[VOTE] Apache Artemis 1.2.0

classic Classic list List threaded Threaded
64 messages Options
1234
Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

dkulp

> On Dec 21, 2015, at 1:08 PM, Clebert Suconic <[hidden email]> wrote:
>
> In particular, in org_apache_activemq_artemis_jlibaio_LibaioContext.c,
> I see right at the top:
>
> #ifndef _GNU_SOURCE
> // libaio, O_DIRECT and other things won't be available without this define
> #define _GNU_SOURCE
> #endif
>
>
> this has nothing to do with LGPL licenses.. or even libaio on this
> instance.. That's how you enable O_DIRECT, with O_DIRECT being a Linux
> extension non conformant with POSIX.  the header is there for any
> source code using it.
>
>> …linked to libraries that are LGPL and only LGPL, which is not allowed per ASF policy.
>>
>
> Dynamic linked.. it doesn't not include libaio.

But is still REQUIRED for building the library…   Thus, the library still has a REQUIRED dependency on a LGPL’d library  to work.    Thus, it must be optional and the user must take explicit actions to enable it during build and runtime.


--
Daniel Kulp
[hidden email] - http://dankulp.com/blog
Talend Community Coder - http://coders.talend.com

Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

clebertsuconic
> But is still REQUIRED for building the library…   Thus, the library still has a REQUIRED dependency on a LGPL’d library  to work.    Thus, it must be optional and the user must take explicit actions to enable it during build and runtime.

>>has a REQUIRED dependency on a LGPL’d library  to work<<

if the library is not available, we won't use it...  the use is optional.

The compilation is required to validate method calls. It would be
actually possible to compile the library without the actual libaio
installed, but I don't want to go that route as it would complicate
the build and code. But the use is still dynamic linked.
Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

chirino
In reply to this post by dkulp
To me this seem the same as linking to glibc which is an LGLP lib and
a lib which just about every app on linux links to.  Surely even apr
links to it right?

This can also be seen a linking against hibernate, but not distributing it.

On Mon, Dec 21, 2015 at 1:10 PM, Daniel Kulp <[hidden email]> wrote:

>
>> On Dec 21, 2015, at 1:08 PM, Clebert Suconic <[hidden email]> wrote:
>>
>> In particular, in org_apache_activemq_artemis_jlibaio_LibaioContext.c,
>> I see right at the top:
>>
>> #ifndef _GNU_SOURCE
>> // libaio, O_DIRECT and other things won't be available without this define
>> #define _GNU_SOURCE
>> #endif
>>
>>
>> this has nothing to do with LGPL licenses.. or even libaio on this
>> instance.. That's how you enable O_DIRECT, with O_DIRECT being a Linux
>> extension non conformant with POSIX.  the header is there for any
>> source code using it.
>>
>>> …linked to libraries that are LGPL and only LGPL, which is not allowed per ASF policy.
>>>
>>
>> Dynamic linked.. it doesn't not include libaio.
>
> But is still REQUIRED for building the library…   Thus, the library still has a REQUIRED dependency on a LGPL’d library  to work.    Thus, it must be optional and the user must take explicit actions to enable it during build and runtime.
>
>
> --
> Daniel Kulp
> [hidden email] - http://dankulp.com/blog
> Talend Community Coder - http://coders.talend.com
>



--
Hiram Chirino
Engineering | Red Hat, Inc.
[hidden email] | fusesource.com | redhat.com
skype: hiramchirino | twitter: @hiramchirino
Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

clebertsuconic
On Mon, Dec 21, 2015 at 1:22 PM, Hiram Chirino <[hidden email]> wrote:
> To me this seem the same as linking to glibc which is an LGLP lib and
> a lib which just about every app on linux links to.  Surely even apr
> links to it right?

