bug#40697: 25.3; sort-lines not accepting reverse parameter

[Noam Postavsky]

Roland Hughes <roland@logikalsolutions.com> writes:

> On 4/19/20 8:08 AM, Eli Zaretskii wrote:
>>> Cc: stepnem@gmail.com, 40697@debbugs.gnu.org
>>> From: Roland Hughes <roland@logikalsolutions.com>
>>> Date: Sat, 18 Apr 2020 14:48:19 -0500
>>>
>>> The wording in the help text is rather expert friendly. A user coming in
>>> cold to Emacs is going to interpret "prefix argument" as
>>>
>>> [GOLD]-7 REVERSE sort-lines [RETURN]
>> I can only say that I used the terminology and style that are used
>> everywhere in our doc strings.
>
> Yeah, I know. Not slamming you. Just pointing out the reason more
> people, especially from non-Linux platforms don't use Emacs is the
> "expert friendly" doc and culture around it. One must know all of the
> secret handshakes to join the club. New people, especially from a
> Windows world, aren't willing to do that. The GUI version of Emacs was
> a giant step in the right direction. Combined with EDT emulation it
> can eliminate needing to know most of the super secret handshakes.

Um, how many users "coming in cold to Emacs" will be coming from EDT, do
you think?  It seems to me that adding stuff to the docs about EDT style
args everywhere will only add more confusing "secret handshakes".

> Perhaps the "correct" solution here is for EDT or the sort-
> maintainers to add -reverse versions of each method so they are a
> single command with a common naming format requiring no super secret
> handshakes.
>
> [GOLD]-7 sort-lines [RETURN]
>
> [GOLD]-7 sort-lines-reverse [RETURN]
>
> same with numbers and a few of the others. They would just be a
> wrapper layer, but they would provide an easily digestible interface
> for those not looking to become one with the internals of Emacs or
> Linux.

Not sure where are any EDT maintainers at the moment, but I agree it
does sound useful for the emulation to be extended in this way.