Re: bug#45596: 12.2.4; Wrong DPI calculation for mixed-DPI multi-monitor setups

[Braun, Michael]

Ikumi:

I did not expect my two issues to intersect like this, but the discussion about frame dimensions and DPI seems relevant to an issue I had a couple of years ago. Preview-latex certainly has been working on Aquamacs (emacs 25), but not exactly perfectly.  I’m not much of a developer, but I thought you might find this experience of an Aquamacs end-user helpful.

The problem was that in Aquamacs, previews were not being generated with the correct aspect ratios and resolutions.  At the time it seemed like a Retina display problem, and a mailing list thread from 2016 indicated that Retina-specific issues would not be fixed on the AucTeX side because of GNU policies (it’s a long thread, but maybe start here?  https://lists.gnu.org/archive/html/auctex/2016-05/msg00021.html).

So I came up with a workaround on the Aquamacs side, such as this pull request, which seems related to the discussion about updating preview.el.in.

https://github.com/aquamacs-emacs/aquamacs-emacs/pull/208

But none of that helped with preview resolution, either because 96 DPI was hard-coded in,  and/or because AucTeX was using the pixel dimensions both to determine the resolution of the image (as generated from Ghostscript), *and* to determine how big the image should be on the screen.  And the problem looks worse on Retina displays because even though my 15in MBP’s display is physically capable of 2880 x 1800, the maximum MacOS-allowed resolution was 1920 x 1200.  I work at 1680 x 1050, and 1680 is exactly what (display-pixel-width) returns.

As another workaround, I defined preview-scale-function with the following, so I can use the same .emacs on my desktop and laptop.

(defun mb-preview-scale ()

  (if (< (display-pixel-width) 2500) .8 1)

  )

This was all back in 2020.  Since Aquamacs is in the process of being updated to Emacs 28, I put the issue off to the side (perhaps svg support would be an option?).  Maybe it’s time to revisit.

Michael

On Jun 15, 2022, at 11:47 AM, Ikumi Keita <ikumi@ikumi.que.jp> wrote:

[EXTERNAL SENDER]

Hi Tassilo,

Tassilo Horn <tsdh@gnu.org> writes:

Grzegorz Kowzan <grzegorz@kowzan.eu> writes:
Hi Grzegorz,

I guess the problem is that mm-size in the `frame-monitor-attributes'
return value differs from `display-mm-height' and `display-mm-width'.

1) Yes, `display-pixel/mm-width/height` functions are defined in terms
of Xlib calls, whereas `frame-monitor-attributes` calls Gdk functions
(if Emacs was compiled with Gtk support) and only as a fallback it
calls Xlib. This is why we have this inconsistency. The standard DPI
for Xorg server is 96 regardless of the actual value, so for a given
resolution and assumed DPI Xorg gives fake mm width and
height. Essentially, the current calculation of DPI in auctex looks
pointless to me...

Ok, I see.

With pure Gtk port we get actual physical pixel width/height and
actual mm width/height, so I guess when auctex detects pgtk port, it
can either ignore these values and hardcode DPI of 96 or make use of
the actual DPI. I suppose there is something to be said for not
changing auctex's behaviour and hardcoding the value, but I would
prefer to use the actual DPI at least as an option (see below).

That's what I did.  Essentially I've applied your patch with an
additional fboundp check for `frame-monitor-attributes'.  If it's
available (24.4), then your variant is used, if not, the previous
variant basically always saying 96 DPI is used.

Isn't it a good time to carry out this TODO in preview.el.in? ;-)
,----
| (defun preview-get-dpi ()
|   ;; TODO: Remove false-case when required emacs version is bumped to
|   ;; 24.4 or newer as this is the version where
|   ;; `frame-monitor-attributes' has been introduced.
|   (if (fboundp 'frame-monitor-attributes)
|       (let* ((monitor-attrs (frame-monitor-attributes))
|              (mm-dims (cdr (assoc 'mm-size monitor-attrs)))
|              (mm-width (nth 0 mm-dims))
|              (mm-height (nth 1 mm-dims))
|              (pixel-dims (cdddr (assoc 'geometry monitor-attrs)))
|              (pixel-width (nth 0 pixel-dims))
|              (pixel-height (nth 1 pixel-dims)))
|         (cons (/ (* 25.4 pixel-width) mm-width)
|               (/ (* 25.4 pixel-height) mm-height)))
|     (cons (/ (* 25.4 (display-pixel-width))
|              (display-mm-width))
|           (/ (* 25.4 (display-pixel-height))
|              (display-mm-height)))))
`----

And I'd like to comment on two minor aspects.
1. The function `cdddr' used above isn't yet available for emacs 25.1,
  so it should be replaced with `cl-cdddr'. I noticed this thanks to a
  byte compile log which Michael Braun sent me off list. Thank you,
  Michael! (And this means that preview-latex didn't work at all for
  emacs 25.1 for this 1.5 years!)

2. Though bug#20171[1] is still open, I hope it has already been
  resolved as well with this bug#45596. If so, let's close it, too.

Best,
Ikumi Keita
#StandWithUkraine #StopWarInUkraine

[1] https://debbugs.gnu.org/cgi/bugreport.cgi?bug=20171