That's correct... just like anything else in the Linux world.

as a matter of fact this code requires glibc as well.
Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

chirino
In reply to this post by chirino
On Mon, Dec 21, 2015 at 1:22 PM, Hiram Chirino <[hidden email]> wrote:
> To me this seem the same as linking to glibc which is an LGLP lib and
> a lib which just about every app on linux links to.  Surely even apr
> links to it right?
>
> This can also be seen a linking against hibernate, but not distributing it.
>

And it seems other apache projects are doing that.  See:

https://github.com/apache/tapestry5/blob/5.0/tapestry-hibernate/pom.xml

--
Hiram Chirino
Engineering | Red Hat, Inc.
[hidden email] | fusesource.com | redhat.com
skype: hiramchirino | twitter: @hiramchirino
Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

dkulp
In reply to this post by John D. Ament-2

> On Dec 21, 2015, at 12:41 PM, John D. Ament <[hidden email]> wrote:
>
> On Mon, Dec 21, 2015 at 12:34 PM Clebert Suconic <[hidden email]>
> wrote:
>
>>> Nothing's stopping you from including them in the binary release.  They
>>> should be excluded in the source release.
>>
>>
>> It's been easier to keep these .so there. I'm about to give up
>> maintaining 32 bits. but right now you would need to log on 32 bits..
>> compile it.. log on 64 bits.. compile it..to make a full binary
>> distribution from the source.
>>
>> removing the .so will only complicate things.. I don't think we should
>> be so purist on this matter.
>>
>
> I think you're thinking about removing the .so's from the git repo.  I'm
> not requesting that.  They simply can't be in the source release tar.gz/zip
> archives.


Back to this part, the DO have to be removed from the source  tar.gz.

Per Roy Fielding:

"Apache releases open source and ONLY open source.  Our releases are absolutely
forbidden to contain anything other than the open source code that is in our
vcs-of-record, meaning code that is in the form most likely to be edited by
recipients for the sake of modifying the product, and in some specific cases
the generated (and open) source code of build scripts."

http://mail-archives.apache.org/mod_mbox/incubator-general/201203.mbox/%3CC3656B87-A6DC-4D3D-B1EB-29911B7A8070%40gbiv.com%3E

So yes, this part MUST be done.

--
Daniel Kulp
[hidden email] - http://dankulp.com/blog
Talend Community Coder - http://coders.talend.com

Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

clebertsuconic
> Per Roy Fielding:
>
> "Apache releases open source and ONLY open source.  Our releases are absolutely
> forbidden to contain anything other than the open source code that is in our
> vcs-of-record, meaning code that is in the form most likely to be edited by
> recipients for the sake of modifying the product, and in some specific cases
> the generated (and open) source code of build scripts."
>


Ok, fair enough.. -1 (non binding) from me then...

I will remove it from the source distribution.
Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

John D. Ament-2
In reply to this post by dkulp
On Dec 21, 2015 1:42 PM, "Daniel Kulp" <[hidden email]> wrote:
>
>
> > On Dec 21, 2015, at 12:41 PM, John D. Ament <[hidden email]>
wrote:
> >
> > On Mon, Dec 21, 2015 at 12:34 PM Clebert Suconic <
[hidden email]>
> > wrote:
> >
> >>> Nothing's stopping you from including them in the binary release.
They

> >>> should be excluded in the source release.
> >>
> >>
> >> It's been easier to keep these .so there. I'm about to give up
> >> maintaining 32 bits. but right now you would need to log on 32 bits..
> >> compile it.. log on 64 bits.. compile it..to make a full binary
> >> distribution from the source.
> >>
> >> removing the .so will only complicate things.. I don't think we should
> >> be so purist on this matter.
> >>
> >
> > I think you're thinking about removing the .so's from the git repo.  I'm
> > not requesting that.  They simply can't be in the source release
tar.gz/zip
> > archives.
>
>
> Back to this part, the DO have to be removed from the source  tar.gz.
>
> Per Roy Fielding:
>
> "Apache releases open source and ONLY open source.  Our releases are
absolutely
> forbidden to contain anything other than the open source code that is in
our
> vcs-of-record, meaning code that is in the form most likely to be edited
by
> recipients for the sake of modifying the product, and in some specific
cases
> the generated (and open) source code of build scripts."
>
>
http://mail-archives.apache.org/mod_mbox/incubator-general/201203.mbox/%3CC3656B87-A6DC-4D3D-B1EB-29911B7A8070%40gbiv.com%3E
>
> So yes, this part MUST be done.

I'd recommend tabling the lgpl license issue and bringing it up on
legal-discuss.

