Re: [lmi] Switch to using C++11 uniform initialization in the ctor initializer lists?

[Greg Chicares]

On 2018-08-28 12:53, Vadim Zeitlin wrote:
> On Tue, 28 Aug 2018 11:58:33 +0000 Greg Chicares <address@hidden> wrote:
> 
> [...]
> GC> In 'tier_view_editor.hpp', I now think your version is better here:
> GC> 
> GC> inline tier_entity_adapter::tier_entity_adapter
> GC>     (std::vector<double>& limits
> GC>     ,std::vector<double>& values
> GC>     )
> GC>     :limits_ (&limits)
> GC>     ,values_ (&values)
> GC> 
> GC> because that code just initializes a member with a (non-const) reference
> GC> of the same type.
> 
>  FWIW I agree.

Okay, then that's one change in my patchset that must not
overwrite your preferable version. To leave nothing to
chance, I've pushed it as:

  
http://git.savannah.nongnu.org/cgit/lmi.git/commit/?h=odd/uniform_i12n&id=21f9e16b9b0bc17e8700add4cf8daae9842b65d2

> GC> I don't understand these files well enough to modify them extensively:
> GC>   database_view_editor.cpp
> GC>   multidimgrid*.?pp
> 
>  I don't claim any deep knowledge of them neither, but the changes to them
> seem to be trivial, so I'm confident enough in their correctness.

All right.

> GC> Therefore, if you're highly confident that the changes you've made to the
> GC> product-editor code are all correct, I'm willing to cherry-pick and push
> GC> your xanadu/uniform-init branch now...
> 
>  I could update my branch to remove the differences with odd/uniform_i12n
> concerning the base class ctor initializers, would you like me to do this?

Yes please.

> GC> >  I'd like to avoid doing white-space changes in the initializer list 
> right
> GC> > now as it risks conflicting with my other changes, so I'd prefer to 
> commit
> GC> > this one first, so that I could rebase over it and submit the PDF
> GC> > pagination changes too, that I've finally finished (this is the good 
> news;
> GC> 
> GC> ...and hold my odd/uniform_i12n branch back until you've submitted those
> GC> pagination changes.
> 
>  There should be no problems if you commit them now too, they are related
> and I think it'd be better to finish with this question once and for all.
Although I've commonly asked you for patches and then committed
them myself after review and testing, I think it might make more
sense to ask you to handle this. That way, you can merge our two
changesets together, or commit them separately, or do something
more wonderful than my limited git knowledge allows me to conceive;
and you can make certain that whatever you push doesn't clash with
any other changes you're planning for later. I've separately tested
your changeset and the combination of mine and yours, so I'm good.
Please just let me know when you're completely finished and the
final version has been pushed to savannah origin/master.