Re: screen-gamma not always effective?

[Philip Rooke]

Richard Stallman <address@hidden> writes:

> In principle it would be more correct to have such code, but in
> practice it does not really matter, because the normal use of
> screen-gamma is that you find out the right value

I agree and that is my own use case.

> and set it once and for all.

I agree, but what I have failed to do today is work out how to set it
once and for all.  (Please accept the fact I'm new to this stuff).

I have now tried a number of methods:

- creating a stub ~/.emacs containing just

  (setq default-frame-alist '((screen-gamma . 1.0)))

- creating a stub ~/.emacs containing just

  (setq default-frame-alist '((screen-gamma . 1.0)))
  (setq initial-frame-alist '((screen-gamma . 1.0)))

- setting both default-frame-alist and initial-frame-alist variables
  through the customize function into a simple stub ~/.emacs

- setting specific X resources in ~/.Xresources (and in all cases
  trying in all permutations, X server restarts, xrdb reloads etc) eg

  emacs.screenGamma: 1.0
  Emacs.ScreenGamma: 1.0
  emacs*screenGamma: 1.0
  ...

- setting the same X resources through command line options
  to a specific command line emacs invocation.

In all cases I am seeing the same effect: those parts of the screen
showing text are displayed in the gamma corrected colour, the other
parts of the screen remain remain in the brighter uncorrected colour.
The parts of the screen showing the brighter colour include parts of
the scroll bar (when displaying files longer than one screen-full of
text) and also the mini-buffer when entering text.