>
> --
> Daniel Kulp
> [hidden email] - http://dankulp.com/blog
> Talend Community Coder - http://coders.talend.com
>
Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

clebertsuconic
In reply to this post by John D. Ament-2
> [ERROR] Command was /bin/sh -c cd
> "/Users/johnament/Downloads/apache-artemis-1.2.0 2/artemis-commons" &&
> /Library/Java/JavaVirtualMachines/jdk1.8.0_51.jdk/Contents/Home/jre/bin/java



The compilation issue is because you used a space ' ' on your directory

had you used apache-artemis-1.2.0_2 (instead of space) it would have
worked fine. I replicated it here with a " " as well on a mac (dunno
about other envs.. probably the same).
Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

John D. Ament-2
So... create a ticket to track this? I can compile other stuff with spaces
generally fine.  Not sure what the issue is here w/ a space.

On Mon, Dec 21, 2015 at 2:30 PM Clebert Suconic <[hidden email]>
wrote:

> > [ERROR] Command was /bin/sh -c cd
> > "/Users/johnament/Downloads/apache-artemis-1.2.0 2/artemis-commons" &&
> >
> /Library/Java/JavaVirtualMachines/jdk1.8.0_51.jdk/Contents/Home/jre/bin/java
>
>
>
> The compilation issue is because you used a space ' ' on your directory
>
> had you used apache-artemis-1.2.0_2 (instead of space) it would have
> worked fine. I replicated it here with a " " as well on a mac (dunno
> about other envs.. probably the same).
>
Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

clebertsuconic
On Mon, Dec 21, 2015 at 2:37 PM, John D. Ament <[hidden email]> wrote:
> So... create a ticket to track this? I can compile other stuff with spaces
> generally fine.  Not sure what the issue is here w/ a space.


yep.. I don't think it's a release blocker. although if someone is
able to pick it up we can have it before next respin.
Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

John D. Ament-2
On Mon, Dec 21, 2015 at 2:44 PM Clebert Suconic <[hidden email]>
wrote:

> On Mon, Dec 21, 2015 at 2:37 PM, John D. Ament <[hidden email]>
> wrote:
> > So... create a ticket to track this? I can compile other stuff with
> spaces
> > generally fine.  Not sure what the issue is here w/ a space.
>
>
> yep.. I don't think it's a release blocker. although if someone is
> able to pick it up we can have it before next respin.
>

The space issue comes from this configuration in the root pom


<activemq-surefire-argline>-Djava.util.logging.manager=org.jboss.logmanager.LogManager

 -Dlogging.configuration=file:${activemq.basedir}/tests/config/logging.properties
         -Djava.library.path=${activemq.basedir}/artemis-native/bin/
-Djgroups.bind_addr=localhost
-Dorg.apache.activemq.artemis.api.core.UDPBroadcastEndpointFactory.localBindAddress=localhost
         -Djava.net.preferIPv4Stack=true
      </activemq-surefire-argline>

I'll see if there's a way to fix it.
Reply | Threaded
Open this post in threaded view
|

[CANCEL][VOTE] Apache Artemis 1.2.0

Martyn Taylor
In reply to this post by John D. Ament-2
Cancelling vote, due to binaries in the source distro.

We are getting very close to the holiday period now, so let's pick the
1.2.0 release up after the new year.

Happy holidays all!

Regards
Martyn

On 21/12/15 16:42, John D. Ament wrote:

