svn commit: r437089 - in head/devel: . hhdate hhdate/files

[Ian Lepore]

On Wed, 2017-03-29 at 17:19 +0000, Alexey Dokuchaev wrote:
> On Tue, Mar 28, 2017 at 09:05:27PM -0500, Mark Linimon wrote:
> > 
> > On Tue, Mar 28, 2017 at 04:57:08PM +0200, Michael Gmelin wrote:
> > > 
> > > These are in to align to
> > > 
> > > TEST_TEST_TARGET=test, otherwise it would look like this:
> > > 
> > > TEST_USES=      compiler:c++11-lib
> > > TEST_TEST_TARGET=test
> > > 
> > > (replaced tabs by spaces for email)
> > > 
> > > Would this actually be preferred?
> > Keep in mind, "preferred" by danfe does not imply "preferred" by me,
> > or anyone else other than danfe.
> FWIW, I didn't use the word "preferred" at all in my reply, and yet
> Michael said it "makes sense".
> 
> > 
> > If we expanded the PH to cover all danfe's objections, it would be
> > 600 pages long.
> Mark, come on.  I'm not enforcing anything, and in fact you can find
> lots of examples of what I've said above already in the ports.  You
> make it sound like I'm trying to formalize whether one must or must not
> put a  after (len(knob) % 8) == 7) knobs, like OPTIONS_DEFAULT,
> which I'm certainly not.
> 
> Again, I'm just trying to keep things consistent, not enforce any rules.
> Also, as you've surely noticed, when I try to explain why things are
> (IMHO) better be done one way than another I *always* try to bring some
> rationale behind it, and I'm open for a discussion.  If people disagree,
> they are more than welcome to jump in and convince me otherwise.
> 

And right there is the problem:  The implicit assumption in your mind
(and believe me, it comes through loud and clear in every message you
post), that for something to be correct, YOU must be convinced it is
correct.

-- Ian

> I don't want to control ports' style(9), but I do want ports to be 1)
> consistent, and 2) pretty looking (as long as #1 holds).  You might
> argue that #2 is subjective, but history shows that most of the ports
> are (to my liking), and just some of them aren't.  So again (sorry),
> it's mostly about consistency, not my "preferred" way of doing things.
> 
> ./danfe
>