Re: Emacs's popularity

[Tim X]

Richard Riley <rileyrgdev@gmail.com> writes:

> "Lennart Borgman" <lennart.borgman@gmail.com> writes:
>
>> On Mon, Dec 15, 2008 at 11:28 PM, Richard Riley <rileyrgdev@gmail.com> wrote:
>>
>>>>> But Vim is not only installed; it's really used a lot. In Debian Vim has
>>>>> always been a bit more popular than Emacs but in the first half of 2007
>>>>> Vim really got popular (around Vim 7.1 and Debian 4.0 release). This
>>>>> "used actively" graph compares vim-common, emacs21-bin-common and
>>>>> emacs22-bin-common packages:
>>>>>
>>>>>    http://preview.tinyurl.com/5thmmx
>>>>
>>>> That is a bit strange since the vi emulator Viper in Emacs is now so good.
>>>>
>>>
>>> Not strange at all Lennart, Why would someone run the Emacs OS to run
>>> emulated vim  when they can run the real thing in 100th of the
>>> footprint?
>>
>> Exactly why do you think the footprint matter?
>
> Are you serious?
>
> Memory usage, start speed, response, and all thinks linked.
>
> And with the boom in netbooks and OSen on USB sticks, emacs finds itself
> more and more pushed into that "heavy dinosaur" category. I'm not saying
> I *agree* with it, just those are my observations.
>
> Personally I detest Vi and clones.

I think its an apples and oranges comparison. People who like vi*
editors are less likely to like emacs and vice versa. They serve
different audiences and different ways of working. I was a vi user for
nearly 15 years before I switched to emacs and I've now been using emacs
for 10 years. They way I work with each editor is completely different. 

When I started using emacs I had to change the way I worked because
emacs required it. Instead of opening my editor, doing some work,
closing it and moving on, with emacs, I open it, leave it open and only
close it when I'm logging out, rebooting etc. 

The biggest hurdle to emacs isn't due to 'odd' user interface defaults,
odd terminology or most of the claims that periodically come up in this
group. The hurdle is in the fact that emacs is significantly different
to other text editors. While it has much of the same functionality, the
way you use it is very different and requires something even more
difficult than learning new commands - it requires a paradigm shift and
a change in habits. I failed in my first few attempts simply because I
didn't really have the inclination or drive to make that shift and learn
new habits. Changing defaults, turning on CUA mode, changing terminology
etc is unlikely to have any real impact - this is just surface stuff. In
fact, it cold have a negative impact in that people are likely to make
comparisons based on the common low level functionality and decide there
is nothing about emacs that is any better than, lets say, vi, textmate,
etc. 

I also remain unconvinced by the claims that emacs is dying and needs to
become more popular. Even in the late 80s and 90s, the majority of
people I worked with in the Unix world used vi and its clones. Few used
emacs. Those that did were very dedicated and strong supporters, but
they certainly weren't in the majority - not anywhere I worked at
least. I don't think emacs is ever going to be what wold be called a
popular editor. 

With respect to it dying, I do't think the facts support the
claim. Emacs continues to see steady development. It remains stable and
powerful and still has a very dedicated following. In fact, I suspect
emacs is seeing more new development in recent years than it saw for
much of the 90s. The various forums, such as this newgroup seems to get
a respectable number of posts and I've not seen any great decline over
the last 10 years. The many new enhancements currently available ini
emacs 23, such as support for anti-aliased fonts, utf-8, improved
internationalisation, GTK based interface, built-in support for GPG,
dbus, json, svg graphics, new modes etc don't make me feel like its a
project thats about to die. While some may find criticism for its
development model, it should also be noted that although improvements
are always possible, its a model that has worked for a very long time -
a lot longer than the many other open source and free software projects
that have been cited in this thread. 

I would also be very wary regarding any effort to make emacs
'popular'. All too often, popularity is obtained by appealing to the
lowest common denominator. Emacs has traditionally been focused on the
needs of developers and I think that is still where the focus should
be. 

One of the advantages that emacs has is in the fact it can easily be
customized to suit new environments. I remember when I started doing
java when java was new. At the time, there wasn't a decent editor for
working in Java. One of the very first and possibly the best was JDE,
the emacs java development environment. Now there are numerous editors
that are ideal for working with java and to some extent, JDE may seem a
bit antiquated or limited in comparison. However, the point to note is
that it was one of the first really good environments available on
multiple platforms and provided a good solution while more powerful java
development environments were being implemented and refined. I have been
lucky enough not to have to do any java development for nearly 10 years
now, so I can't really say how JDE compares, but I suspect it looks
somewhat primitive to developers used to existing purpose built
environments. However, this is often a theme with emacs - in the end, it
may not have the best available development environment for your fav
language, but generally, it will be one of the first and I suspect
often, many of the more sophisticated environments used what emacs did
as inspiration. Other fine examples of this can be seen in modes like
SLIME, which is still the best environment I've used for lisp and is as
good as any of the commercial environments. Now, I'm using it with
clojure - as yet, there is no other environment for working with clojure
which comes even close.

I would therefor suggest that rather than spend time and effort trying
to make emacs more popular or attempting to find some miracle formular
consisting of terminology changes, manual updates, new defaults etc,
time would be better spent on improving packages that can help create
really high quality development environments. For example, the CEDET
project and semantic, parsers for different languages, etc are likely to
do a lot more to ensure emacs' continued development and existance. Such
packages and modes are non-trivial and take considerable effort, but
hold the promise of really great things. There are no simple things that
can be done that will achieve much and I would suggest even minor
contributions to significant projects like CEDET will provide far more
benefit than anything else.

peace,

Tim



-- 
tcross (at) rapttech dot com dot au