Re: Emacs's popularity (was: Distributed Maintenance for Emacs)

[David L]

On Thu, Dec 18, 2008 at 9:50 AM, Drew Adams <drew.adams@oracle.com> wrote:
<snip>
>> > Why the need to make Emacs the most popular? It's good to
>> > make Emacs better, but what's the popularity contest about?
>> > Perhaps Americans on average listen to
>> > Britney Spears more than Mozart or Muddy Waters. So what?
>>
>> A tool with a shrinking user base also has a shrinking pool
>> of potential developers that will continue to make it better.
>> Fewer developers to make it better will make it less competitive
>> with other tools causing it to lose more users.  And so on.
>> There are also fewer people to file
>> bug reports and help other users with problems.
>
> You miss the point. No one has argued that Emacs should be unpopular or that
> unpopularity is better than popularity. No one has said that having fewer 
> users
> and developers is better than having more.
>
> I said "Why the need to make Emacs _the most_ popular?". And I argued against
> chasing popularity for its own sake, as opposed to trying to improve Emacs - a
> difference in aim.

I agree, it doesn't need to be the "most popular" as a goal in itself.
But if the reason that it is not the most popular is usability, features,
and bugs (as opposed to a different target audience or different
goals), then that is a problem.  I'm not saying that is necessarily the
case, but it is the reason given by new developers at my company
for not adopting it despite tutorials on its use.

Obviously, emacs primary goal is not to be an IDE
but it does have a lot of great features as an editor and I like using
it as an IDE more than any other that I've tried including eclipse and
kdevelop.  But that's because the the editor is very good and the
other IDE features are marginally good enough or I don't care about
them.  Other people at my company care more about things like
ECB features (which isn't even part of the standard emacs) than
I do, and I have to admit that I would probably bail on emacs too
if I cared more about ECB, completion, etc.


>
>> Another reason that is would be nice if it was growing in popularity
>> is to give people like me some ammunition when many people
>> in my company are saying emacs is obsolete and I should start
>> using eclipse.
>
> Let them use Eclipse. Learn from Eclipse. Improve Emacs using Eclipse as
> inspiration, if that's appropriate.
>
> If you want to help people learn how Emacs can help them, then show them (and
> argue from) the specific merits of Emacs. Show and describe Emacs features.
> Don't try to sell them on Emacs because it is the most popular.
>
I tried... they found ECB inadequate and difficult to set up and use
and the gdb interface buggy (I should mention at this point that
I believe the emacs gdb guy is one of the most responsive
open source developers I've ever dealt with, so this should in
no way be taken to reflect negatively on him).

> Trying to persuade others to use Emacs because of its relative popularity is a
> misguided approach. I don't know many good developers who would be persuaded 
> to
> use a tool just because of its popularity. Talk to them about features, not
> popularity.
I'm not trying to persuade them to use it based on its popularity.
I'm trying to justify my continued use of it despite poor competitiveness
as an IDE.  Basically they don't care about the things that I really like
about emacs and I am willing to live with its shortcomings as an
IDE whereas they are not.

But I do think it is fair for a company to require proficiency in one
or two of the more popular tools.  It means a higher chance that
new hires will already have experience with the tools and a
broader base of users to ask for help and developers to improve
it.


>
>> There is something to be said for standardizing tools within
>> a company.
>
> Is there? Depends what you mean. If you mean standards, yes. If you mean how
> tools interface, yes.
>
> But if you mean which tools an individual uses to accomplish tasks such as
> editing, email, etc., then no, I'm not convinced.
>
Look, I had this same argument with my management, and I was
on your side.  But they had some good points.  Training and support
for tools costs time and money.  If I have a problem with a tool,
it's nice if IT or somebody can help me.

>
>> I always get frustrated when I sit down with somebody to help them debug
>> something and don't know my way around eclipse and I know they feel
>> the same with emacs.
>
> So learn to use Eclipse. Do you need Emacs to be _the most popular_ just so 
> you
> don't need to learn Eclipse?

I think eclipse sucks.  It does't make sense for me to learn it for the 5% of
my time that I spend working with other people.  And because I am very
productive at writing and debugging code, there is no serious pressure
for me to learn eclipse (yet).

>> If a company forces developers to use a certain
>> tool or a few more popular tools (which many companies do), it's
>> like people are forced to listen to Britney Spears (to use your
>> analogy).
>
> Uh, and that's a good thing?
No, that's a horrible thing.  But it happens.

> If you're an Emacs lover, then it's about improving Emacs. If you're an 
> Eclipse
> lover, then it's about improving Eclipse. (And I assume that one can be both.)
> Improvement, not popularity.
Open source software is different than your music analogy.  Popularity
tends to help to improve open source software.  There is a positive
feedback loop between popularity and improvement... there are other
effects that can dampen this feedback, so it's not the only factor.
Popularity for the sake of popularity is bad, but prioritizing features and
bugs that are contributing to reduced popularity might improve
emacs faster.

Cheers...

         David

PS - Does emacs have a feature request and bug reporting system
like bugzilla yet?  I heard rumors that it might be getting one.  With
respect to this discussion, it would be nice if when emacs chokes
during a demonstration of its features, I could point to the bugzilla
page that shows the problem has been reported and is being
worked on and is a blocker for the release of version x.y which
is scheduled to be release on such and such a date.