vector, Matrix and Complex Number arithmetics.
> [...]
> I am also working on some other math stuff - Fractional Numbers,
> Statistics - and think this could also be helpful for Java programms
> though there are specialised APIs already out there.

I'm a huge fan of these sorts of functionality.  We struggle
with what to do here because if we, indeed, add stuff which will
become part of JDK 7 that involves new API's then a JSR will need
to be written for it, etc.

Obviously there are times when we do want to add API's which will
become part of the core platform forever.  Other times certain
API's may be happier as libraries... and I'm confident that
the Java Module System will help us with that management issue.

I think that in either case you can start with these API's
as a library.  The next question is: "does it make sense for
LinAlg to be an experimental project within OpenJDK"?
Perhaps it would be good to get a couple more opinions on
the matter before making a formal project proposal.

Regards,

--Tom


From Kelly.Ohair at Sun.COM  Wed Aug 22 01:07:05 2007
From: Kelly.Ohair at Sun.COM (Kelly O'Hair)
Date: Tue, 21 Aug 2007 18:07:05 -0700
Subject: OpenJDK Mercurial Trsntiion Update 1
Message-ID: <46CB8C39.8090007@sun.com>


I decided to do this update with a blog that I could edit instead of
an email:

    http://weblogs.java.net/blog/kellyohair/archive/2007/08/openjdk_mercuri.html


-kto



From cresley_battlelite at hotmail.com  Wed Aug 22 05:28:12 2007
From: cresley_battlelite at hotmail.com (cresley battlelite)
Date: Wed, 22 Aug 2007 17:28:12 +1200
Subject: Too many open files exception ignored by
	UnixPrintServiceLookup.execCmd()
Message-ID: <BAY114-W1661C4E3CC5914A0C73744F9D50@phx.gbl>

When I request a javax.print.PrintJob to print on Unix, the PrintService checks to see if the printer is accepting jobs and can return a false negative if there are too many open files because the following Exception is ignored (see j2se/src/solaris/classes/sun/print/UnixPrintServiceLookup.java)
 
java.security.PrivilegedActionException: java.io.IOException: Too many open filesCaused by: java.io.IOException: Too many open files
I would have preferred a PrintException 
 
        } catch (PrivilegedActionException e) {        }
should change to (or similar) 
 
        } catch (PrivilegedActionException e) {                throw new PrintException("Unable to determine if printer is accepting jobs - " + e.getMessage());
        }
Is there a better area than discuss to discuss such things?
 
  
Cresley
_________________________________________________________________
Connect to the next generation of MSN Messenger?
http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-us&source=wlmailtagline
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070822/03746974/attachment.html>

From Phil.Race at Sun.COM  Wed Aug 22 14:11:39 2007
From: Phil.Race at Sun.COM (Phil Race)
Date: Wed, 22 Aug 2007 07:11:39 -0700
Subject: Too many open files exception ignored by
	UnixPrintServiceLookup.execCmd()
In-Reply-To: <BAY114-W1661C4E3CC5914A0C73744F9D50@phx.gbl>
References: <BAY114-W1661C4E3CC5914A0C73744F9D50@phx.gbl>
Message-ID: <46CC441B.6050502@sun.com>


 >Is there a better area than discuss to discuss such things?

yes. openjdk-discuss is not for discussion of specific bugs.

There are group specific aliases  and  2d-dev  is the more appropriate 
one in
this case since "2d" owns printing.

But the openjdk aliases (including 2d-dev) are really for people who are 
working
with the source code. Sounds like you may just want to report an 
apparent bug
I recommend this is better directed to JAVA2D-INTEREST at JAVA.SUN.COM

or if you want to submit a bug, go to bugs.sun.com

BTW the real question here is why do you have so many open files .. 
65536 is the
usual limit and that's a lot of open files.

