Re: :extend t inheritance

[Eli Zaretskii <eliz@gnu.org>]

> From: Ingo Lohmar <ingo.lohmar@posteo.net>
> Date: Wed, 23 Oct 2019 22:34:05 +0200
> 
> I am messing around with the face :extend attribute and I am wondering
> about sth.  The elisp manual says
> 
>      The value should be ‘t’ to display the empty space between end of
>      line and edge of the window using this face, or ‘nil’ to not use
>      this face for the space between the end of the line and the edge of
>      the window.  When Emacs merges several faces for displaying the
>      empty space beyond end of line, only those faces with ‘:extend’
>      non-‘nil’ will be merged.
> 
> I briefly read the original discussion of this feature, but did not find
> a reason for this peculiar behavior when it comes to inheritance; if
> there was a reason given, could someone point me to it?

I don't think I understand what you are saying or asking.  The text
that you cite doesn't mention inheritance at all.  Could you clarify
the issue?

> Background: I manage all faces without themes on my own, and I rely
> heavily on inheritance.  As it stands now, IIUC (and according to some
> tests), I have to specify :extend t on ~30 faces (mostly diff, ediff,
> smerge, magit-ediff and -blame etc) that all derive from 3 or 4 base
> faces.  It seems to me that this renders the whole point of inheritance
> moot.
> 
> My natural reaction would have been to specify :extend t on some base
> faces, and for faces that inherit from them, explicitly say :extend nil
> only for those faces for which I do not want that extension.  Isn't that
> at least as flexible, and more in the spirit of inheriting attributes?

Are you saying that the :extend attribute isn't inherited via
:inherit?  Or are you saying something else?  In any case, could you
please show a recipe or an example of what you tried and how the
results didn't match your expectations?

Thanks.