> Sorry but -1 (non-binding)
>
> There are binary files in the source release, under artemis-native/bin/
>
> Ideally, the source release would include the C/C++ code required to build
> these so's, but I'm not sure that they're anywhere in the artemis codebase.
>
> In addition, compiling on a mac, latest patches, I can't build the source
> release on 1.8.0_u51 with the following error:
>
> Error: Could not find or load main class
> 2.artemis-commons....tests.config.logging.properties
>
>
> Results :
>
>
> Tests run: 0, Failures: 0, Errors: 0, Skipped: 0
>
>
> [ERROR] Failed to execute goal
> org.apache.maven.plugins:maven-surefire-plugin:2.18.1:test (default-test)
> on project artemis-commons: Execution default-test of goal
> org.apache.maven.plugins:maven-surefire-plugin:2.18.1:test failed: The
> forked VM terminated without properly saying goodbye. VM crash or
> System.exit called?
>
> [ERROR] Command was /bin/sh -c cd
> "/Users/johnament/Downloads/apache-artemis-1.2.0 2/artemis-commons" &&
> /Library/Java/JavaVirtualMachines/jdk1.8.0_51.jdk/Contents/Home/jre/bin/java
> -Djava.util.logging.manager=org.jboss.logmanager.LogManager
> -Dlogging.configuration=file:/Users/johnament/Downloads/apache-artemis-1.2.0
> 2/artemis-commons/../tests/config/logging.properties
> -Djava.library.path=/Users/johnament/Downloads/apache-artemis-1.2.0
> 2/artemis-commons/../artemis-native/bin/ -Djgroups.bind_addr=localhost
> -Dorg.apache.activemq.artemis.api.core.UDPBroadcastEndpointFactory.localBindAddress=localhost
> -Djava.net.preferIPv4Stack=true -jar
> '/Users/johnament/Downloads/apache-artemis-1.2.0
> 2/artemis-commons/target/surefire/surefirebooter197336116502841627.jar'
> '/Users/johnament/Downloads/apache-artemis-1.2.0
> 2/artemis-commons/target/surefire/surefire8924091653426997785tmp'
> '/Users/johnament/Downloads/apache-artemis-1.2.0
> 2/artemis-commons/target/surefire/surefire_03004869270640953611tmp'
>
> [ERROR] -> [Help 1]
>
> On Fri, Dec 18, 2015 at 12:36 PM Martyn Taylor <[hidden email]> wrote:
>
>> Hello all.
>>
>> I'd like to propose an Apache Artemis 1.2.0 release.
>>
>> Since 1.1.0 we've had some significant improvements to performance,
>> particularly around the persistence layer.  In addition new features
>> have been added such as LDAP and OSGi support.  There has also been more
>> improvements to the OpenWire protocol.
>>
>> The release notes can be found here:
>>
>> https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315920&version=12333274
>>
>>
>> The binary distributions can be found here:
>>
>> https://repository.apache.org/content/repositories/orgapacheactivemq-1076/org/apache/activemq/apache-artemis/1.2.0/
>>
>>
>> The source archives can be found here:
>>
>> https://repository.apache.org/content/repositories/orgapacheactivemq-1076/org/apache/activemq/apache-artemis/1.2.0/
>>
>>
>> The Maven repository is here:
>> https://repository.apache.org/content/repositories/orgapacheactivemq-1076/
>>
>> The source tag:
>>
>> https://git-wip-us.apache.org/repos/asf?p=activemq-artemis.git;a=tag;h=refs/tags/1.2.0
>>
>>
>> The project website for that version has been staged to:
>> http://people.apache.org/~martyntaylor/
>>
>> The vote will remain open for 72 hours.
>>
>> [ ] +1 approve the release as Apache Artemis 1.2.0
>> [ ] +0 no opinion
>> [ ] -1 disapprove (and reason why)
>>
>> Here's my (non-binding) +1
>>
>> Regards
>>
>> Martyn
>>

Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

andytaylor
In reply to this post by Claus Ibsen
Claus,

fyi, Ive sent a PR with some of your suggestions, see
https://github.com/apache/activemq-artemis/pull/287.

I'll take a look at the stuff ive missed in the new year.

Merry Christmas.

Andy

On 21/12/15 07:45, Claus Ibsen wrote:

> On Mon, Dec 21, 2015 at 1:03 AM, Clebert Suconic
> <[hidden email]> wrote:
>>> Also the output of the help command in the html looks ugly. Its not
>>> spaced correctly.
>>
>> I'm not sure where you mean that?
>>
>
> Open the readme.html in your browser. And then compare that to the
> --help output from the CLI. At least in my browser the formatting in
> the readme.html is not presented nicely.
>
> eg remember this is about first impression, new users is likely to
> open the readme as the 1st thing.
>
>
>
>>
>>>
>>> And its a bit unclear that after creating a broker you must cd into
>>> that directory to start it. eg some people may think you can start the
>>> broker from the unzip like you can do in Apache ActiveMQ. I would like
>>> to see the readme more clear on this. Maybe a better step by step
>>> showing the commands and output etc so its really clear.
>>
>> +1
>>
>>>
>>> Also why not use a .md file instead of .html ?
>>
>> I'm confused what's best here. .html are easy to read (thinking of
>> beginners just downloading the distro). pehaps we could switch to .md
>> and have a converter tool? (although I wouldn't know which one to use
>> from maven).
>>
>>>
>>> In JMX the broker is listed as brokerName="0.0.0.0" but I was not
>>> given any opportunity to name the broker when I created it from the
>>> command line.
>>
>> the cli will only ask for the host if needed (that is if you're using
>> clustered). otherwise you can use options (use ./artemis help create
>> for that).
>>
>>>
>>> And I cannot see the broker name in the log when the broker start up.
>>> That name should IMHO be allowed to enter when creating, and maybe
>>> default to the directory name or something? And show the name in the
>>> log.
>>
>>
>> That's used for the JMX only.
>>
>
> Its still a part of naming the broker. IMHO you should not hide that
> fact. People should be able to see from the logs what the name would
> be in JMX so they are sure they find the correct broker.
>
> Now its also seldom to use 2+ message brokers in the same JVM. Not
> sure that is really best practice.
>
> Since the broker name is "dynamic" then it makes tooling harder to
> predict the mbean name tree. As 99,9% uses 1 broker in the JVM why not
> have it default as before. And then if you really want 2+ then you
> need to enable configure the managementBrokerName to make them unique
> in that tree.
>
>
>
>>>
>>> Also I would like to be abel to easily enable jolokia out of the box.
>>> Apache ActiveMQ has that, but Artemis does not.
>>
>> It doest.. it's part of the web interface.
>
> If Jolokia is enabled then please log the jolokia url in the logs when
> you start up. See how we do that in ActiveMQ.
>
> I can find some details in the user guide
> http://activemq.apache.org/artemis/docs/1.1.0/management.html
>
> And btw there is typo in the docs -
>
>   example to query thebrokers version
>
> -->
>
> there is no space between thebrokers
>
>
>
>>
>>>
>>> The web console is likely a bit disappointment, its just the
>>> documentation. People in 2015 would expect a management console. So
>>> maybe point that out in the docs that there is none (yet).
>>
>> +1
>>
>>>
>>> When you create a queue its logged at INFO level and the message is
>>> not with a capital letter, eg trying -> Trying
>>>
>>
>> +1 I was so used to this message that I never though about that.
>>
>>> And when the broker shutdown, I would like to see an INFO logging that
>>> says the uptime of the broker. See how we do that in ActiveMQ.
>>
>>
>> +1 Nice feature.
>>
>>
>>> Also IMHO you should log at INFO level when the broker is being
>>> shutdown, so the user can see from exactly in the log when the
>>> shutdown starts, in case there is problems during so.
>>>
>>> eg as shown above where I started ActiveMQ in the foreground, and then
>>> hit CTRL+C (eg ^C).
>>>
>>
>>
>> I will open the JIRAs for these on monday. (Or maybe commit those if
>> it's an easy fix).
>>
>>
>> I don't see a reason to withdraw the release based on your feedback.
>> Great stuff but I don't see a need to hold it (hence your +1 already)
>
> Yes its a +1 - great to see a new release out.
>

Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

John D. Ament-2
Just wondering, does anyone plan to raise the LGPL question w/ legal
discuss?  If we're waiting for the new year to do the next release, would
be good to at least start the discussion.