-phil.

cresley battlelite wrote:
> When I request a javax.print.PrintJob to print on Unix, 
> the PrintService checks to see if the printer is accepting jobs and 
> can return a false negative if there are too many open files because 
> the following Exception is ignored (see 
> j2se/src/solaris/classes/sun/print/UnixPrintServiceLookup.java)
>  
> java.security.PrivilegedActionException: java.io.IOException: Too many 
> open files
> Caused by: java.io.IOException: Too many open files
>
> I would have preferred a PrintException
>  
>         } catch (PrivilegedActionException e) {
>         }
>
> should change to (or similar) 
>  
>         } catch (PrivilegedActionException e) {
>                 throw new PrintException("Unable to determine if 
> printer is accepting jobs - " + e.getMessage());
>         }
>
> Is there a better area than discuss to discuss such things?
>  
>   
> Cresley
>
> ------------------------------------------------------------------------
> Connect to the next generation of MSN Messenger  Get it now! 
> <http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-us&source=wlmailtagline>



From mr at sun.com  Fri Aug 24 18:15:49 2007
From: mr at sun.com (Mark Reinhold)
Date: Fri, 24 Aug 2007 11:15:49 -0700
Subject: Search facility for the list archives? 
In-Reply-To: david.herron@sun.com; Wed, 15 Aug 2007 14:04:58 PDT;
	<46C36A7A.9010303@sun.com> 
Message-ID: <20070824181549.4BCB03B5A@eggemoggin.niobe.net>

> Date: Wed, 15 Aug 2007 14:04:58 -0700
> From: david.herron at sun.com

> ...
> 
> It seems to me that going to google and entering:
> 
>     site:mail.openjdk.java.net search string
> 
> would be a very good way to handle this.  But there's a robot.txt saying
> otherwise...
> 
>      wget http://mail.openjdk.java.net/robots.txt
> 
> Yields:
> 
>     User-agent: *
>     Disallow: /
> 
> There must be a reason for this to which I'm not privy.  To me it makes
> absolute perfect sense to index this stuff because it's the search
> engines that draw the most traffic to a web site.

You're completely right.

There used to be a reason for that robots.txt.  The new openjdk.java.net
domains were actually online a few days before the launch at JavaOne, and
we (obviously) didn't want search engines to index the site prior to the
launch.  We removed the robots.txt files from openjdk.java.net but forgot
to removing the one on mail.o.j.n.

I've now removed it.  Give Google a couple of days, and you should be all
set.

- Mark


From Igor.Nekrestyanov at Sun.COM  Fri Aug 24 18:19:28 2007
From: Igor.Nekrestyanov at Sun.COM (Igor Nekrestyanov)
Date: Fri, 24 Aug 2007 22:19:28 +0400
Subject: Search facility for the list archives?
In-Reply-To: <20070824181549.4BCB03B5A@eggemoggin.niobe.net>
References: <20070824181549.4BCB03B5A@eggemoggin.niobe.net>
Message-ID: <46CF2130.3090601@sun.com>

Can we also remove https://openjdk.dev.java.net/robots.txt?
So sources will be searchable too.

-igor

Mark Reinhold wrote:
>> Date: Wed, 15 Aug 2007 14:04:58 -0700
>> From: david.herron at sun.com
>>     
>
>   
>> ...
>>
>> It seems to me that going to google and entering:
>>
>>     site:mail.openjdk.java.net search string
>>
>> would be a very good way to handle this.  But there's a robot.txt saying
>> otherwise...
>>
>>      wget http://mail.openjdk.java.net/robots.txt
>>
>> Yields:
>>
>>     User-agent: *
>>     Disallow: /
>>
>> There must be a reason for this to which I'm not privy.  To me it makes
>> absolute perfect sense to index this stuff because it's the search
>> engines that draw the most traffic to a web site.
>>     
>
> You're completely right.
>
> There used to be a reason for that robots.txt.  The new openjdk.java.net
> domains were actually online a few days before the launch at JavaOne, and
> we (obviously) didn't want search engines to index the site prior to the
> launch.  We removed the robots.txt files from openjdk.java.net but forgot
> to removing the one on mail.o.j.n.
>
> I've now removed it.  Give Google a couple of days, and you should be all
> set.
>
> - Mark
>   



From mr at sun.com  Sat Aug 25 03:45:06 2007
From: mr at sun.com (Mark Reinhold)
Date: Fri, 24 Aug 2007 20:45:06 -0700
Subject: Search facility for the list archives? 
In-Reply-To: igor.nekrestyanov@sun.com; Fri, 24 Aug 2007 22:19:28 +0400;
	<46CF2130.3090601@sun.com> 
Message-ID: <20070825034507.04E17B3C4@callebaut.niobe.net>

> Date: Fri, 24 Aug 2007 22:19:28 +0400
> From: igor.nekrestyanov at sun.com

> Can we also remove https://openjdk.dev.java.net/robots.txt?
> So sources will be searchable too.

