Preference to disable input completion?

[Jean-Marc Lasgouttes]

Le 01/12/2020 à 16:44, Richard Kimberly Heck a écrit :
>> No, the forces of FUD and reaction managed to stop progress (again).
> 
> I don't think we reached a conclusion here. The proposal was to enable
> it by default, but to set it to off for users migrating from 2.3.x who
> haven't enabled it (via prefs2prefs). Patch attached again.

It is a compromise, but better than nothing.

OTOH, we really should find a way to mitigate the completion mess. There 
is not reason it cannot be disabled. As far as preferences go, there is 
room for disabling many of them IMO:

	/// minimum length of words to complete
	unsigned int completion_minlength = 6;

	///
	bool completion_cursor_text = true;
	///
	double completion_inline_delay = 0.2;
	///
	bool completion_inline_math = true;
	///
	bool completion_inline_text = false;
	///
	int completion_inline_dots = -1;
	///
	double completion_popup_delay = 2.0;
	///
	bool completion_popup_math = true;
	///
	bool completion_popup_text = false;
	///
	bool completion_popup_after_complete = true;

Everything is configurable, except the important stuff!

Scott, I am not sure you still have this set of preference files 
contributed by users. It would be nice to know whether people use thoese 
settings. It seems that there are choices to be made. For example I do 
not know why it could be good to disable the |> cursor with 
completion_cursor_text=false.

JMarc