RE: Indenting text bugs.

["Drew Adams" <drew.adams@oracle.com>]

Hi Lennart,

> >> > Isn't that a very inconvenient default? Almost every window
> >> > manager across platforms intercept that to switch windows...
> >
> > No, they do not intercept `ESC TAB'.
> 
> But if you use a vi emulator in Emacs like Viper (which many uses)
> then you can not use ESC TAB since ESC is caught by the vi emulator.

Hey, if something else steals ESC, then sure, it's not available for this. That
goes without saying.

But with the normal ESC map, this is not a problem.

We're not, here, going to list every possible Emacs library that might modify
the ESC map are we?

If you bind `C-h f' to `forward-char', then it won't do `describe-function'
anymore. But that's not a reason not to inform help-gnu-emacs readers that `C-h
f' is normally `describe-function'. Same thing here.
 
> > But you need only do this to make Windows pass `M-TAB' 
> > along to Emacs:
> >
> > (when (fboundp 'w32-register-hot-key)
> >  (w32-register-hot-key [M-tab]))
> 
> And this does not always work (as I have told many, many times). To
> make it work correctly you need the patched version of Emacs+EmacsW32
> (and then the recipe is a bit different).

Yes, you've said that many, many times. But never (AFAIK) with any explanation
or demonstration. You've hinted that there might be some corner cases that are
problematic, but I've never heard anything more concrete.

What can I say? It's always worked for me. No doubt I just haven't run into
those corner cases.

I don't doubt that you are speaking the truth when you say you've run into a
problem, but if you aren't more specific than that, then saying that is not very
helpful.

Unless you provide some helpful info about it, this is like Chicken Little
screaming "The sky is falling!". That might scare some people away from trying
it, when trying it might actually be useful for them.

I understand that from your perspective this is a problem and Emacs+EmacsW32 is
the solution. But I've never experienced the problem, and AFAIK you have never
explained the problem.

How about a simple recipe starting with `emacs -Q' in unpatched, vanilla Emacs,
to demonstrate the problem? Then users can decide if it's something they need to
worry about.