R4481 Checkstyle Parameter Rename Issue

[Brent Putman]

On 9/14/16 12:19 PM, Cantor, Scott wrote:
> On 9/14/16, 12:18 PM, "dev on behalf of Tom Zeller" <dev-bounces at shibboleth.net on behalf of tzeller at dragonacea.biz> wrote:
>
>     
>>    > I don't think we compile with debug info?
>     
>>    Turns out we do (set maven-compiler-plugin debug to true in the parent-v3 POM).
> Oh. Dunno if that's normal.

Ah.  That was the other thing I was meaning to ask and/or look into re:
the c: namespace usage of param names. When I first saw that, I was
mildly surprised that that even worked.  I always kind of thought that
ctor and method parameter *names* weren't included in the byte code by
default. You only got them with debug turned on.  That may not be
correct at all.  However, I wonder if the c:name ctor stuff still works
if debug is turned off.  If not, then that's another reason why that
seems unwise to me personally.

Aside from that: Looks like debug is also turned on in the v2 parent.
So I guess we've been doing this as SOP for a long time.  I wasn't
aware of this.  Does anyone remember why we did that?  I just wonder if
we are incurring any bytecode size or performance penalties, etc for
building with debug=true.

Just turning it off, though, might be risky, obviously. So not saying
we should anytime soon.  Just wondering about the reasons, and tradeoffs.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://shibboleth.net/pipermail/dev/attachments/20160914/1bb6e46f/attachment.html>