So we're in this weird state where most of our content is on
openjdk.java.net, which is hosted on machines that we control
completely, except for the read-only Subversion repository,
which is on openjdk.DEV.java.net, which runs on top of the
regular old java.net infrastructure.

There are, unfortunately, aspects of the java.net infrastructure
over which we have essentially no control, and I'm afraid that
this robots.txt file is one of them.  Sorry.

In the long run, once we've made our Mercurial repositories
public -- and that will be on machines that we control -- then
we can do whatever we want.  We'll also, eventually, have an
OpenGrok instance set up for the OpenJDK sources.  That should
be much more efficient than Google for searching through the
source code.

- Mark


From alexanderschunk at t-online.de  Tue Aug 28 20:36:00 2007
From: alexanderschunk at t-online.de (alexanderschunk at t-online.de)
Date: 28 Aug 2007 20:36 GMT
Subject: OpenJDK: Java application Startup 
Message-ID: <1IQ7nM-1zJbou0@fwd35.t-online.de>

Hi,

i would like to contribute to OpenJDK by improving Java application start up times.

There are two different kinds of application start up:

cold start:
-----------

this is a start immediately after system reboot 

warm start:
-------------

this is one ore more starts of an Java application when the VM is already loaded and the Java application can load data into the disk cache which has faster access then the reading the data from the disk itself.

The Problem:
-------------
currently Java applicatins - in particula on Windows - take much longer to load then normal .exe applications. this is because native .exe applications with precompiled headers or libs seem to be loaded faster than corresponding Java apps.

solutions:
-----------

Take advantage of disk cache
allow .exe generation of Java apps 
improve loading of Java application related ressources, files etc.


I dont know if there is any active development on this for OpenJDK at present but i think this is  crucial for any Java relase.

- Best Alex
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070828/4e621574/attachment.html>

From roman.kennke at aicas.com  Tue Aug 28 20:46:14 2007
From: roman.kennke at aicas.com (Roman Kennke)
Date: Tue, 28 Aug 2007 22:46:14 +0200
Subject: OpenJDK: Java application Startup
In-Reply-To: <1IQ7nM-1zJbou0@fwd35.t-online.de>
References: <1IQ7nM-1zJbou0@fwd35.t-online.de>
Message-ID: <1188333974.6029.10.camel@mercury>

Hi,

> The Problem:
> -------------
> currently Java applicatins - in particula on Windows - take much longer to load then normal .exe applications. this is because native .exe applications with precompiled headers or libs seem to be loaded faster than corresponding Java apps.
> 
> solutions:
> -----------
> 
> Take advantage of disk cache

In which way? Pre-load the VM after system reboot?

> allow .exe generation of Java apps 

How would that help anything?

> improve loading of Java application related ressources, files etc.

Not sure what could be improved here.
> 
> I dont know if there is any active development on this for OpenJDK at present but i think this is  crucial for any Java relase.

I think the Java kernel project (or how it's called, I don't know) aims
in this direction. I might be wrong though.

/Roman

-- 
Dipl.-Inform. (FH) Roman Kennke, Software Engineer, http://kennke.org
aicas Allerton Interworks Computer Automated Systems GmbH
Haid-und-Neu-Stra?e 18 * D-76131 Karlsruhe * Germany
http://www.aicas.com   * Tel: +49-721-663 968-0
USt-Id: DE216375633, Handelsregister HRB 109481, AG Karlsruhe
Gesch?ftsf?hrer: Dr. James J. Hunt



From Sandeep.Konchady at Sun.COM  Tue Aug 28 21:04:19 2007
From: Sandeep.Konchady at Sun.COM (Sandeep Konchady)
Date: Tue, 28 Aug 2007 14:04:19 -0700
Subject: OpenJDK: Java application Startup
In-Reply-To: <1188333974.6029.10.camel@mercury>
References: <1IQ7nM-1zJbou0@fwd35.t-online.de>
	<1188333974.6029.10.camel@mercury>
Message-ID: <46D48DD3.4010909@Sun.COM>


As Roman mentioned, there is some activity around improving performance 
of Java through Java Kernel project. This forum on java.net has some 
interesting information, in case you have not already seen it.

http://forums.java.net/jive/thread.jspa?threadID=22378&tstart=0

As for creating .exe from Java program, it has already been done by gcj, 
but I have not seen or heard about significant performance gain or 
startup improvements by doing this.

