[RNG] Build succeeds locally but fails on Travis-CI

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

[RNG] Build succeeds locally but fails on Travis-CI

Gilles Sadowski
Hi.

Does someone know how to fix that?
See:
   https://travis-ci.org/apache/commons-rng/jobs/177715168

Thanks,
Gilles


---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [RNG] Build succeeds locally but fails on Travis-CI

Jörg Schaible
Gilles wrote:

> Hi.
>
> Does someone know how to fix that?
> See:
>    https://travis-ci.org/apache/commons-rng/jobs/177715168
>
> Thanks,
> Gilles


Delete your local repository (well, org/apache/commons/commons-rng) and
you'll face the same. You cannot declare a dependency for a plutin that does
not yet exist. Maven resolves the plugin's deps way earlier in the reactor
before the it calculates the build sequence for the subprojects.

And you cannot rely on the fact the the build tools are actually build
first. AFAICS the reactor will not take the dependencies to plugins into
account when it calculates the build sequence. The list of modules in the
parent has to be unique only, the sequence is based on the dependency
calculation. If you are using multiple threads to build, nobody will wait
for the build tools to complete.

- Jörg


---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [RNG] Build succeeds locally but fails on Travis-CI

Gilles Sadowski
On Tue, 22 Nov 2016 00:48:04 +0100, Jörg Schaible wrote:

> Gilles wrote:
>
>> Hi.
>>
>> Does someone know how to fix that?
>> See:
>>    https://travis-ci.org/apache/commons-rng/jobs/177715168
>>
>> Thanks,
>> Gilles
>
>
> Delete your local repository (well, org/apache/commons/commons-rng)
> and
> you'll face the same. You cannot declare a dependency for a plutin
> that does
> not yet exist. Maven resolves the plugin's deps way earlier in the
> reactor
> before the it calculates the build sequence for the subprojects.

What I've been asking about for weeks is whether this ("build-tools")
use of a module is a valid one.

As I understand it, all configurations for the reports have been
bundled in that "module".  Its use is _purely_ internal to the
maven build process.
So Artem's idea looked fine _iff_ maven can indeed handle such
"pre-dependencies".

If not, this "build-tools" module must disappear (as there is
nothing to release about it, once "mvn site" has run).

If this is the case, how should I go about to repair the build?


Thanks,
Gilles

> And you cannot rely on the fact the the build tools are actually
> build
> first. AFAICS the reactor will not take the dependencies to plugins
> into
> account when it calculates the build sequence. The list of modules in
> the
> parent has to be unique only, the sequence is based on the dependency
> calculation. If you are using multiple threads to build, nobody will
> wait
> for the build tools to complete.
>
> - Jörg
>


---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [RNG] Build succeeds locally but fails on Travis-CI

garydgregory
Why not put all this stuff in your parent POM?

Gary

On Mon, Nov 21, 2016 at 4:03 PM, Gilles <[hidden email]>
wrote:

> On Tue, 22 Nov 2016 00:48:04 +0100, Jörg Schaible wrote:
>
>> Gilles wrote:
>>
>> Hi.
>>>
>>> Does someone know how to fix that?
>>> See:
>>>    https://travis-ci.org/apache/commons-rng/jobs/177715168
>>>
>>> Thanks,
>>> Gilles
>>>
>>
>>
>> Delete your local repository (well, org/apache/commons/commons-rng) and
>> you'll face the same. You cannot declare a dependency for a plutin that
>> does
>> not yet exist. Maven resolves the plugin's deps way earlier in the reactor
>> before the it calculates the build sequence for the subprojects.
>>
>
> What I've been asking about for weeks is whether this ("build-tools")
> use of a module is a valid one.
>
> As I understand it, all configurations for the reports have been
> bundled in that "module".  Its use is _purely_ internal to the
> maven build process.
> So Artem's idea looked fine _iff_ maven can indeed handle such
> "pre-dependencies".
>
> If not, this "build-tools" module must disappear (as there is
> nothing to release about it, once "mvn site" has run).
>
> If this is the case, how should I go about to repair the build?
>
>
> Thanks,
> Gilles
>
>
> And you cannot rely on the fact the the build tools are actually build
>> first. AFAICS the reactor will not take the dependencies to plugins into
>> account when it calculates the build sequence. The list of modules in the
>> parent has to be unique only, the sequence is based on the dependency
>> calculation. If you are using multiple threads to build, nobody will wait
>> for the build tools to complete.
>>
>> - Jörg
>>
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>
>


