Cute, but leads to awkward diffs... in fact, the only reason I ever
care about style is for diffs.
Andrus is as usual probably right, though.
On 3/3/06, Todd O'Bryan <toddobrya..ac.com> wrote:
> As a non-committer who's just watching the discussion...
>
> I think the Eclipse CheckStyle plugin will automatically reformat
> code. You could come up with a standard for the project and each
> developer could change the code to their personal preferences on
> checkout and back to the project format on check-in. Best of both
> worlds.
>
> Todd
>
> On Mar 3, 2006, at 2:18 AM, Andrus Adamchik wrote:
>
> > I think arguing about formatting and code style is opening a huge
> > can of worms. I think Eclipse formatting preferences strike a nice
> > balance, although even that can be contentious and lead to a
> > lengthy discussion of wether a "\n" is needed before "{" or not ;-)
> >
> > For instance I'd code Kevin's example like this:
> >
> > return (whatever) ? something : anotherthing;
> >
> > But this doesn't mean that we need to follow one or the other style
> > religiously. I don't care, as long as the code works. I am neutral
> > in regards to the checkstyle addition to the build script. I am
> > just afraid that we'll get bogged down in the discussion of what
> > the right style is and never accomplish anything of value.
> >
> > Andrus
> >
> >
> >
> > On Mar 2, 2006, at 10:16 PM, Cris Daniluk wrote:
> >
> >> Apache has some style guides, I think.... I recall not liking them
> >> much, though :)
> >>
> >> Are we bound to them? Obviously this is low priority stuff, but now
> >> that we have a lot more people actively participating, it does make
> >> sense to at least have going forward standards, kinda like what I'm
> >> working on for docs.
> >>
> >> On 3/2/06, Kevin Menard <kmenar..ervprise.com> wrote:
> >>> I guess that would require us to standardize a coding style. I
> >>> know we
> >>> have the eclipse prefs, but I think Checkstyle gives you a lot more
> >>> control.
> >>>
> >>> So, while I'm not ready to write all of these out, I was
> >>> wondering what
> >>> the concensus on something like the following is:
> >>>
> >>> public Thing blah()
> >>> {
> >>> if (whatever)
> >>> {
> >>> return something;
> >>> }
> >>> else
> >>> {
> >>> return anotherthing;
> >>> }
> >>> }
> >>>
> >>> I personally feel the else should be removed. But, that's because I
> >>> normally get fooled by such things. In my mind, it's easier to
> >>> consider
> >>> whatever to be a special case and thus it goes in the if() while
> >>> the else
> >>> body is the regular case (and there's no need to assert it). I'm
> >>> asking
> >>> because there's a lot of cases similar to this in the code. If
> >>> others
> >>> agree with me, I'll go through and clean it up. Otherwise, I can
> >>> cope
> >>> with the existing style.
> >>>
> >>> --
> >>> Kevin
> >>>
> >>>
> >>> On Thu, 02 Mar 2006 12:45:10 -0500, Cris Daniluk
> >>> <cris.danilu..mail.com>
> >>> wrote:
> >>>
> >>>> One thing that may be helpful is to setup PMD and Checkstyle...
> >>>> both
> >>>> require extensive configuration to eliminate annoying false
> >>>> positives,
> >>>> but when tuned, they usually present helpful information.
> >>>> Obviously no
> >>>> tool is perfect, but for us, about 25% of the reported PMD issues
> >>>> deserved attention, and 75% of the checkstyle issues did.
> >>>>
> >>>> Maybe Findbugs would be a better choice than PMD... either way,
> >>>> I've
> >>>> used PMD more for analytics, and checkstyle more for pure
> >>>> style/formatting myself.
> >>>>
> >>>> Ideas?
> >>>>
> >>>> Cris
> >>>
> >>>
> >>
> >
>
>
This archive was generated by hypermail 2.0.0 : Fri Mar 03 2006 - 09:01:46 EST