http://www.linuxjournal.com/article/4860

- Sandeep

Roman Kennke wrote:
> Hi,
>
>   
>> The Problem:
>> -------------
>> currently Java applicatins - in particula on Windows - take much longer to load then normal .exe applications. this is because native .exe applications with precompiled headers or libs seem to be loaded faster than corresponding Java apps.
>>
>> solutions:
>> -----------
>>
>> Take advantage of disk cache
>>     
>
> In which way? Pre-load the VM after system reboot?
>
>   
>> allow .exe generation of Java apps 
>>     
>
> How would that help anything?
>
>   
>> improve loading of Java application related ressources, files etc.
>>     
>
> Not sure what could be improved here.
>   
>> I dont know if there is any active development on this for OpenJDK at present but i think this is  crucial for any Java relase.
>>     
>
> I think the Java kernel project (or how it's called, I don't know) aims
> in this direction. I might be wrong though.
>
> /Roman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070828/974b41fd/attachment.html>

From David.Herron at Sun.COM  Tue Aug 28 21:33:45 2007
From: David.Herron at Sun.COM (David Herron)
Date: Tue, 28 Aug 2007 14:33:45 -0700
Subject: OpenJDK: Java application Startup
In-Reply-To: <46D48DD3.4010909@Sun.COM>
References: <1IQ7nM-1zJbou0@fwd35.t-online.de>
	<1188333974.6029.10.camel@mercury> <46D48DD3.4010909@Sun.COM>
Message-ID: <46D494B9.5030908@sun.com>

There's also an internal working group who've been focusing on 
performance and startup time.  Perhaps they should externalize that 
work?  The past performance improvements that have been gained in Java 
startup have come from that group.

Some other issues with compiling a java app to native code are: it's no 
longer cross platform (may or may not be important); and how do you 
handle dynamically downloaded classes?  The latter requires having some 
kind of .class->.exe runtime compilation.

Finally, your query should be redirected to a group other than "discuss" 
... but where?  The existing mailing lists doesn't have an appropriate 
place, and the solutions also cuts across the whole platform.

I think that by the existing group/project structure
    http://openjdk.java.net/groups/
    http://openjdk.java.net/projects/
there should be a "Performance Group" formed.

- David Herron



Sandeep Konchady wrote:
>
> As Roman mentioned, there is some activity around improving 
> performance of Java through Java Kernel project. This forum on 
> java.net has some interesting information, in case you have not 
> already seen it.
>
> http://forums.java.net/jive/thread.jspa?threadID=22378&tstart=0
>
> As for creating .exe from Java program, it has already been done by 
> gcj, but I have not seen or heard about significant performance gain 
> or startup improvements by doing this.
>
> http://www.linuxjournal.com/article/4860
>
> - Sandeep
>
> Roman Kennke wrote:
>> Hi,
>>
>>   
>>> The Problem:
>>> -------------
>>> currently Java applicatins - in particula on Windows - take much longer to load then normal .exe applications. this is because native .exe applications with precompiled headers or libs seem to be loaded faster than corresponding Java apps.
>>>
>>> solutions:
>>> -----------
>>>
>>> Take advantage of disk cache
>>>     
>>
>> In which way? Pre-load the VM after system reboot?
>>
>>   
>>> allow .exe generation of Java apps 
>>>     
>>
>> How would that help anything?
>>
>>   
>>> improve loading of Java application related ressources, files etc.
>>>     
>>
>> Not sure what could be improved here.
>>   
>>> I dont know if there is any active development on this for OpenJDK at present but i think this is  crucial for any Java relase.
>>>     
>>
>> I think the Java kernel project (or how it's called, I don't know) aims
>> in this direction. I might be wrong though.
>>
>> /Roman

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070828/c5227613/attachment.html>

From aph at redhat.com  Wed Aug 29 10:06:04 2007
From: aph at redhat.com (Andrew Haley)
Date: Wed, 29 Aug 2007 11:06:04 +0100
Subject: OpenJDK: Java application Startup
In-Reply-To: <46D48DD3.4010909@Sun.COM>
References: <1IQ7nM-1zJbou0@fwd35.t-online.de>
	<1188333974.6029.10.camel@mercury> <46D48DD3.4010909@Sun.COM>
Message-ID: <18133.17676.188122.153524@zebedee.pink>

Sandeep Konchady writes:
 > 
 > As Roman mentioned, there is some activity around improving performance 
 > of Java through Java Kernel project. This forum on java.net has some 
 > interesting information, in case you have not already seen it.
 > 
 > http://forums.java.net/jive/thread.jspa?threadID=22378&tstart=0
 > 
 > As for creating .exe from Java program, it has already been done by gcj, 
 > but I have not seen or heard about significant performance gain or 
 > startup improvements by doing this.

