Re: What is `image-compute-scaling-factor' for?

[<tomas@tuxteam.de>]

[-- Attachment #1: Type: text/plain, Size: 1708 bytes --]

On Thu, Dec 24, 2020 at 01:52:21PM +0300, Evgeny Zajcev wrote:
> чт, 24 дек. 2020 г. в 13:37, Eli Zaretskii <eliz@gnu.org>:
> 
> > On December 24, 2020 11:30:57 AM GMT+02:00, Evgeny Zajcev <
> > lg.zevlg@gmail.com> wrote:
> > > What is this `image-compute-scaling-factor' function for?
> > >
> > > It operates with some insane "typical character" definition.  In old
> > > times
> > > this "typical character" thing might had some common sense, but
> > > nowadays
> > > with huge monitors and dense DPI, this "typical character" thing is
> > > insane
> > >
> > > What should `auto' image-scaling-factor *really* do?
> > >
> > > Thanks
> >
> > Doesn't the doc string of create-image (which does the scaling) answer
> > your question?
> >
> 
> Sorry, I don't understand it, why 10 pixels? In my case `auto' creates very
> large images (i.e. (frame-char-width) ==> 30), so scaling factor is 3.
> This creates very large images by default.
> 
> I know I can set `image-scaling-factor` to whatever I need.  However I want
> to understand the default behaviour decision, why 10 pixels is considered
> as "typical character" width?
> 
> I think, by default `create-image' should create images that look good when
> inserted into buffer, that is *not* what I get by default.

The way I read it (but I concur it's somewhat unclear to me, too) is
that the images are scaled proportionally to the (average) default
font size -- unless that is "too small", i.e. <= 10 pixels. In that
case I think (but am not sure) the images are shown as-is (i.e.
scaled by 1).

I too feel a bit confused by this explanation, but I still can't
quite put my finger on why.

Cheers
 - t

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 198 bytes --]