Thanks Dave.   My oversight.

I think Eric's argument is reasonable.  I understand the argument in principle, but how it plays out in an intensive, deep debugging session may be different.

So the rule now is that comments must not be _duplicated_ when methods are overridden.   The writer has two choices: a) empty comment (comment deferred up hierarchy) or b) comment specific to the overriding behavior.  By corollary, we should be stricter in our use of the @Override annotation (which I gather Eric is picking up on anyway).

I'm happy with the change (at least for now :-).

--Steve

On 03/05/2012, at 10:55 AM, David P Grove wrote:

Steve Blackburn <Steve.Blackburn@anu.edu.au> wrote on 05/02/2012 07:29:53 PM:

> From: Steve Blackburn <Steve.Blackburn@anu.edu.au>

> To: jikesrvm-core@lists.sourceforge.net,
> Date: 05/02/2012 07:31 PM
> Subject: Re: [rvm-core] [rvm-commits] jikesrvm: 13 new changesets
>
> Sorry for not noticing this change earlier.

>
> I don't understand the rationale for the removal of comments that
> are duplicated on overridden methods (ie method in subclass carries
> the same comment as method in superclass).   Sorry if I'm being
> stupid here, but this duplication seems helpful/necessary (at least
> with the Eclipse IDE) because the comments are ostensibly useful,
> and when I'm walking through a class it seems unhelpful to have them
> elided (even though I can push up through the hierarchy (sometimes
> deep) to find them.  

>
> To be clear, I'm all for standardization and application good
> practice, but it's just not immediately obvious to me that removing
> these comments is a good thing.  (I'm also very appreciative of
> Eric's substantial effort in pushing these changes through).

>
> Can someone enlighten me?

>

Hi Steve,

The motivation was discussed mid-April on this thread: http://sourceforge.net/mailarchive/message.php?msg_id=29121555

Summarizing, as long as the methods are tagged with @Override then there's an argument that avoiding the duplicated comments is better, since it avoids inconsistent maintenance operations that result in unintended differences between comments on sub/superclass methods.

--dave

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/_______________________________________________
Jikesrvm-core mailing list
Jikesrvm-core@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jikesrvm-core