Mailing-List: contact dev-help@maven.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "Maven Developers List" <dev@maven.apache.org>
Content-Type: text/plain; charset=utf-8
Mime-Version: 1.0 (Mac OS X Mail 8.2 \(2102\))
Subject: Re: [VOTE] Maven 3.3.6
From: Jason van Zyl <jason@takari.io>
In-Reply-To: 
 <CAMYdmatKQTyzKGt969QVH+O5VkxUKW-62OVcE=9FVL5fzh8QTw@mail.gmail.com>
Date: Sun, 2 Aug 2015 22:09:53 -0400
Content-Transfer-Encoding: quoted-printable
Message-Id: <66A12A17-EFA2-4A01-BE27-9FCC60BD3393@takari.io>
References: <A7AA6DB0-9FC8-4E9B-AB63-D5CB8537E3CE@takari.io>
 <55BCC4B3.3060608@gmx.de>
 <CAMYdmastMgZx_npGRXV2=0VmYBxGmhRpmKDQxEBm9FUpGr=YYg@mail.gmail.com>
 <CAMYdmatN5fEAGAVi76fEwm4-GRfFtZd1vfainKkO34wYs_jhwQ@mail.gmail.com>
 <op.x2q5naklkdkhrr@robertscholte.dynamic.ziggo.nl>
 <CAMYdmatKQTyzKGt969QVH+O5VkxUKW-62OVcE=9FVL5fzh8QTw@mail.gmail.com>
To: Maven Developers List <dev@maven.apache.org>

I=E2=80=99m not sure a unit test really matters as it was a conscious =
change in a public signature which in and of itself is fine. If the =
signature changed then a unit test would have to change accordingly. But =
unless we have something akin to Clirr or a semantic version check then =
we don=E2=80=99t know how this will impact consumers of the public API =
which is the problem Karl-Heinz noticed with the enforcer. We have tests =
and they should have detected the issue. Our CI is setup is not helping =
us detect these issues. This is what I=E2=80=99m going to work on this =
week in a way I think is effective. I don=E2=80=99t think unit tests are =
of any help in this particular case but do anything you think will help.

I=E2=80=99ll cut the release next Monday.