--
E-Mail: [hidden email] | [hidden email]
Java Persistence with Hibernate, Second Edition
<https://www.amazon.com/gp/product/1617290459/ref=as_li_tl?ie=UTF8&camp=1789&creative=9325&creativeASIN=1617290459&linkCode=as2&tag=garygregory-20&linkId=cadb800f39946ec62ea2b1af9fe6a2b8>

<http:////ir-na.amazon-adsystem.com/e/ir?t=garygregory-20&l=am2&o=1&a=1617290459>
JUnit in Action, Second Edition
<https://www.amazon.com/gp/product/1935182021/ref=as_li_tl?ie=UTF8&camp=1789&creative=9325&creativeASIN=1935182021&linkCode=as2&tag=garygregory-20&linkId=31ecd1f6b6d1eaf8886ac902a24de418%22>

<http:////ir-na.amazon-adsystem.com/e/ir?t=garygregory-20&l=am2&o=1&a=1935182021>
Spring Batch in Action
<https://www.amazon.com/gp/product/1935182951/ref=as_li_tl?ie=UTF8&camp=1789&creative=9325&creativeASIN=1935182951&linkCode=%7B%7BlinkCode%7D%7D&tag=garygregory-20&linkId=%7B%7Blink_id%7D%7D%22%3ESpring+Batch+in+Action>
<http:////ir-na.amazon-adsystem.com/e/ir?t=garygregory-20&l=am2&o=1&a=1935182951>
Blog: http://garygregory.wordpress.com
Home: http://garygregory.com/
Tweet! http://twitter.com/GaryGregory
Reply | Threaded
Open this post in threaded view
|

Re: [RNG] Build succeeds locally but fails on Travis-CI

Jörg Schaible
Hi Gary,

Gary Gregory wrote:

> Why not put all this stuff in your parent POM?

Parent project is the last to build. IMHO we need an own independent
component for that. IIRC, those resources are not really commons-rng
specific and might be useful for other commons components also. We already
have build related components in commons (proper) and we always used lazy
consensus for a release of those (commons-parent, commons-build-plugin,
commons-skin, ...). commons-build seems abandonned anyway, as it is still a
Maven 1 project. Why not simply replace that one?

Cheers,
Jörg



---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [RNG] Build succeeds locally but fails on Travis-CI

Rob Tompkins
In reply to this post by Gilles Sadowski
I'll have to dig into this later, but at first look it seems that an environment variable is available to Jenkins that isn't in the Travis build.

-Rob

> On Nov 21, 2016, at 5:58 PM, Gilles <[hidden email]> wrote:
>
> Hi.
>
> Does someone know how to fix that?
> See:
>  https://travis-ci.org/apache/commons-rng/jobs/177715168
>
> Thanks,
> Gilles
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [RNG] Build succeeds locally but fails on Travis-CI

garydgregory
In reply to this post by Jörg Schaible
I was thinking the parent POM just had stuff that needed to be inherited
but since I have not looked it sounds like there is actual code in there?

I am about to push a new version of commons-build-plugin. Is that what you
are referring to?

Gary

On Mon, Nov 21, 2016 at 5:00 PM, Jörg Schaible <[hidden email]>
wrote:

> Hi Gary,
>
> Gary Gregory wrote:
>
> > Why not put all this stuff in your parent POM?
>
> Parent project is the last to build. IMHO we need an own independent
> component for that. IIRC, those resources are not really commons-rng
> specific and might be useful for other commons components also. We already
> have build related components in commons (proper) and we always used lazy
> consensus for a release of those (commons-parent, commons-build-plugin,
> commons-skin, ...). commons-build seems abandonned anyway, as it is still a
> Maven 1 project. Why not simply replace that one?
>
> Cheers,
> Jörg
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>
>