John

On Wed, Dec 23, 2015 at 11:04 AM Andy Taylor <[hidden email]> wrote:

> Claus,
>
> fyi, Ive sent a PR with some of your suggestions, see
> https://github.com/apache/activemq-artemis/pull/287.
>
> I'll take a look at the stuff ive missed in the new year.
>
> Merry Christmas.
>
> Andy
>
> On 21/12/15 07:45, Claus Ibsen wrote:
> > On Mon, Dec 21, 2015 at 1:03 AM, Clebert Suconic
> > <[hidden email]> wrote:
> >>> Also the output of the help command in the html looks ugly. Its not
> >>> spaced correctly.
> >>
> >> I'm not sure where you mean that?
> >>
> >
> > Open the readme.html in your browser. And then compare that to the
> > --help output from the CLI. At least in my browser the formatting in
> > the readme.html is not presented nicely.
> >
> > eg remember this is about first impression, new users is likely to
> > open the readme as the 1st thing.
> >
> >
> >
> >>
> >>>
> >>> And its a bit unclear that after creating a broker you must cd into
> >>> that directory to start it. eg some people may think you can start the
> >>> broker from the unzip like you can do in Apache ActiveMQ. I would like
> >>> to see the readme more clear on this. Maybe a better step by step
> >>> showing the commands and output etc so its really clear.
> >>
> >> +1
> >>
> >>>
> >>> Also why not use a .md file instead of .html ?
> >>
> >> I'm confused what's best here. .html are easy to read (thinking of
> >> beginners just downloading the distro). pehaps we could switch to .md
> >> and have a converter tool? (although I wouldn't know which one to use
> >> from maven).
> >>
> >>>
> >>> In JMX the broker is listed as brokerName="0.0.0.0" but I was not
> >>> given any opportunity to name the broker when I created it from the
> >>> command line.
> >>
> >> the cli will only ask for the host if needed (that is if you're using
> >> clustered). otherwise you can use options (use ./artemis help create
> >> for that).
> >>
> >>>
> >>> And I cannot see the broker name in the log when the broker start up.
> >>> That name should IMHO be allowed to enter when creating, and maybe
> >>> default to the directory name or something? And show the name in the
> >>> log.
> >>
> >>
> >> That's used for the JMX only.
> >>
> >
> > Its still a part of naming the broker. IMHO you should not hide that
> > fact. People should be able to see from the logs what the name would
> > be in JMX so they are sure they find the correct broker.
> >
> > Now its also seldom to use 2+ message brokers in the same JVM. Not
> > sure that is really best practice.
> >
> > Since the broker name is "dynamic" then it makes tooling harder to
> > predict the mbean name tree. As 99,9% uses 1 broker in the JVM why not
> > have it default as before. And then if you really want 2+ then you
> > need to enable configure the managementBrokerName to make them unique
> > in that tree.
> >
> >
> >
> >>>
> >>> Also I would like to be abel to easily enable jolokia out of the box.
> >>> Apache ActiveMQ has that, but Artemis does not.
> >>
> >> It doest.. it's part of the web interface.
> >
> > If Jolokia is enabled then please log the jolokia url in the logs when
> > you start up. See how we do that in ActiveMQ.
> >
> > I can find some details in the user guide
> > http://activemq.apache.org/artemis/docs/1.1.0/management.html
> >
> > And btw there is typo in the docs -
> >
> >   example to query thebrokers version
> >
> > -->
> >
> > there is no space between thebrokers
> >
> >
> >
> >>
> >>>
> >>> The web console is likely a bit disappointment, its just the
> >>> documentation. People in 2015 would expect a management console. So
> >>> maybe point that out in the docs that there is none (yet).
> >>
> >> +1
> >>
> >>>
> >>> When you create a queue its logged at INFO level and the message is
> >>> not with a capital letter, eg trying -> Trying
> >>>
> >>
> >> +1 I was so used to this message that I never though about that.
> >>
> >>> And when the broker shutdown, I would like to see an INFO logging that
> >>> says the uptime of the broker. See how we do that in ActiveMQ.
> >>
> >>
> >> +1 Nice feature.
> >>
> >>
> >>> Also IMHO you should log at INFO level when the broker is being
> >>> shutdown, so the user can see from exactly in the log when the
> >>> shutdown starts, in case there is problems during so.
> >>>
> >>> eg as shown above where I started ActiveMQ in the foreground, and then
> >>> hit CTRL+C (eg ^C).
> >>>
> >>
> >>
> >> I will open the JIRAs for these on monday. (Or maybe commit those if
> >> it's an easy fix).
> >>
> >>
> >> I don't see a reason to withdraw the release based on your feedback.
> >> Great stuff but I don't see a need to hold it (hence your +1 already)
> >
> > Yes its a +1 - great to see a new release out.
> >
>
>
Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

clebertsuconic
On Wed, Dec 23, 2015 at 1:55 PM, John D. Ament <[hidden email]> wrote:
> Just wondering, does anyone plan to raise the LGPL question w/ legal
> discuss?  If we're waiting for the new year to do the next release, would
> be good to at least start the discussion.