> On Aug 2, 2015, at 5:46 PM, Anton Tanasenko <atg.sleepless@gmail.com> =
wrote:
>=20
> Sure thing, I'll update pull request.
> Thanks for reminding.
>=20
> 2015-08-02 21:00 GMT+03:00 Robert Scholte <rfscholte@apache.org>:
>=20
>> Even though the changes seem too simple, I'd like to see unittests to
>> prevent this from happening again.
>> I don't mind writing them.
>>=20
>> Robert
>>=20
>> Op Sat, 01 Aug 2015 17:05:37 +0200 schreef Anton Tanasenko <
>> atg.sleepless@gmail.com>:
>>=20
>>=20
>> I've created yet another PR for MNG-5805:
>>> https://github.com/apache/maven/pull/62
>>> Sorry guys, I was certain I had run m-enforcer-p its as well with =
3.3.5
>>> candidate.
>>>=20
>>> 2015-08-01 16:57 GMT+03:00 Anton Tanasenko =
<atg.sleepless@gmail.com>:
>>>=20
>>> Oh not again,
>>>> The LifecyclePhase#toString() method was added by me along with
>>>> toLegacyMap.
>>>> Looks like I didn't take care of 'mojos' field possibly being null. =
This
>>>> seems to be happening when an extension specifies a lifecycle phase =
but
>>>> doesn't assign any mojo executions to it.
>>>>=20
>>>>=20
>>>> 2015-08-01 16:08 GMT+03:00 Karl Heinz Marbaise <khmarbaise@gmx.de>:
>>>>=20
>>>> Hi,
>>>>>=20
>>>>> +0 from me...
>>>>>=20
>>>>> Reference Documentation updated =
(http://maven.apache.org/ref/3-LATEST/)
>>>>> ..
>>>>>=20
>>>>> checked with maven-invoker-plugin, maven-install-plugin, with =
Maven
>>>>> itself ;-)...
>>>>> and with some of my own projects without any issue...
>>>>>=20
>>>>>=20
>>>>> Unfortunately i have found an issue with maven-enforcer-plugin
>>>>> (trunk:r1693704) where two integrations test are failing with =
Maven
>>>>> 3.3.6
>>>>> which do not fail with Maven 3.3.3...
>>>>>=20
>>>>> I have attached both log files...
>>>>>=20
>>>>> Maven Enforcer accesses methods in maven-core which are marked as
>>>>> deprecated...(based on indirect changes):
>>>>>=20
>>>>> This has been added with 3.3.6...
>>>>>=20
>>>>>=20
>>>>> =
./maven-core/src/main/java/org/apache/maven/lifecycle/mapping/LifecyclePha=
se.java
>>>>>=20
>>>>>=20
>>>>> The following is called indirectly...
>>>>>=20
>>>>>    @Deprecated
>>>>>    public static Map<String, String> toLegacyMap( Map<String,
>>>>> LifecyclePhase> lifecyclePhases )
>>>>>    {
>>>>>        if ( lifecyclePhases =3D=3D null )
>>>>>        {
>>>>>            return null;
>>>>>        }
>>>>>=20
>>>>>        if ( lifecyclePhases.isEmpty() )
>>>>>        {
>>>>>            return Collections.emptyMap();
>>>>>        }
>>>>>=20
>>>>>        Map<String, String> phases =3D new LinkedHashMap<>();
>>>>>        for ( Map.Entry<String, LifecyclePhase> e:
>>>>> lifecyclePhases.entrySet() )
>>>>>        {
>>>>>            phases.put( e.getKey(), e.getValue().toString() );
>>>>>        }
>>>>>        return phases;
>>>>>    }
>>>>>=20
>>>>>=20
>>>>> So the culprit seemed to be this line:
>>>>>=20
>>>>>            phases.put( e.getKey(), e.getValue().toString() );
>>>>>=20
>>>>> where e.getValue() could be null...but they are not the problem.
>>>>>=20
>>>>> The problem is located in the toString() method of LifecyclePhase:
>>>>>=20
>>>>>   @Override
>>>>>    public String toString()
>>>>>    {
>>>>>        StringBuilder sb =3D new StringBuilder();
>>>>>        boolean first =3D true;
>>>>>        for ( LifecycleMojo mojo: getMojos() )
>>>>>        {
>>>>>            if ( first )
>>>>>            {
>>>>>                first =3D false;
>>>>>            }
>>>>>            else
>>>>>            {
>>>>>                sb.append( "," );
>>>>>            }
>>>>>            sb.append( mojo.getGoal() );
>>>>>        }
>>>>>        return sb.toString();
>>>>>    }
>>>>>=20
>>>>> The call to getMojos() is the real problem cause the call to
>>>>> "toLegacyMap()" is a static method without relationship to the =
object
>>>>> instance which could explain the non existings content of =
getMojos()...
>>>>>=20
>>>>> This looks to me that maven-enforcer needed to be fixed...But WDYT =
?
>>>>>=20
>>>>> Kind regards
>>>>> Karl Heinz Marbaise
>>>>>=20
>>>>>=20
>>>>> On 7/31/15 5:41 AM, Jason van Zyl wrote:
>>>>>=20
>>>>> Hi,
>>>>>>=20
>>>>>> Time to release Maven 3.3.6!
>>>>>>=20
>>>>>> Here is a link to the issues resolved:
>>>>>>=20
>>>>>>=20
>>>>>> =
https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=3D1231692=
2&version=3D12333058
>>>>>>=20
>>>>>> Staging repo:
>>>>>> https://repository.apache.org/content/repositories/maven-1201/
>>>>>>=20
>>>>>> The distributable binaries and sources for testing can be found =
here:
>>>>>>=20
>>>>>>=20
>>>>>> =
https://repository.apache.org/content/repositories/maven-1201/org/apache/m=
aven/apache-maven/3.3.6/
>>>>>>=20
>>>>>> Specifically the zip, tarball, and source archives can be found =
here:
>>>>>>=20
>>>>>>=20
>>>>>> =
https://repository.apache.org/content/repositories/maven-1201/org/apache/m=
aven/apache-maven/3.3.6/apache-maven-3.3.6-bin.zip
>>>>>>=20
>>>>>>=20
>>>>>> =
https://repository.apache.org/content/repositories/maven-1201/org/apache/m=
aven/apache-maven/3.3.6/apache-maven-3.3.6-bin.tar.gz
>>>>>>=20
>>>>>>=20
>>>>>> =
https://repository.apache.org/content/repositories/maven-1201/org/apache/m=
aven/apache-maven/3.3.6/apache-maven-3.3.6-src.zip
>>>>>>=20
>>>>>>=20
>>>>>> =
https://repository.apache.org/content/repositories/maven-1201/org/apache/m=
aven/apache-maven/3.3.6/apache-maven-3.3.6-src.tar.gz
>>>>>>=20
>>>>>> Source release checksum(s):
>>>>>> apache-maven-3.3.6-src.zip sha1:
>>>>>> ae409472561584c50691e672539b3eb0f11e806a
>>>>>>=20
>>>>>> Staging site:
>>>>>> http://people.apache.org/~jvanzyl/maven-3.3.6/
>>>>>>=20
>>>>>> Vote open for 72 hours.
>>>>>>=20
>>>>>> [ ] +1
>>>>>> [ ] +0
>>>>>> [ ] -1
>>>>>>=20
>>>>>> Thanks,
>>>>>>=20
>>>>>> The Maven Team
>>>>>>=20
>>>>>>=20
>>>>>=20
>>>>> Kind regards
>>>>> Karl Heinz Marbaise
>>>>>=20
>>>>>=20
>>>>>=20
>>>>>=20
>>>>> =
---------------------------------------------------------------------
>>>>> To unsubscribe, e-mail: dev-unsubscribe@maven.apache.org
>>>>> For additional commands, e-mail: dev-help@maven.apache.org
>>>>>=20
>>>>>=20
>>>>=20
>>>>=20
>>>> --
>>>> Regards,
>>>> Anton.
>>>>=20
>>>>=20
>>>=20
>>>=20
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscribe@maven.apache.org
>> For additional commands, e-mail: dev-help@maven.apache.org
>>=20
>>=20
>=20
>=20
> --=20
> Regards,
> Anton.

Thanks,

Jason

----------------------------------------------------------
Jason van Zyl
Founder, Takari and Apache Maven
http://twitter.com/jvanzyl
http://twitter.com/takari_io
---------------------------------------------------------

Lastly, "Impossible." The lamest of the lame excuses! Difficult maybe, =
or impractical, or too expensive, but rarely is anything impossible.

  -- Yvon Chouinard, Let my People Go Surfing


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@maven.apache.org
For additional commands, e-mail: dev-help@maven.apache.org