Re: Upcoming loss of usability of Emacs source files and Emacs.

["Stephen J. Turnbull" <stephen@xemacs.org>]

Eli Zaretskii writes:

 > I'm not talking about input methods here, I'm talking about
 > display.

OK, I see.  Again, Japanese kind of spoils you, since the JIS
standards incorporate a bunch of Latin characters as well as a few
non-Latin scripts (Greek and Cyrillic at least).  But I can't seem to
find curly quotes, so maybe I misremembered the display repertoire.

 > In any case, localized versions of Windows (or any OS) are a thing
 > of the past;

Not in Japan.  Windows is very much localized (for example, in
recognizing Japanese charsets in IE), although most of the components
are properly internationalized.

 > > It's not just quotation marks.  It's foreign words spelled correctly,
 > > ellipses, the occasional math symbol (the lemniscate (infinity) is
 > > popular among the more flaky of my correspondents), emoticons, and
 > > various other symbols (dice, playing card suit symbols, enclosed
 > > characters such as circled numerals).  There are too many of them to
 > > excuse with "smart quotes"; users are using "input methods" of some
 > > sort for these characters that don't exist on their keyboards.
 > 
 > Not IME.  When I type "1st", "2nd", etc., one particularly popular mailer,
 > which will remain unnamed, automatically makes the "st" and "nd" parts
 > rendered as superscripts.  Ellipses appear if I type "...", em-dashes
 > appear if I type "--", and "→" if I type "-->".  If I type "naive", it gets
 > converted to "naïve".  Emoticons appear automatically if I type their
 > ASCII-art equivalents.  Likewise with "(c)", "(tm)", "(r)", and "(e)" (I'll
 > let you guess what the last one gets me).  There are special
 > auto-conversions for math symbols, too numerous to mention.  Plus, I've
 > counted more than a dozen of other such automatic conversions that I can
 > turn on or off.  And -- no less important -- each auto-converted text gets
 > a small widget shown near it, which allows to undo that particular
 > conversion with a single mouse click.
 > 
 > Given all that, why would one need an input method, except when typing in
 > some complex script?

That *is* an input method, since it's not just remapping keystrokes to
different characters, but rather using sequences of keystrokes to
represent single characters.  That's precisely the way one inputs
Japanese, including the little widget.  OTOH, C-x 8 is just
paleolithic, it's the non-programmer's equivalent to C-q <octal>.

 > We in Emacs have yet a way to go until we get there.  Are there any
 > motivated volunteers reading this who'd like to provide something like
 > that in Emacs text modes?

I don't think we're really that far off.  Quail is pretty close.  I
think we need four things: (1) tables for these characters, (2) a way
to combine tables so that we can overlay quote or sub/superscript
translation on other translation tables and perhaps overlay AltGr-
style "native keyboard" methods (ie, non-Quail), (3) a way to
"decompose" a character back into a sequence of characters, and (4)
perhaps somewhat less "highlighting" of partial translations (eg, in
Latin input, Quail underlines a character in the process of being
composed -- I find that annoying).

I suppose we could also provide a little widget for (3), on displays
that can handle it.

 > > Let's try loosening up and find out, no?
 > 
 > Yes, let's.  But let's not forget about the inconveniences this will bring
 > to some, either.  Let's not dismiss them.  Let's provide graceful fallbacks
 > for them.  As we are doing now.  Let's not be too religious about these
 > changes.

 > >  > I just think that we shouldn't dismiss so easily the issues these
 > >  > changes bring with them.  IOW, we should see this issue in its
 > >  > delicate balance.
 > > 
 > > You're telling *me*?
 > 
 > Not just you, everyone who reads this list.  This ain't private mail, and
 > I'm not talking to you alone.

Point taken.  Nevertheless, in English your phrasing, taken in its
original context, strongly implies that *I* dismissed and *I'm*
unaware of the delicate balance.  I take exception to those
implications, and ask that you take the kind of care with them that
you request of me with respect to hyperbole.

 > > It's certainly true that I've engaged in hyperbole.
 > 
 > That's my sole point: don't.  Hyperboles don't help a bit here.

I'll try, but really, in adult conversation in the three languages I
speak well enough to recognize exaggeration, it is extremely frequent.
And I'm hardly the only offender on this list.  Like it or not,
readers are going to have to deal with it.

 > IOW, your mental model of what I think and do in these matters seems to be
 > flawed.

You're absolutely right.  I stand corrected, and I apologize.

 > My point is that we should make these steps carefully, one step at
 > a time, and provide fallbacks and ways to switch this off where
 > possible on every step.

Once again, your phrasing excludes me.  This is *our* point.

 > Which is what we've been doing.  You seem to be advocating to charge
 > ahead and never look back,

Not at all.  Surely you have noticed how often I've used the word
"experiment"!  It was I who risked insulting Paul by asking for an
explicit statement that the change would be backed out on request.  I
could go on.  If you haven't noticed that, maybe you should try
harder.

Sure, I've been emphatic, but mostly in response to self-described
reactionaries and other *very* conservative posters, who don't even
want to see experiments.

It's also true that I think that we should not display different
characters from what's in the buffer in programming modes (eg
displaying left single quotation mark when the buffer contains a grave
accent), and therefore I'm in favor of experimenting with curly quotes
in Lisp sources.