On Apr 28, 2013, at 8:02 AM, Remko Popma wrote:

> Trivialities
> 
> 1) (note)
> In o.a.l.l.c.c.p.ResolverUtil line 279 
> I put the cast back in this line removing the cast broke the build on my 
> machine:
> 
> //Do not remove the cast on the next line as removing it will cause a compile 
> error on Java 7.
> final BundleWiring wiring = (BundleWiring) FrameworkUtil.getBundle(
>                 ResolverUtil.class).adapt(BundleWiring.class);

Indeed, when I switched to Java 7 it failed for me, too. Looks like they 
changed the API for that method ... not something they usually do. However, I 
think you may need to look into this further. The JavaDoc for OSGI for Java 7 
(http://www.osgi.org/javadoc/r5/core/org/osgi/framework/Bundle.html) clearly 
says the adapt(Class<A>) method does not return Object, it returns A (in this 
case, BundleWiring). This should not fail. There might be a problem with the 
bundle plugin. Perhaps there's a newer version, or for Java 7 it's using a beta 
version of the OSGI version.

> 
> 
> 2) (request)
> On my Japanese Windows 7, I have trouble building the site (mvn site).
> This seems to be caused by findbugs-maven-plugin encountering encoding issues.
> The problem goes away if I modify api/pom.xml to add <fork>false</fork> to 
> the findbugs configuration. I am guessing the <locales>en</locales> tag
> in the global pom.xml is not passed on if the process is forked.
> 
> Does anyone mind if I commit this change (adding <fork>false</fork>
> to the findbugs-maven-plugin configuration in api/pom.xml)?
> 
> 
> 3) (request)
> checkstyle.xml has a check "RegexpSingleline" that checks if a line has 
> trailing spaces.
> This is a bit embarrassing, but I cannot figure out how to modify the
> autoformatter on my IDE (Eclipse) so that it does not add spaces in javadoc
> comments like this:
> /**
>  * Some comment line
>  *                    <-- Eclipse auto-format adds a single space after the 
> '*'
>  * @param someParam param explanation
>  */
> 
> Checkstyle flags all these comments as problems.
> Can we consider these to be false positives and add an exception to this rule
> so that a space after a '*' character is not flagged as a trailing space?
> (I still need to figure out the regular expression for this.)
> 
> 

Attachment: smime.p7s
Description: S/MIME cryptographic signature

Reply via email to