Re: Inaccuracy in NEWS.28

[Eli Zaretskii]

> From: Po Lu <luangruo@yahoo.com>
> Date: Tue, 01 Mar 2022 12:54:01 +0800
> 
> NEWS.28 says:
> 
>   Note that building with Cairo means using Pango instead of libXFT for
>   font support.  Since Pango 1.44 has removed support for bitmapped
>   fonts, this may require you to adjust your font settings.
> 
> But we don't use Pango anywhere except to parse GTK input method
> information and the output of the GTK font dialog.
> 
> Cairo text display is unrelated to Pango, and bitmapped fonts will work
> there as long as they are supported by FreeType.  Bitmapped fonts will
> also continue to work with the X server-side font driver, which is
> available in Cairo builds.
> 
> So would anyone mind me simply deleting that part of NEWS.28, since it
> is plain out wrong?

I'm not sure it's so plain wrong.  This text came out of prolonged
discussions we had back in January 2020 on emacs-devel, and it started
with this report:

  https://lists.gnu.org/archive/html/emacs-devel/2020-01/msg00583.html

Please read that discussion, and also the followups:

  https://lists.gnu.org/archive/html/emacs-devel/2020-01/msg00620.html
  https://lists.gnu.org/archive/html/emacs-devel/2020-01/msg00739.html
  https://lists.gnu.org/archive/html/emacs-devel/2020-01/msg00870.html

Maybe the description we now have in NEWS.28 is inaccurate, but the
problems were real.  So if you think mentioning Pango there is
incorrect (but please explain why, as I don't think I agree with your
reasoning: AFAIK Pango is the text layout engine used by Cairo), then
perhaps we should say something different there to convey the same
message.  And the message is still valid, AFAIK.