camel-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Babak Vahdat <babak.vah...@swissonline.ch>
Subject Re: [HEADS UP] Bigger changes in parent/pom.xml
Date Fri, 07 Sep 2012 17:37:12 GMT
That sounds just great :-)

Babak

Am 07.09.12 18:05 schrieb "Christian Müller" unter
<christian.mueller@gmail.com>:

>I polished/cleaned the maven-surefire-plugin definitions. At the moment, I
>run a full test to check whether I do not break anything. Will commit the
>change after the test was running successful.
>Afterwards I will have a look at all the other duplicated plugin
>definitions, version definitions, ...
>
>Best,
>Christian
>
>On Fri, Sep 7, 2012 at 2:29 PM, Babak Vahdat
><babak.vahdat@swissonline.ch>wrote:
>
>> This is for sure a step in the right direction, as IMHO the Camel's
>>maven
>> setup needs pretty a lot of face-lifting (parts of them already being
>> mentioned by you). Some other points comming into my mind are:
>>
>> - There are places where we repeat ourselves again and again, for
>>example
>> because of derby.log while running unit-test:
>>
>>           <systemProperties>
>>             <property>
>>               <name>derby.stream.error.file</name>
>>               <value>target/derby.log</value>
>>             </property>
>>           </systemProperties>
>>
>> Which we could better say just ONCE inside parent POM using inheritance
>>of
>> pluginManagement.
>>
>> - Get rid of all those hard-coded version values being used in
>>components
>> and better extract them all up to the parent pom so that upgrading to
>>the
>> latest & greatest third-party can go smoother, as not too many people
>>would
>> go into components/came-xyz/pom.xml to check if there's any dependency
>>we
>> could / should upgrade.
>>
>> - If possible, it would be great to leverage a set of checkstyle rules
>>for
>> the POMs themselves (maybe there're already some ASL software out there
>>for
>> these kinds of stuff) as we do already today for the Java source (the
>> "sourcecheck" profile), then we could keep on a unique formatting of the
>> POMs such as:
>>
>> - No tab inside POM
>> - Max line length of XXX chars
>> - Indention using X spaces
>> - etc.
>>
>> - Also IMHO we should better get rid of ALL those
>> <exclude>**/XXXTest.*</exclude>:
>>
>>                 <artifactId>maven-surefire-plugin</artifactId>
>>                 <configuration>
>>                     <forkMode>pertest</forkMode>
>>
>> <forkedProcessTimeoutInSeconds>300</forkedProcessTimeoutInSeconds>
>>                     <excludes>
>>
>>                         <exclude>**/XXXTest.*</exclude>
>>                     </excludes>
>>                 </configuration>
>>
>> as sticking to @Ignore Annotation (JUnit) or @Test(enabled=false)
>>(TestNG)
>> is much more conventional. And frankly we would then spot the tests
>>being
>> skipped much easier.
>>
>> Last but not least many thanks for looking into this :-)
>>
>> Babak
>>
>>
>>
>>
>> --
>> View this message in context:
>> 
>>http://camel.465427.n5.nabble.com/HEADS-UP-Bigger-changes-in-parent-pom-x
>>ml-tp5718769p5718776.html
>> Sent from the Camel Development mailing list archive at Nabble.com.
>>
>
>
>
>--



Mime
View raw message