Eclipse Community Forums
Forum Search:

Search      Help    Register    Login    Home
Home » Eclipse Projects » Orbit » JSR166 (java.util.concurrent) backport for Orbit
JSR166 (java.util.concurrent) backport for Orbit [message #15230] Fri, 05 December 2008 05:52 Go to next message
Eclipse User
Originally posted by: news.eclipse.org.08.lemmster.de

Hi,

for bug #257381 ("[Discovery] Reduce BREE to Java 1.4") [1] ECF needs a
backport of java.util.concurrent. The Distributed Computing Laboratory
at Emory University [2] provides such a backport and in my local tests
the lib seems to keep its promises.
Is anybody interested in such a backport in Orbit? Does it make sense to
have it in Orbit? If not, I won't go through the hassle of creating an
Orbit-esque version and simple embed it in the ECF consumer bundle.

Markus

[1] https://bugs.eclipse.org/257381
[2] http://backport-jsr166.sourceforge.net/
Re: JSR166 (java.util.concurrent) backport for Orbit [message #15262 is a reply to message #15230] Mon, 08 December 2008 19:49 Go to previous messageGo to next message
Eclipse User
Originally posted by: david_williams.linux.vnet.ibm.com

Markus Alexander Kuppe wrote:
> ...
> Is anybody interested in such a backport in Orbit? Does it make sense to
> have it in Orbit? If not, I won't go through the hassle of creating an
> Orbit-esque version and simple embed it in the ECF consumer bundle.
> ...
Even if no one is interested (right now) I'd encourage you to put it in
Orbit anyway ... since it encourages re-use later.
But I agree, it's not critically important to put in Orbit.

How much hassle is it, anyway? Just curious what the "inhibitors" are to
putting things in Orbit.

Too bad the code can't be written to run on Java 5 without the "extra"
plugin and only require the backport if running on Java 4 (or, is that
the plan?).

Thanks,
Re: JSR166 (java.util.concurrent) backport for Orbit [message #15295 is a reply to message #15262] Tue, 09 December 2008 08:07 Go to previous messageGo to next message
Eclipse User
Originally posted by: news.eclipse.org.08.lemmster.de

David Williams wrote:

> Too bad the code can't be written to run on Java 5 without the "extra"
> plugin and only require the backport if running on Java 4 (or, is that
> the plan?).

Hi David,

I don't see a way to write either the bundle or consumers in a way that
it falls back to java.util.concurrency on >= 1.5. If other do, please
speak up. :)
Thus the idea momentarily is to have just one bundle for all >= 1.4
VMs. backports 1.4 is upwards compatibility and will run on any vm >=
1.4 [1]. I've verified this by running the concurrency tests on
different VM versions.
There is a performance penalty though. On VMs >= 1.5 backports-1.4
obviously doesn't exploit newer VM performance features. For the ECF use
case, this is totally acceptable though.

I've created an initial take at ECF's OSU repo [3][4].

[1]
http://concurrency.markmail.org/message/wx6pxm3fynlwirwr?q=b ackport&page=2
[3] http://ecf1.osuosl.org/
[4] plugins/edu.emory.mathcs.backport

How do we continue from here? Can you file a CQ for inclusion in Orbit?

Markus
Re: JSR166 (java.util.concurrent) backport for Orbit [message #15328 is a reply to message #15295] Tue, 09 December 2008 12:28 Go to previous messageGo to next message
Eclipse User
Originally posted by: david_williams.linux.vnet.ibm.com

Markus Alexander Kuppe wrote:
> How do we continue from here? Can you file a CQ for inclusion in Orbit?
>
>
The thing to do is to file a CQ for your own Eclipse Project, then once
that's approved, an "administrative" one is open to put it into Orbit.
Re: JSR166 (java.util.concurrent) backport for Orbit [message #15394 is a reply to message #15328] Wed, 10 December 2008 06:09 Go to previous messageGo to next message
Eclipse User
Originally posted by: news.eclipse.org.08.lemmster.de

David Williams wrote:
> Markus Alexander Kuppe wrote:
>> How do we continue from here? Can you file a CQ for inclusion in Orbit?
>>
> The thing to do is to file a CQ for your own Eclipse Project, then once
> that's approved, an "administrative" one is open to put it into Orbit.

