Daniel Kulp wrote:
> Question 1:
> Would people be willing to accept patches that just fix warnings that show 
> up in eclipse?
> 
> Right now, using the same settings I use for Celtix, there are 622 
> warnings in the spec, sdo, and sca projects.    I hate seeing warnings, 
> so I have two options:
> 
> 1) Turn off most of the eclipse warnings (with I think sucks)
> 2) Fix the code and submit patches
> 
> In general, I don't like submitting "warnings" fixes along with patches 
> that provide real code changes.   The real code changes tend to get lost.   
> However, I just want to check to see if people are interested in warnings 
> patches before I spend time on it.     (FYI: Celtix is completely 
> warnings free, which include a bunch of checkstyle and pmd rules which 
> I'm not applying to tuscany.   I have NO idea how bad tuscany would be if 
> those types of checks were enabled.   One step at a time.....)
> 

In general I think this kind of improvement is good but it may depend on
the severity warning and the fix. I get a lot in IDEA as well and every
so often get the urge and go in and fix some. I have been marking the
commit as "cosmetic" if it is not meant to change anything except to fix
formatting or codestyle warnings.

I agree that we should not mix "cosmetic" stuff with real code changes -
that just makes things more complex and by its nature "cosmetic" stuff
should not be urgent.

Please submit patches for this.

> 
> Question 2:
> In the sdo-api package, there are some annotations for SuppressWarnings 
> like:
> @SuppressWarnings({"ClassLoader2Instantiation"})
> and
> @SuppressWarnings({"AccessOfSystemProperties"})
> Those warning values aren't "standard" ones that javac recognizes nor does 
> the eclipse compiler.   I'm just curious as to what compiler/tool they 
> are targeted at.    (just a curiosity question, nothing more)
> 

They are added by IDEA when it detects checkstyle type issues. If it is
causing a problem we can just remove them.

Out of curiosity, does Eclipse have an equivalent?
--
Jeremy

Reply via email to