--
E-Mail: [hidden email] | [hidden email]
Java Persistence with Hibernate, Second Edition
<https://www.amazon.com/gp/product/1617290459/ref=as_li_tl?ie=UTF8&camp=1789&creative=9325&creativeASIN=1617290459&linkCode=as2&tag=garygregory-20&linkId=cadb800f39946ec62ea2b1af9fe6a2b8>

<http:////ir-na.amazon-adsystem.com/e/ir?t=garygregory-20&l=am2&o=1&a=1617290459>
JUnit in Action, Second Edition
<https://www.amazon.com/gp/product/1935182021/ref=as_li_tl?ie=UTF8&camp=1789&creative=9325&creativeASIN=1935182021&linkCode=as2&tag=garygregory-20&linkId=31ecd1f6b6d1eaf8886ac902a24de418%22>

<http:////ir-na.amazon-adsystem.com/e/ir?t=garygregory-20&l=am2&o=1&a=1935182021>
Spring Batch in Action
<https://www.amazon.com/gp/product/1935182951/ref=as_li_tl?ie=UTF8&camp=1789&creative=9325&creativeASIN=1935182951&linkCode=%7B%7BlinkCode%7D%7D&tag=garygregory-20&linkId=%7B%7Blink_id%7D%7D%22%3ESpring+Batch+in+Action>
<http:////ir-na.amazon-adsystem.com/e/ir?t=garygregory-20&l=am2&o=1&a=1935182951>
Blog: http://garygregory.wordpress.com
Home: http://garygregory.com/
Tweet! http://twitter.com/GaryGregory
Reply | Threaded
Open this post in threaded view
|

Re: [RNG] Build succeeds locally but fails on Travis-CI

Jörg Schaible-5
Hi Gary,

Gary Gregory wrote:

> I was thinking the parent POM just had stuff that needed to be inherited
> but since I have not looked it sounds like there is actual code in there?

Just ressources, but they are referenced in the plugins' classpaths.

> I am about to push a new version of commons-build-plugin. Is that what you
> are referring to?

Yes and no. Yes for the release procedure and no for the component itself.
Look into svn, there's also an old commons-build.

Cheers,
Jörg


---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [RNG] Build succeeds locally but fails on Travis-CI

sebb-2-2
Surely the commons build plugin is only needed to simplify setting up
some files?

If it were not there, the files could still be updated manually.

The doc page [1] makes it clear that the only _potential_ build-time
dependency is the copy-javadoc-files goal.
However AFAIK that is not used anywhere.

So it is not actually used at build time currently.

[1] http://commons.apache.org/proper/commons-build-plugin/index.html


On 22 November 2016 at 07:29, Jörg Schaible
<[hidden email]> wrote:

> Hi Gary,
>
> Gary Gregory wrote:
>
>> I was thinking the parent POM just had stuff that needed to be inherited
>> but since I have not looked it sounds like there is actual code in there?
>
> Just ressources, but they are referenced in the plugins' classpaths.
>
>> I am about to push a new version of commons-build-plugin. Is that what you
>> are referring to?
>
> Yes and no. Yes for the release procedure and no for the component itself.
> Look into svn, there's also an old commons-build.
>
> Cheers,
> Jörg
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: [RNG] Build succeeds locally but fails on Travis-CI

Jörg Schaible-5
Hi Sebb,

sebb wrote:

> Surely the commons build plugin is only needed to simplify setting up
> some files?
>
> If it were not there, the files could still be updated manually.
>
> The doc page [1] makes it clear that the only _potential_ build-time
> dependency is the copy-javadoc-files goal.
> However AFAIK that is not used anywhere.
>
> So it is not actually used at build time currently.
>
> [1] http://commons.apache.org/proper/commons-build-plugin/index.html

Actually it does not matter. It is deployed to central normally, so you
*could* use it at build time independently whether it makes currently sense
or not.

We could achieve the same with these configuration resources that are used
by the plugins in the build process of commons-rng and that are possibly
interesting for other components, too. Do we have shared (copied)
configuration files (currently for checkstyle, clirr, and pmd) between the
different components and do we want to share them physically?

Cheers,
Jörg


---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]