We had such discussion long ago with legal. I couldn't find that email
on my inbox but we specifically asked questions about it. We were ok
as I remember. Maybe someone else (Martyn?) will have it on their
inboxes. For that reason I don't want to go over the same issue we had
asked before.

The use of libaio is optional anyways and the system works as
expected. what also covers other questions we had here on this thread.
Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

clebertsuconic
also, there has also been questions about it during the donation
process.. licenses reviewed.. etc.. so I don't think we need to open a
new discussions over this. the binary inclusion on the source was
something that was fixed now.

The dependency on libaio on the C code is through through dynamic
linked library, and is the same as any C code depending on libc or
gcc.

On Wed, Dec 23, 2015 at 1:58 PM, Clebert Suconic
<[hidden email]> wrote:

> On Wed, Dec 23, 2015 at 1:55 PM, John D. Ament <[hidden email]> wrote:
>> Just wondering, does anyone plan to raise the LGPL question w/ legal
>> discuss?  If we're waiting for the new year to do the next release, would
>> be good to at least start the discussion.
>
>
> We had such discussion long ago with legal. I couldn't find that email
> on my inbox but we specifically asked questions about it. We were ok
> as I remember. Maybe someone else (Martyn?) will have it on their
> inboxes. For that reason I don't want to go over the same issue we had
> asked before.
>
> The use of libaio is optional anyways and the system works as
> expected. what also covers other questions we had here on this thread.



--
Clebert Suconic
Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

dkulp

Question: If I grab Artemis 1.1.0 tarbal/zip and start up the broker “out of the box”, does it use libaio or not?  If I specifically have to configure something (pass a flag, edit a config file, etc…) to enable use if the LGPL library, then fine.    However, if it’s something that occurs completely automatically without the user even knowing that it’s occurring, then I have a major problem with it.  It needs to be something that the user has to explicitly CHOOSE to use.

Dan



> On Dec 23, 2015, at 2:02 PM, Clebert Suconic <[hidden email]> wrote:
>
> also, there has also been questions about it during the donation
> process.. licenses reviewed.. etc.. so I don't think we need to open a
> new discussions over this. the binary inclusion on the source was
> something that was fixed now.
>
> The dependency on libaio on the C code is through through dynamic
> linked library, and is the same as any C code depending on libc or
> gcc.
>
> On Wed, Dec 23, 2015 at 1:58 PM, Clebert Suconic
> <[hidden email]> wrote:
>> On Wed, Dec 23, 2015 at 1:55 PM, John D. Ament <[hidden email]> wrote:
>>> Just wondering, does anyone plan to raise the LGPL question w/ legal
>>> discuss?  If we're waiting for the new year to do the next release, would
>>> be good to at least start the discussion.
>>
>>
>> We had such discussion long ago with legal. I couldn't find that email
>> on my inbox but we specifically asked questions about it. We were ok
>> as I remember. Maybe someone else (Martyn?) will have it on their
>> inboxes. For that reason I don't want to go over the same issue we had
>> asked before.
>>
>> The use of libaio is optional anyways and the system works as
>> expected. what also covers other questions we had here on this thread.
>
>
>
> --
> Clebert Suconic

--
Daniel Kulp
[hidden email] - http://dankulp.com/blog
Talend Community Coder - http://coders.talend.com

Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

John D. Ament-2
Are you referring to the bin or src distribution?