Done, https://dev.eclipse.org/ipzilla/show_bug.cgi?id=2882
Re: JSR166 (java.util.concurrent) backport for Orbit [message #15524 is a reply to message #15328] Tue, 16 December 2008 02:33 Go to previous messageGo to next message
Eclipse User
Originally posted by: news.eclipse.org.08.lemmster.de

David Williams wrote:
> Markus Alexander Kuppe wrote:
>> How do we continue from here? Can you file a CQ for inclusion in Orbit?
>>
>>
> The thing to do is to file a CQ for your own Eclipse Project, then once
> that's approved, an "administrative" one is open to put it into Orbit.

Hi David,

CQ 2882 [1] has been approved under the terms of parallel IP. Are you
going to file the Orbit CQ now?

Markus

[1] http://dev.eclipse.org/ipzilla/show_bug.cgi?id=2882
Re: JSR166 (java.util.concurrent) backport for Orbit [message #15554 is a reply to message #15524] Tue, 16 December 2008 03:08 Go to previous messageGo to next message
Eclipse User
Originally posted by: david_williams.linux.vnet.ibm.com

Markus Alexander Kuppe wrote:
> David Williams wrote:
>
> Hi David,
>
> CQ 2882 [1] has been approved under the terms of parallel IP. Are you
> going to file the Orbit CQ now?
>
> Markus
>
> [1] http://dev.eclipse.org/ipzilla/show_bug.cgi?id=2882
>
>
Sure, I'll help you out. I'm assuming no one else on your project is an
Orbit committer?
I won't be speedy getting it into Orbit, but probably can by January.
Sound ok?
Re: JSR166 (java.util.concurrent) backport for Orbit [message #15585 is a reply to message #15554] Tue, 16 December 2008 03:21 Go to previous messageGo to next message
Eclipse User
Originally posted by: news.eclipse.org.08.lemmster.de

David Williams wrote:
> Markus Alexander Kuppe wrote:
>> David Williams wrote:
>> Hi David,
>>
>> CQ 2882 [1] has been approved under the terms of parallel IP. Are you
>> going to file the Orbit CQ now?
>>
>> Markus
>>
>> [1] http://dev.eclipse.org/ipzilla/show_bug.cgi?id=2882
>>
>>
> Sure, I'll help you out. I'm assuming no one else on your project is an
> Orbit committer?
> I won't be speedy getting it into Orbit, but probably can by January.
> Sound ok?

Chris Aniszczyk is both a committer on Orbit as well as ECF.

Markus
Re: JSR166 (java.util.concurrent) backport for Orbit [message #15616 is a reply to message #15585] Tue, 16 December 2008 03:49 Go to previous messageGo to next message
Eclipse User
Originally posted by: david_williams.linux.vnet.ibm.com

Markus Alexander Kuppe wrote:
> Chris Aniszczyk is both a committer on Orbit as well as ECF.
>
>
Well, I wouldn't want to take the honor away from Chris :)
I opened the Orbit CQ, but Chris would be the better choice to actually
put it in there and list as technical contact,
since he'll have a long term interest in it.
If he can't for some reason, I'd still be glad to help.
Re: JSR166 (java.util.concurrent) backport for Orbit [message #15647 is a reply to message #15616] Tue, 23 December 2008 16:12 Go to previous messageGo to next message
Eclipse User
Originally posted by: eclipse-news.lemmster.de

David Williams wrote:
> Markus Alexander Kuppe wrote:
>> Chris Aniszczyk is both a committer on Orbit as well as ECF.
>>
>>
> Well, I wouldn't want to take the honor away from Chris :)
> I opened the Orbit CQ, but Chris would be the better choice to actually
> put it in there and list as technical contact,
> since he'll have a long term interest in it.
> If he can't for some reason, I'd still be glad to help.
>

Just so everyone knows, concurrent backport 3.1 has made it into Orbit
[1]. Thanks Chris!

Markus


[1]
:pserver:anonymous@dev.eclipse.org:/cvsroot/tools/org.eclipse.orbit/edu.emory.mathcs.backport
New Orbit build (was Re: JSR166 (java.util.concurrent) backport for Orbit) [message #16418 is a reply to message #15647] Tue, 20 January 2009 07:24 Go to previous message
Eclipse User
Originally posted by: news.eclipse.org.08.lemmster.de

Markus Kuppe wrote:
> David Williams wrote:
>> Markus Alexander Kuppe wrote:
>>> Chris Aniszczyk is both a committer on Orbit as well as ECF.
>>>
>>>
>> Well, I wouldn't want to take the honor away from Chris :)
>> I opened the Orbit CQ, but Chris would be the better choice to actually
>> put it in there and list as technical contact,
>> since he'll have a long term interest in it.
>> If he can't for some reason, I'd still be glad to help.
>>
>
> Just so everyone knows, concurrent backport 3.1 has made it into Orbit
> [1]. Thanks Chris!
>
> Markus
>
>
> [1]
> :pserver:anonymous@dev.eclipse.org:/cvsroot/tools/org.eclipse.orbit/edu.emory.mathcs.backport

Any plans to make edu.emory.mathcs.backports available in an Orbit build?

