On Wed, Sep 16, 2020 at 11:30:18AM +0100, João Távora wrote:
> On Wed, Sep 16, 2020 at 11:20 AM <tomas@tuxteam.de> wrote:
> 
> > On Wed, Sep 16, 2020 at 10:54:43AM +0100, João Távora wrote:
> >
> > [...]
> >
> > > Good point: these are opinionated things by nature, so they
> > > should be named as themes are, somewhat freely according
> > > to the vision of their authors. This also reveals that the
> > > thing shouldn't be called a "mode" at all. It should be done
> > > with custom-themes and if that proves a limitation, then
> > > custom-themes have to be improved, maybe even reinvented.
> >
> > We can set up our own giant feedback loops, too :-)
> >
> 
> Not sure what you mean, I just meant the "custom-themes"
> infrastructure should be enough to accommodate enough of
> the proposed "modern-mode".  Not sure if it is (as I don't use it).

Sorry -- was a lame self-reference. I characterised at some
point that whatever is currently considered "ergonomic" is
but part of a giant feedback loop involving (among many other
things, like genuine UI research) big corp's marketing departments.

What you wrote offered itself as another kind of feedback
loop: opinionated "custom-themes" (or whatever we end up
naming that) informing the Emacs core about whatever tech
or concepts are necessary to keep the whole ship together.

> I think reasonably solutions with a lot of value and relatively
> little code are often in front of our eyes. Such was the case
> with icomplete being a good basis for fido-mode, which seems
> good enough that people are even recommending it. I'm
> almost always wary of giants or grand reinventions of things.
> For the "base" Emacs experience that is, in their setups people
> can use all the ivys, dooms, helms and magits they want.

So Emacs has been doing that all along, it seems... And still,
this kind of discussions seem necessary and fruitful.

Cheers
 - t