Re: looking for some volunteer

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

Re: looking for some volunteer

Tibor Digana
Hi Dejan,

Good to hear. In our Jira are two issues related to my original email:
https://issues.apache.org/jira/browse/SUREFIRE-1654
https://issues.apache.org/jira/browse/SUREFIRE-1494
If you have any questions, feel free to ask here.
Pls run the build locally (mvn -P run-its install) until our TravisCI build
would be ready for you. The build takes cca 1 hour to complete.
We should solve these two issue in two separate pull requests.
I guess these tasks require very good preparation and analysis of ITs
before making any changes.

@Enrico can you pls investigate TravisCI, why it still behaves so much
differently from Jenkins CI and fails?
Is it really due to the deployed SNAPSHOT versions at ASF Nexus?

Thx

Cheers
Tibor


On Sun, Mar 24, 2019 at 3:23 PM Dejan Stojadinovic <[hidden email]>
wrote:

> Hi Tibor,
>
> I volunteer for this. I have a solid experience with maven usage and also
> contributed few (easy) maven commits:
> https://github.com/apache/maven/commits?author=dejan2609
>
> Regards,
> Dejan Stojadinović
>
> On 2019/03/23 21:47:04, Tibor Digana <[hidden email]> wrote:
> > It's going to be very pedant work for someone who want to help us in
> > Surefire.
> >
> > I am looking for some volunteer who will remove the deprecated config
> param
> > `forkMode` in favor of `forkCount`. All our ITs should use `forkCount`
> > since then.
> >
> > Additionally, the volunteer should deprecate `surefire-junit4` provider.
> > Instead `surefire-junit47` provider should be selected as default
> provider
> > for JUnit4 tests. Again the ITs should be changed and plugin dependencies
> > should use `surefire-junit4` provider in place where it was not
> specified.
> >
> > It's easy to do. When you see the usages of `.forkMode()` method - only
> 20
> > in `surefire-its` and 43 usages of `<forkMode>` in
> > `surefire-its/src/test/resources`.
> >
> > Successful build must prove that the changes have been applied correctly.
> > The changes should be in a pull request on GitHub.
> >
> > Cheers
> > Tibor
> >
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>
>
Reply | Threaded
Open this post in threaded view
|

Re: looking for some volunteer

Tibor Digana
Hi Dejan,

Take your time.
We don't need it in days. We are talking about weeks/months.
I know that we will make code review and the development period would be
prolonged but that's expected.
What is a little bit more important than the time is the quality.

Cheers
Tibor

On Tue, Mar 26, 2019 at 9:24 AM Dejan Stojadinovic <[hidden email]>
wrote:

> Thanx Tibor, I left short comments on JIRA issues (just to mark them).
>  I hope I will squeeze first github PR in a few days.
>
> Regards,
> Dejan
>
> On 2019/03/24 23:19:08, Tibor Digana <[hidden email]> wrote:
> > Hi Dejan,
> >
> > Good to hear. In our Jira are two issues related to my original email:
> > https://issues.apache.org/jira/browse/SUREFIRE-1654
> > https://issues.apache.org/jira/browse/SUREFIRE-1494
> > If you have any questions, feel free to ask here.
> > Pls run the build locally (mvn -P run-its install) until our TravisCI
> build
> > would be ready for you. The build takes cca 1 hour to complete.
> > We should solve these two issue in two separate pull requests.
> > I guess these tasks require very good preparation and analysis of ITs
> > before making any changes.
> >
> > @Enrico can you pls investigate TravisCI, why it still behaves so much
> > differently from Jenkins CI and fails?
> > Is it really due to the deployed SNAPSHOT versions at ASF Nexus?
> >
> > Thx
> >
> > Cheers
> > Tibor
> >
> >
> > On Sun, Mar 24, 2019 at 3:23 PM Dejan Stojadinovic <[hidden email]>
> > wrote:
> >
> > > Hi Tibor,
> > >
> > > I volunteer for this. I have a solid experience with maven usage and
> also
> > > contributed few (easy) maven commits:
> > > https://github.com/apache/maven/commits?author=dejan2609
> > >
> > > Regards,
> > > Dejan Stojadinović
> > >
> > > On 2019/03/23 21:47:04, Tibor Digana <[hidden email]> wrote:
> > > > It's going to be very pedant work for someone who want to help us in
> > > > Surefire.
> > > >
> > > > I am looking for some volunteer who will remove the deprecated config
> > > param
> > > > `forkMode` in favor of `forkCount`. All our ITs should use
> `forkCount`
> > > > since then.
> > > >
> > > > Additionally, the volunteer should deprecate `surefire-junit4`
> provider.
> > > > Instead `surefire-junit47` provider should be selected as default
> > > provider
> > > > for JUnit4 tests. Again the ITs should be changed and plugin
> dependencies
> > > > should use `surefire-junit4` provider in place where it was not
> > > specified.
> > > >
> > > > It's easy to do. When you see the usages of `.forkMode()` method -
> only
> > > 20
> > > > in `surefire-its` and 43 usages of `<forkMode>` in
> > > > `surefire-its/src/test/resources`.
> > > >
> > > > Successful build must prove that the changes have been applied
> correctly.
> > > > The changes should be in a pull request on GitHub.
> > > >
> > > > Cheers
> > > > Tibor
> > > >
> > >
> > > ---------------------------------------------------------------------
> > > 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: Surefire CNFE

mgainty
In reply to this post by Tibor Digana
CNFE when running surefire means
Isolated (TEST) Classloader cannot load the dependency containing that class
In your case that dependency is junit

if it was up to me i would discontinue use of Isolated Classloader in favour of using system classloader
the only reason Isolated cl was used was the path limitation for System ClassLoader (initially 4k)
now all paths including CLASSPATH can be 32k so restricted classpath length should not be an issue

ymmv
martin-


Reply | Threaded
Open this post in threaded view
|

Re: Surefire CNFE

Tibor Digana
I don't know what Jira issue you are refferring to.
The ICL is used due to user's filtering and the multiple fork VMs.

On Fri, Aug 23, 2019 at 12:40 AM Martin Gainty <[hidden email]> wrote:

> CNFE when running surefire means
> Isolated (TEST) Classloader cannot load the dependency containing that
> class
> In your case that dependency is junit
>
> if it was up to me i would discontinue use of Isolated Classloader in
> favour of using system classloader
> the only reason Isolated cl was used was the path limitation for System
> ClassLoader (initially 4k)
> now all paths including CLASSPATH can be 32k so restricted classpath
> length should not be an issue
>
> ymmv
> martin-
>
>
>