Markus
Re: JSR166 (java.util.concurrent) backport for Orbit [message #564677 is a reply to message #15230] Mon, 08 December 2008 19:49 Go to previous message
David Williams is currently offline David Williams
Messages: 176
Registered: July 2009
Senior Member
Markus Alexander Kuppe wrote:
> ...
> Is anybody interested in such a backport in Orbit? Does it make sense to
> have it in Orbit? If not, I won't go through the hassle of creating an
> Orbit-esque version and simple embed it in the ECF consumer bundle.
> ...
Even if no one is interested (right now) I'd encourage you to put it in
Orbit anyway ... since it encourages re-use later.
But I agree, it's not critically important to put in Orbit.

How much hassle is it, anyway? Just curious what the "inhibitors" are to
putting things in Orbit.

Too bad the code can't be written to run on Java 5 without the "extra"
plugin and only require the backport if running on Java 4 (or, is that
the plan?).

Thanks,
Re: JSR166 (java.util.concurrent) backport for Orbit [message #564699 is a reply to message #15262] Tue, 09 December 2008 08:07 Go to previous message
Markus Kuppe is currently offline Markus Kuppe
Messages: 177
Registered: July 2009
Senior Member
David Williams wrote:

> Too bad the code can't be written to run on Java 5 without the "extra"
> plugin and only require the backport if running on Java 4 (or, is that
> the plan?).

Hi David,

I don't see a way to write either the bundle or consumers in a way that
it falls back to java.util.concurrency on >= 1.5. If other do, please
speak up. :)
Thus the idea momentarily is to have just one bundle for all >= 1.4
VMs. backports 1.4 is upwards compatibility and will run on any vm >=
1.4 [1]. I've verified this by running the concurrency tests on
different VM versions.
There is a performance penalty though. On VMs >= 1.5 backports-1.4
obviously doesn't exploit newer VM performance features. For the ECF use
case, this is totally acceptable though.

I've created an initial take at ECF's OSU repo [3][4].

[1]
http://concurrency.markmail.org/message/wx6pxm3fynlwirwr?q=b ackport&page=2
[3] http://ecf1.osuosl.org/
[4] plugins/edu.emory.mathcs.backport

How do we continue from here? Can you file a CQ for inclusion in Orbit?

Markus
Re: JSR166 (java.util.concurrent) backport for Orbit [message #564722 is a reply to message #15295] Tue, 09 December 2008 12:28 Go to previous message
David Williams is currently offline David Williams
Messages: 176
Registered: July 2009
Senior Member
Markus Alexander Kuppe wrote:
> How do we continue from here? Can you file a CQ for inclusion in Orbit?
>
>
The thing to do is to file a CQ for your own Eclipse Project, then once
that's approved, an "administrative" one is open to put it into Orbit.
Re: JSR166 (java.util.concurrent) backport for Orbit [message #564771 is a reply to message #15328] Wed, 10 December 2008 06:09 Go to previous message
Markus Kuppe is currently offline Markus Kuppe
Messages: 177
Registered: July 2009
Senior Member
David Williams wrote:
> Markus Alexander Kuppe wrote:
>> How do we continue from here? Can you file a CQ for inclusion in Orbit?
>>
> The thing to do is to file a CQ for your own Eclipse Project, then once
> that's approved, an "administrative" one is open to put it into Orbit.

Done, https://dev.eclipse.org/ipzilla/show_bug.cgi?id=2882
Re: JSR166 (java.util.concurrent) backport for Orbit [message #564858 is a reply to message #15328] Tue, 16 December 2008 02:33 Go to previous message
Markus Kuppe is currently offline Markus Kuppe
Messages: 177
Registered: July 2009
Senior Member
David Williams wrote:
> Markus Alexander Kuppe wrote:
>> How do we continue from here? Can you file a CQ for inclusion in Orbit?
>>
>>
> The thing to do is to file a CQ for your own Eclipse Project, then once
> that's approved, an "administrative" one is open to put it into Orbit.

Hi David,

CQ 2882 [1] has been approved under the terms of parallel IP. Are you
going to file the Orbit CQ now?

