Re: Motif support

[Arthur Miller <arthur.miller@live.com>]

[-- Attachment #1: Type: text/plain, Size: 3283 bytes --]

Madhu <enometh@meer.net> writes:

> * Arthur Miller <DB9PR09MB498654A6B8A8D5B84A9B7CC8967D9@DB9PR09MB4986.eurprd09.prod.outlook.com> :
> Wrote on Wed, 22 Dec 2021 16:58:18 +0100:
>
>> I feel said. Is there really anyone in the world who cares for Motif?
>>
>> Emacs C source is full of spaghetti-if-defs all over the place. It
>> would be nice if it was cleaned of some old stuff not cared by more
>> than 5 people.
>

Actually, I percieve that for freedom of software it is equally important to
have documentation for the sources as the sources themselves. Situation with
docs is somewhat addressed in last years or so, but there were times when many
GNU projects came out with very few comments, usually no dev docs, and scarce
descriptions about internals. I said several times to Eli Z. to write a book
about Emacs internals; that would probably be the best boost to Emacs
development.

As time goes by, projects are growing, Emacs too, and it takes more and more
time to understand it for people new to the source. If you have a family and
full time job, you might not have the time for all the detective work that is
sometimes needed. It can be unfortunate, if you know what to do to implement
something, but have to spend lot of time to investigate how Emacs does it.

The popular belief is that Emacs development is hold back by lack of "modern
development" in form of some public git repository and pr-modell as used on
github & co. I don't think that is what hold more people to contribute to
Emacs. There is SXEmacs, a fork of XEmacs, that has git workflow as the popular 
request is for GNU Emacs. They even have tools to make it easier to work with
SXEmacs itself, and they also made some populistic changes of the time, like
reformatted the code to kernel style and removed Windows port and implemented
FFI. Yet, they have no active devs seems like. What is the problem? I don't know,
but I would rather say the complexity and that it is hard and time consuming to
do deeper changes if you are not quite familiar with internals. It will be
interesting to see if there will be a flood of developers incoming after a
change to sourcehut or whatever is chosen for future development. I would be
happy if I am wrong of course.

> This is an example of one of the subtle ways in which a "freedoma
> narrative" is introduced to remove freedom from software, destroy value,
> and devalue the work of others by people who are not invested in the
> code at all.

When it comes to Motif, the fact is that it is a dead toolkit. It is alive just
so its devs can milk money out of some existing customers and old code bases,
but I can't imagine anyone clear in their minds to use it for a new
project. Motif hasn't been relevant for like 15+ years, maybe 20, I don't
know. I don't understand why Debian maintainers build Emacs with Motif, when
even Debian stable comes with Gtk, but it is their choice of course.

Just for the fun I have attached comparison, side by side of SXEmacs and GNU
Emacs, SXEmacs uses Motif to render it's gui and pure X I guess to render
fonts. In my eyes it looks very ugly, I don't think it is representable of what
GNU Emacs can do to give to anyone an Emacs built on Motif nowadays. (I don't
have a build of GNU Emacs with Motif, so I took SXEmacs).


[-- Attachment #2: sxe-vs-gnu.png --]
[-- Type: image/png, Size: 184576 bytes --]