Re: flyspell + longlines: hang/wait

[David Reitter]

On 30 Oct 2005, at 01:30, Miles Bader wrote:
> The "opposition" such as it is, seems just as religious in its zeal.

Recently, we have arrived at a few good non-religious arguments.

On the "pro DTWW" side, there's speed, line-number referencing  
(compiler errors e.g.), collaborative editing with others (who prefer  
different line length, but need to read/edit my text), variable needs  
in terms of frame width (when we use  the most common form of system- 
level UI, a multi-window desktop).

On the "no DTWW" side, RMS has made a very good point: GNU tools are  
based on the assumption of relatively short lines. Display-time word  
wrapping does away with single line-feeds as end-of-line markers, and  
something like the output of "grep" isn't too useful any more.

I for my part, don't use grep or more on my LaTeX files very often.  
But that'll be different for different people, and I respect that.
I hope that eventually, tools like grep, wc and more will move on,  
recognizing that there isn't one standard 65 or 80 character wide  
terminal any more...

> As a _user_, I generally support display-time word-wrapping (call it
> "DTWW"), because have to deal with editing such text too, and
> traditionally it's been a pain to do that with Emacs (though
> longlines-mode has done a great job with everything I've thrown at it
> so far).

I see longlines as a good workaround. The bugs with variable width  
fonts should be fixed and some essential things should be  
reimplemented in C (sorry, can't do it, otherwise I'd prepare a patch  
today).

> As an Emacs hacker, though, my concern with DTWW is what interaction
> it would have with the assumptions made by lisp code; supporters seem
> to generally assume that the only modification required will be some
> support in `line-move' to make C-n/C-p move by physical lines instead
> of logical lines when DTWW is turned on in a buffer.

You'll never know until you've tried it out.