Re: Android port of Emacs

[Alan Mackenzie]

Hello, Eli.

On Sat, Jun 24, 2023 at 09:54:09 +0300, Eli Zaretskii wrote:
> > From: Po Lu <luangruo@yahoo.com>
> > Cc: rms@gnu.org,  emacs-devel@gnu.org
> > Date: Sat, 24 Jun 2023 09:19:30 +0800

> > Eli Zaretskii <eliz@gnu.org> writes:

> > > This description doesn't fit the reality.  People who use the NS port
> > > are quite disappointed by the problems that don't get solved, and I as
> > > the maintainer cannot remain indifferent to their plight.  It breaks
> > > my heart that I can do almost nothing to facilitate the solution of
> > > those problems.  And mind you, no one thinks the NS problems are
> > > "unimportant", we are just unable to fix them, even though we all
> > > agree they are grave and should be fixed ASAP.

> > > So you might be indifferent to such problems, or consider them
> > > "unimportant", but IMO any responsible maintainer will not; I
> > > certainly am not.  And if we envision such problems to happen sooner
> > > rather than later to the Android port, we may wish to decide whether
> > > we want this now, before the danger becomes real -- as it became with
> > > other similar features.

> > I'm not indifferent to those problems with the NS port: I hope that some
> > day, they will be fixed.  However, Emacs development _should_ be
> > indifferent to them, as its goal is to develop a high quality text
> > editor for the GNU operating system, and not specifically others.

> > Which seems especially reasonable, considering that the reality of
> > volunteer-driven development is that every port will usually be stuck
> > with one maintainer, at least until that maintainer steps down.

> That's exactly the issue at hand, thank you.  I'd like very much to
> hear people speak up about these aspects, not anything else.

> Lars, Stefan, Robert, Martin, Mattias, Philip, Dmitry, João, and all
> the rest of frequent contributors and developers: where are you?  I'd
> love to hear your opinions on this.

Well, I'm not sure whether or not I count as a frequent contributor, but
I'll answer anyway.

I think you're right to be a bit concerned - how many people in Emacs
have even the slightest hacking experience on Android?  I think an answer
of one would be concerning, but it seems not unlikely to be the case.

Again, how stable is Android at the moment?  Again, I don't know, not
being the owner of an Android device, but my impression is "not very", at
least in the sense that GTK, Wayland, NS, and so on, also don't seem very
stable, more like "rapidly changing".  All of these versions seem likely
to be needing a lot of love for the forseeable future.

All due respect to Po Lu for what he's already done for Emacs (including
reporting lots of CC Mode bugs ;-), but that has been over a period of
only two years.  My impression is that the longer somebody has been an
Emacs contributor, the longer they're likely to remain a long-time
contributor.  We could really do with a statistician to support/refute
that notion.

At the same time, having Android support in Emacs core would clearly be
beneficial.  Particularly if large-scale software development is about to
start happening on Android.

On whether or not to include Android support in Emacs, I feel very unsure
- I can see the arguments both for and against, and feel unable to
evaluate them.  I suspect most contributors here likewise feel uncertain,
and that is the main reason so few have so far expressed a view.

If it would be practicable firstly to try out Android support separate
from the Emacs code, and then move it in the event of a positive
experience I would suggest that.  But that would cost at least as much
extra effort as simply having it separate.

As I said, I don't feel able to express a strong opinion on the matter,
but that's ducking the issue.  My gut feeling is that the arguments for a
separate Android Emacs version are stronger than those for integrated
support, but that's not a judgment I could really defend very strongly.

-- 
Alan Mackenzie (Nuremberg, Germany).