Re: svn commit: r1730433 - /mina/site/trunk/content/mina-project/developer-guide.mdtext

[Jeff MAURY <je...@jeffmaury.com>]

Should we really promote skipping tests which is a bad practice ?

Jeff


On Mon, Feb 15, 2016 at 12:45 AM, <el...@apache.org> wrote:

> Author: elecharny
> Date: Sun Feb 14 23:45:54 2016
> New Revision: 1730433
>
> URL: http://svn.apache.org/viewvc?rev=1730433&view=rev
> Log:
> Fixed the developper guide
>
> Modified:
>     mina/site/trunk/content/mina-project/developer-guide.mdtext
>
> Modified: mina/site/trunk/content/mina-project/developer-guide.mdtext
> URL:
> http://svn.apache.org/viewvc/mina/site/trunk/content/mina-project/developer-guide.mdtext?rev=1730433&r1=1730432&r2=1730433&view=diff
>
> ==============================================================================
> --- mina/site/trunk/content/mina-project/developer-guide.mdtext (original)
> +++ mina/site/trunk/content/mina-project/developer-guide.mdtext Sun Feb 14
> 23:45:54 2016
> @@ -182,7 +182,7 @@ Answer to maven questions :
>  Then some other questions will be asked, about the next version to use.
> The default values should be fine.
>
>  <DIV class="info" markdown="1">
> -**Be Careful**
> +<strong>Be Careful<strong><br>
>
>  Make sure the change made by the release plugin is correct! (pom.xml,
> tags created)
>  </DIV>
> @@ -209,8 +209,10 @@ The first mail tells you that the SNAPSH
>
>  ### Step 4 : perform the release
>
> -The last step before launching a vote is to push the potential release to
> Nexus so that every user can test the created packages. Perform the
> following actions
> +The last step before launching a vote is to push the potential release to
> Nexus so that every user can test the created packages. Perform the
> following actions (note that we have to run a build first for teh javadoc
> to be correctly generated...) :
>
> +    $ mvn clean install -Pserial -DskipTests
> +    ...
>      $ mvn -Pserial,apache-release release:perform
>      ...
>      [INFO] [INFO]
> ------------------------------------------------------------------------
> @@ -250,11 +252,30 @@ Done !
>
>  ### Step 5 : closing the staging release on nexus
>
> -Now, you have to close the staged project on nexus. In order to do that
> you *must* have exported your PGP key to a PGP public server [see](
> http://www.apache.org/dev/openpgp.html)
> +Now, you have to close the staged project on nexus. In order to do that
> you **must** have exported your PGP key to a PGP public server [see](
> http://www.apache.org/dev/openpgp.html)
>
> -Connect to the Nexus server (https://repository.apache.org), login, and
> select the MINA staging repository you just created, then click on the
> 'close' button. You are home...
> +Connect to the [Nexus server](https://repository.apache.org), login, and
> select the MINA staging repository you just created, then click on the
> 'close' button. You are home...
>
>  ### Step 6 : Build the Site
> +You will need to modify the **pom.xml** file to be able to run the **mvn
> site** command. Actually, you have to comment the executions part of the
> maven JXF plugin in the **maven-site-plugin** configuration :
> +
> +    <plugin>
> +      <artifactId>maven-jxr-plugin</artifactId>
> +      <configuration>
> +        <aggregate>true</aggregate>
> +      </configuration>
> +
> +      <!--executions>
> +        <execution>
> +          <phase>install</phase>
> +          <goals>
> +            <goal>jxr</goal>
> +            <goal>test-jxr</goal>
> +          </goals>
> +        </execution>
> +      </executions -->
> +    </plugin>
> +
>
>      $ cd target/checkout
>      $ mvn -Pserial site
>
>
>


-- 
Jeff MAURY


"Legacy code" often differs from its suggested alternative by actually
working and scaling.
 - Bjarne Stroustrup

http://www.jeffmaury.com
http://riadiscuss.jeffmaury.com
http://www.twitter.com/jeffmaury

Re: svn commit: r1730433 - /mina/site/trunk/content/mina-project/developer-guide.mdtext

[Emmanuel Lécharny <el...@gmail.com>]

Le 15/02/16 09:45, Jeff MAURY a écrit :
> Should we really promote skipping tests which is a bad practice ?

This is not what I wrote. This is a very specific case, where we *have*
to run a clean install before running a perform, otherwise you have
failures. Tests have already been ran 2 times in the release process
(dry test and prepare). Here, I'm just skipping test to save the 3
minutes it takes.

What I would rather suggest is that we fix the problem that forces us to
run the clean install at this step...