Markus

[1] http://dev.eclipse.org/ipzilla/show_bug.cgi?id=2882
Re: JSR166 (java.util.concurrent) backport for Orbit [message #564889 is a reply to message #15524] Tue, 16 December 2008 03:08 Go to previous message
David Williams is currently offline David Williams
Messages: 176
Registered: July 2009
Senior Member
Markus Alexander Kuppe wrote:
> David Williams wrote:
>
> Hi David,
>
> CQ 2882 [1] has been approved under the terms of parallel IP. Are you
> going to file the Orbit CQ now?
>
> Markus
>
> [1] http://dev.eclipse.org/ipzilla/show_bug.cgi?id=2882
>
>
Sure, I'll help you out. I'm assuming no one else on your project is an
Orbit committer?
I won't be speedy getting it into Orbit, but probably can by January.
Sound ok?
Re: JSR166 (java.util.concurrent) backport for Orbit [message #564912 is a reply to message #15554] Tue, 16 December 2008 03:21 Go to previous message
Markus Kuppe is currently offline Markus Kuppe
Messages: 177
Registered: July 2009
Senior Member
David Williams wrote:
> Markus Alexander Kuppe wrote:
>> David Williams wrote:
>> Hi David,
>>
>> CQ 2882 [1] has been approved under the terms of parallel IP. Are you
>> going to file the Orbit CQ now?
>>
>> Markus
>>
>> [1] http://dev.eclipse.org/ipzilla/show_bug.cgi?id=2882
>>
>>
> Sure, I'll help you out. I'm assuming no one else on your project is an
> Orbit committer?
> I won't be speedy getting it into Orbit, but probably can by January.
> Sound ok?

Chris Aniszczyk is both a committer on Orbit as well as ECF.

Markus
Re: JSR166 (java.util.concurrent) backport for Orbit [message #564933 is a reply to message #15585] Tue, 16 December 2008 03:49 Go to previous message
David Williams is currently offline David Williams
Messages: 176
Registered: July 2009
Senior Member
Markus Alexander Kuppe wrote:
> Chris Aniszczyk is both a committer on Orbit as well as ECF.
>
>
Well, I wouldn't want to take the honor away from Chris :)
I opened the Orbit CQ, but Chris would be the better choice to actually
put it in there and list as technical contact,
since he'll have a long term interest in it.
If he can't for some reason, I'd still be glad to help.
Re: JSR166 (java.util.concurrent) backport for Orbit [message #564962 is a reply to message #15616] Tue, 23 December 2008 16:12 Go to previous message
Eclipse User
Originally posted by: eclipse-news.lemmster.de

David Williams wrote:
> Markus Alexander Kuppe wrote:
>> Chris Aniszczyk is both a committer on Orbit as well as ECF.
>>
>>
> Well, I wouldn't want to take the honor away from Chris :)
> I opened the Orbit CQ, but Chris would be the better choice to actually
> put it in there and list as technical contact,
> since he'll have a long term interest in it.
> If he can't for some reason, I'd still be glad to help.
>

Just so everyone knows, concurrent backport 3.1 has made it into Orbit
[1]. Thanks Chris!

Markus


[1]
:pserver:anonymous@dev.eclipse.org:/cvsroot/tools/org.eclipse.orbit/edu.emory.mathcs.backport
New Orbit build (was Re: JSR166 (java.util.concurrent) backport for Orbit) [message #565150 is a reply to message #15647] Tue, 20 January 2009 07:24 Go to previous message
Markus Kuppe is currently offline Markus Kuppe
Messages: 177
Registered: July 2009
Senior Member
Markus Kuppe wrote:
> David Williams wrote:
>> Markus Alexander Kuppe wrote:
>>> Chris Aniszczyk is both a committer on Orbit as well as ECF.
>>>
>>>
>> Well, I wouldn't want to take the honor away from Chris :)
>> I opened the Orbit CQ, but Chris would be the better choice to actually
>> put it in there and list as technical contact,
>> since he'll have a long term interest in it.
>> If he can't for some reason, I'd still be glad to help.
>>
>
> Just so everyone knows, concurrent backport 3.1 has made it into Orbit
> [1]. Thanks Chris!
>
> Markus
>
>
> [1]
> :pserver:anonymous@dev.eclipse.org:/cvsroot/tools/org.eclipse.orbit/edu.emory.mathcs.backport

Any plans to make edu.emory.mathcs.backports available in an Orbit build?

Markus
Previous Topic:Consuming Orbit Bundles in a PDE Build
Next Topic:Error with Xalan
Goto Forum:
  


Current Time: Fri Aug 22 02:07:18 EDT 2014

Powered by FUDForum. Page generated in 0.02637 seconds