Re: face-remap.el patch to resize window

[Miles Bader <miles@gnu.org>]

"Drew Adams" <drew.adams@oracle.com> writes:
>> It could be useful for very specific cases,
>> but it seems somehow wrong to have it as a global setting.
>
> No reason given? What good to others is a judgment with no supporting
> rationale, arguments or evidence?

The main problem is that I think it often depends on the buffer contents
(or etc) as to whether such window-adaptation is useful.

For instance, I shrink text for several reasons:

  1) To see more of the buffer at once; in this case, I probably don't
     want the window to shrink.  This is probably the most common case
     for me.

  2) To shrink a short buffer to take up less space, e.g. a help buffer;
     in this case, I often do want the window to shrink, or even better,
     to be re-fit to the buffer contents (like the help display
     functions do initially).

  3) To shrink a buffer that I want to keep on screen so I can keep an
     eye on it, but in many cases, only really care about the last
     bit -- e.g., a compile buffer, or log buffer which is being
     updated.  Often my goal is to make the window showing this buffer
     as small as possible or a little larger.  [and I often shrink the
     font to some absurdly small size, since I care more about motion
     and text-shape than actual contents]

So, really I want the behavior to vary depending on the situation; a
single "adapt windows" or "don't adapt windows" doesn't really suffice
for my usage.  However, a "clever" function that encodes my preferences
(as noted above) might be something I'd like to use.

That's why I like a hook-based situation (see earlier patch) that can
examine the situation and do something appropriate in an easily
user-adaptable way.

-Miles

-- 
P.S.  All information contained in the above letter is false,
      for reasons of military security.