There is a startup time improvement with gcj, but this tends somewhat
to be masked by the large time spent by the shared library loader
doing fixups.  Prelinking helps this, though:


zorro:tmp $ gcj Hello.java --main=Hello -o Hello
zorro:tmp $ time ./Hello
Hello, World!

real    0m0.074s
user    0m0.037s
sys     0m0.038s
zorro:tmp $ sudo prelink Hello
zorro:tmp $ time ./Hello
Hello, World!

real    0m0.045s
user    0m0.017s
sys     0m0.028s
zorro:tmp $ time java Hello
Hello, World!

real    0m0.131s
user    0m0.052s
sys     0m0.043s

Andrew.

-- 
Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, UK
Registered in England and Wales No. 3798903


From Tom.Marble at Sun.COM  Wed Aug 29 11:53:15 2007
From: Tom.Marble at Sun.COM (Tom Marble)
Date: Wed, 29 Aug 2007 06:53:15 -0500
Subject: OpenJDK: Java application Startup
In-Reply-To: <1IQ7nM-1zJbou0@fwd35.t-online.de>
References: <1IQ7nM-1zJbou0@fwd35.t-online.de>
Message-ID: <46D55E2B.9000205@sun.com>

alexanderschunk at t-online.de wrote:
> i would like to contribute to OpenJDK by improving Java application
> start up times.
This is a common concern!

> There are two different kinds of application start up: 
> cold start:
> -----------
> 
> this is a start immediately after system reboot
> warm start:
> -------------
> [...]
> I dont know if there is any active development on this for OpenJDK at
> present but i think this is  crucial for any Java relase.
As David points out we've had a virtual team working on this
for several years.  One of the issues which cannot be separated
from startup is memory footprint.  There is often a trade-off between
these two.  And, to complicate matters, measuring footprint and
accounting for it accurately (shared libs vs. per process memory)
is very difficult to do in a cross platform way.  Hint: TaskManager
does *not* tell the whole story!

We have actually converted many JDK methods to lazy initialization
and done some fairly sophisticated things like re-ordering
classes in rt.jar.

In any case the most easy improvements have probably been made
so this work gets progressively harder.

As Roman points out... there is a new effort to pre-warm the
file system cache called Quickstarter which we are working on:
http://java.sun.com/developer/technicalArticles/javase/consumerjre/#Quickstarter

Regards,

--Tom


From alexanderschunk at t-online.de  Wed Aug 29 19:38:00 2007
From: alexanderschunk at t-online.de (alexanderschunk at t-online.de)
Date: 29 Aug 2007 19:38 GMT
Subject: OpenJDK Performance Group/Project
Message-ID: <1IQTNP-1UGBN20@fwd30.t-online.de>

 I think that by the existing group/project structure
> http://openjdk.java.net/groups/
> http://openjdk.java.net/projects/
> there should be a "Performance Group" formed.

I think this idea from David sounds interesting. Covering all the things of the last discussion i grasped two things:

1.Quickstart development is going on => to be included in a Java 6 update
2. Java Kernel is also in development and will be an extension to the JRE - like an on-demand online installer.

I have run a Profiler test on my LinAlg API and got some good results that were in the ms - milliseconds - realm.

Sure, using .exe format rather than .jar files will break cross-plattform usability of Java apps - however, some people might
need plattform specific generation of executables if they focus only on this specific plattform. 

It would be interesting to know how Java apps actually behave after a system reboot - that is cold start - and how this differs from
warm start behaviour. 

- Alex
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070829/0a1f8854/attachment.html>

From alexanderschunk at t-online.de  Thu Aug 30 06:49:00 2007
From: alexanderschunk at t-online.de (alexanderschunk at t-online.de)
Date: 30 Aug 2007 06:49 GMT
Subject: OpenJDK: Performance Group
Message-ID: <1IQdqm-186e4u0@fwd26.t-online.de>

Hi,

as David Herron already suggested it would be interesting to have a "Performance Group" that deals with Java performance issues, including start up times, application run time performance, memory management of objects while running etc that is all the critical stuff of Java apps. This would help understand why, i.g. some applications crash if they run longer than a day - there was a bug in the bug database on this. 


