Re: [AUCTeX] Embedded previews in high-DPI?

[David Kastrup]

Mosè Giordano <address@hidden> writes:

> Dear all,
>
> I'm a bit confused about this question and I'd like to make it
> clearer, also because I don't have experience with the reported
> problem as I don't have a high-DPI screen.
>
> The point is that on certain systems, OS X combined with the so-called
> "Retina" display, preview images have wrong resolution, right?  If so,
> the proposed patch fixes this problem, in this sense I don't see it as
> a new feature that works only on proprietary systems, that is what FSF
> guidelines discourage, but as a bug fix on a proprietary system, that
> is definitely not forbidden.

No, it doesn't fix the problem of the resolution mismatch.  Instead it
reverts to using the image-io image type (which is _not_ provided by GNU
Emacs but rather by a third-party Mac Emacs port with additional
MacOSX-only support, like the Mac-only proprietary PDFKit library
<URL:https://developer.apple.com/library/mac/documentation/GraphicsImaging/Reference/PDFKit_Ref/index.html>)
in order to provide the images not as bitmaps but rather as PDF files to
be rendered by a proprietary renderer that is not supported by GNU Emacs
(exactly because it is MacOSX-only).

> On the other hand, one may argue that the patch adds a new feature,
> namely support for high-DPI screens,

It doesn't.  It merely exercises the proprietary PDF renderer of MacOSX
when using a MacOSX-only variant of Emacs not maintained by GNU.  That
has the _side-effect_ of hiding the resolution mismatch problems on the
Mac platforms because it stops rendering of bitmaps under the control of
either Emacs or preview-latex altogether.

> which is actually a feature that interests all operating systems,
> since one can install any operating system on a computer with high-DPI
> screen, including on a Macbook without invalidating hardware warranty.
> My question is: did anyone try to use preview-latex on a free
> operating system with a high-DPI screen?  Does it work as expected?
> If it works out-of-the-box, I see again the patch as a bug fix, not a
> new feature, I don't know how it could be seen otherwise in this case.

Have you actually looked at the code?  I don't think it makes sense
discussing it without knowing what it does.

-- 
David Kastrup