On Wed, Dec 23, 2015 at 4:05 PM Daniel Kulp <[hidden email]> wrote:

>
> Question: If I grab Artemis 1.1.0 tarbal/zip and start up the broker “out
> of the box”, does it use libaio or not?  If I specifically have to
> configure something (pass a flag, edit a config file, etc…) to enable use
> if the LGPL library, then fine.    However, if it’s something that occurs
> completely automatically without the user even knowing that it’s occurring,
> then I have a major problem with it.  It needs to be something that the
> user has to explicitly CHOOSE to use.
>
> Dan
>
>
>
> > On Dec 23, 2015, at 2:02 PM, Clebert Suconic <[hidden email]>
> wrote:
> >
> > also, there has also been questions about it during the donation
> > process.. licenses reviewed.. etc.. so I don't think we need to open a
> > new discussions over this. the binary inclusion on the source was
> > something that was fixed now.
> >
> > The dependency on libaio on the C code is through through dynamic
> > linked library, and is the same as any C code depending on libc or
> > gcc.
> >
> > On Wed, Dec 23, 2015 at 1:58 PM, Clebert Suconic
> > <[hidden email]> wrote:
> >> On Wed, Dec 23, 2015 at 1:55 PM, John D. Ament <[hidden email]>
> wrote:
> >>> Just wondering, does anyone plan to raise the LGPL question w/ legal
> >>> discuss?  If we're waiting for the new year to do the next release,
> would
> >>> be good to at least start the discussion.
> >>
> >>
> >> We had such discussion long ago with legal. I couldn't find that email
> >> on my inbox but we specifically asked questions about it. We were ok
> >> as I remember. Maybe someone else (Martyn?) will have it on their
> >> inboxes. For that reason I don't want to go over the same issue we had
> >> asked before.
> >>
> >> The use of libaio is optional anyways and the system works as
> >> expected. what also covers other questions we had here on this thread.
> >
> >
> >
> > --
> > Clebert Suconic
>
> --
> Daniel Kulp
> [hidden email] - http://dankulp.com/blog
> Talend Community Coder - http://coders.talend.com
>
>
Reply | Threaded
Open this post in threaded view
|

Re: [VOTE] Apache Artemis 1.2.0

chirino
In reply to this post by dkulp
I think the binary distro uses the libaio.so if it's installed in your
system.  Since it's optional, the broker should still start up fine
even if libaio is not installed, but it wont get used either.

On Wed, Dec 23, 2015 at 4:05 PM, Daniel Kulp <[hidden email]> wrote:

>
> Question: If I grab Artemis 1.1.0 tarbal/zip and start up the broker “out of the box”, does it use libaio or not?  If I specifically have to configure something (pass a flag, edit a config file, etc…) to enable use if the LGPL library, then fine.    However, if it’s something that occurs completely automatically without the user even knowing that it’s occurring, then I have a major problem with it.  It needs to be something that the user has to explicitly CHOOSE to use.
>
> Dan
>
>
>
>> On Dec 23, 2015, at 2:02 PM, Clebert Suconic <[hidden email]> wrote:
>>
>> also, there has also been questions about it during the donation
>> process.. licenses reviewed.. etc.. so I don't think we need to open a
>> new discussions over this. the binary inclusion on the source was
>> something that was fixed now.
>>
>> The dependency on libaio on the C code is through through dynamic
>> linked library, and is the same as any C code depending on libc or
>> gcc.
>>
>> On Wed, Dec 23, 2015 at 1:58 PM, Clebert Suconic
>> <[hidden email]> wrote:
>>> On Wed, Dec 23, 2015 at 1:55 PM, John D. Ament <[hidden email]> wrote:
>>>> Just wondering, does anyone plan to raise the LGPL question w/ legal
>>>> discuss?  If we're waiting for the new year to do the next release, would
>>>> be good to at least start the discussion.
>>>
>>>
>>> We had such discussion long ago with legal. I couldn't find that email
>>> on my inbox but we specifically asked questions about it. We were ok
>>> as I remember. Maybe someone else (Martyn?) will have it on their
>>> inboxes. For that reason I don't want to go over the same issue we had
>>> asked before.
>>>
>>> The use of libaio is optional anyways and the system works as
>>> expected. what also covers other questions we had here on this thread.
>>
>>
>>
>> --
>> Clebert Suconic
>
> --
> Daniel Kulp
> [hidden email] - http://dankulp.com/blog
> Talend Community Coder - http://coders.talend.com
>



--
Hiram Chirino
Engineering | Red Hat, Inc.
[hidden email] | fusesource.com | redhat.com
skype: hiramchirino | twitter: @hiramchirino
1234