Tom was speaking about publishing some figures and i think it would be interesting for developers - in particular those working on ports of OpenJDK - to see how the JRE or VM and applications behave on different plattforms. 

Since there is currently no mailing-list for discussions on this just consider this as an "experimantal" outline of Davids idea of a Performance group.

May be David can outline his initial idea a bit more.

- Alex
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070830/5c6c2d1c/attachment.html>

From pdoubleya at gmail.com  Fri Aug 31 05:49:09 2007
From: pdoubleya at gmail.com (Patrick Wright)
Date: Fri, 31 Aug 2007 07:49:09 +0200
Subject: JavaSE 7 build 19 is available at the openjdk.java.net website
In-Reply-To: <46D793F9.6040602@sun.com>
References: <46D793F9.6040602@sun.com>
Message-ID: <64efa1ba0708302249k985aee6h2b1d336e9f23b7ad@mail.gmail.com>

Hi

This link is not working
http://download.java.net/jdk7/changes/jdk7-b19.html


Thanks
Patrick


From Xiomara.Jayasena at Sun.COM  Fri Aug 31 06:09:40 2007
From: Xiomara.Jayasena at Sun.COM (Xiomara Jayasena)
Date: Thu, 30 Aug 2007 23:09:40 -0700
Subject: JavaSE 7 build 19 is available at the openjdk.java.net website
In-Reply-To: <64efa1ba0708302249k985aee6h2b1d336e9f23b7ad@mail.gmail.com>
References: <46D793F9.6040602@sun.com>
	<64efa1ba0708302249k985aee6h2b1d336e9f23b7ad@mail.gmail.com>
Message-ID: <46D7B0A4.5070809@sun.com>


It seems to be an intermittent problem, I can see it at times --  I'll 
inform the team that manage the servers.

-Xiomara

Patrick Wright wrote:
> Hi
>
> This link is not working
> http://download.java.net/jdk7/changes/jdk7-b19.html
>
>
> Thanks
> Patrick
>   



From David.Herron at Sun.COM  Fri Aug 31 13:14:00 2007
From: David.Herron at Sun.COM (David Herron)
Date: Fri, 31 Aug 2007 06:14:00 -0700
Subject: OpenJDK: Performance Group
In-Reply-To: <1IQdqm-186e4u0@fwd26.t-online.de>
References: <1IQdqm-186e4u0@fwd26.t-online.de>
Message-ID: <46D81418.6030906@sun.com>

alexanderschunk at t-online.de wrote:
> Hi,
>
> as David Herron already suggested it would be interesting to have a 
> "Performance Group" that deals with Java performance issues, including 
> start up times, application run time performance, memory management of 
> objects while running etc that is all the critical stuff of Java apps. 
> This would help understand why, i.g. some applications crash if they 
> run longer than a day - there was a bug in the bug database on this. 
>
>
> Tom was speaking about publishing some figures and i think it would be 
> interesting for developers - in particular those working on ports of 
> OpenJDK - to see how the JRE or VM and applications behave on 
> different plattforms. 
>
> Since there is currently no mailing-list for discussions on this just 
> consider this as an "experimantal" outline of Davids idea of a 
> Performance group.
>
> May be David can outline his initial idea a bit more.
>
> - Alex

Well it was an idea which popped up in the moment of writing that email, 
so I didn't give it any more thought than that.  However..  Performance 
is one of the key criticisms of the Java platform, which is why the 
performance virtual team has worked on this issue all these years.   
Like Quality it isn't about owning a module of code, it's about 
improving attributes of the overall thing.

To nitpick for a moment, a crash isn't about performance, it's about a 
bug.  It's just that some bugs occur after awhile of running e.g. if 
something were to be corrupted in the heap.

Anyway, mulling for a moment what a performance group might look like 
I'm having a hard time distinguishing what the internal performance 
virtual team does.  They have a set of methodologies they've defined for 
measuring various performance characteristics, most especially they've 
focused on startup time.  They have a set of tests they run to measure 
performance and startup time.  They have a reporting system to gather 
data on performance and startup time.  And the team is cross-functional 
because the issues cut across the whole platform.

- David Herron

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/discuss/attachments/20070831/4c68b8d4/attachment.html>