Re: follow-link in grep buffer

[Stefan Monnier <monnier@iro.umontreal.ca>]

>>> I have added a new option mouse-1-click-in-non-selected-windows
>>> that controls whether mouse-1 click in non-selected windows
>>> will follow links.  Default is t.
>> 
>> I think it should be purely and simply removed.
>> It addresses the "click to focus window" problem but nobody ever complained
>> about it (contrary to the problem of "click to focus frame" which is still
>> open).

> But that's what the x-mouse-click-focus-ignore-position variable is
> for.  And we also have focus-follows-mouse to have Emacs be able to do
> what it takes to give a frame focus.

Then let's use that.

> Anyway, I am complete against automatisms that get things right 70% of
> all the time immediately, with the user being unable to predict the
> behavior 90% of the time, so that he needs to check every time he uses
> the feature what actually happened.  If there is a braindead
> consistent rule that gets things right even only 40% of the time, but
> does not require switching on the brain or crosschecking for
> correcting the thing efficiently about 60% of the time, this is
> preferable in my book.

Agreed, which is why x-mouse-click-focus-ignore-position seems like the
right solution: it's very deterministic (doesn't depend on precisely where
you click, for instance) and safe and unsurprising since when it doesn't do
what the user intended, it just did a bit less than what she intended and
she can trivially get what she wants by clicking one more time.


        Stefan