How Many Emacs Users?

How Many Emacs Users?

[Brad Collins]

I'm writing a investment proposal and it would be helpful to include
the number of Emacs users (Xemacs etc) are out there.

Are we talking in the tens, hundreds, a million or more?

I remember back in the early 90's when I was running an ISP in Hong
Kong and was called by the South China Morning Post (the biggest
English language newspaper in HK) asking how many Internet Users there
were in Hong Kong.  I hadn't slept in a couple of days so I gave a
number off the top of my head.  The number ended up being quoted on a
Web site (I think at Chinese University) and a month later it was
being cited as fact by the Hong Kong government.  After that the
number was just cited as a government statistic.

Any statistic is dubious, but if anyone has any ball park guesses
it would be helpful.

Cheers,

b/

--
Brad Collins
Chenla Labs
Bangkok, Thailand

Re: How Many Emacs Users?

[David Kastrup]

Brad Collins <brad@studiojungle.net> writes:

> I'm writing a investment proposal and it would be helpful to include
> the number of Emacs users (Xemacs etc) are out there.
> 
> Are we talking in the tens, hundreds, a million or more?

Two, and I am one of them.  Apart from that, talking about Emacs being
"usable" is an insult, much like calling your spouse "usable" and
trying to estimate her worth by the number of other suitors.

Let me not to the marriage of true minds admit impediments.

Xpost and Fupto alt.religion.emacs

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: How Many Emacs Users?

[Tim McNamara]

David Kastrup <dak@gnu.org> writes:

> Brad Collins <brad@studiojungle.net> writes:
>
>> I'm writing a investment proposal and it would be helpful to
>> include the number of Emacs users (Xemacs etc) are out there.
>> 
>> Are we talking in the tens, hundreds, a million or more?

How would there be any way to know, unless GNU or FSF keep download
statistics?  And even that woul dnot be reliable, given that there are
many precompiled binary distributions that GNU/FSF now nothing about.
There's no sales receipts for Emacs...

> Two, and I am one of them.  Apart from that, talking about Emacs
> being "usable" is an insult, much like calling your spouse "usable"
> and trying to estimate her worth by the number of other suitors.

WTF are you on about, David?

Re: How Many Emacs Users?

[Pascal Bourguignon]

Tim McNamara <timmcn@bitstream.net> writes:
> David Kastrup <dak@gnu.org> writes:
> 
> > Brad Collins <brad@studiojungle.net> writes:
> >
> >> I'm writing a investment proposal and it would be helpful to
> >> include the number of Emacs users (Xemacs etc) are out there.
> >> 
> >> Are we talking in the tens, hundreds, a million or more?
> 
> How would there be any way to know, unless GNU or FSF keep download
> statistics?  And even that woul dnot be reliable, given that there are
> many precompiled binary distributions that GNU/FSF now nothing about.
> There's no sales receipts for Emacs...

About  half the  unix  users, I'd  say.   The other  half (minus  some
epsilon) would be  vi users.  Well, remove MacOSX  users from the unix
user count...

Between 2 and 4 million I'd say.
-- 
__Pascal_Bourguignon__                     http://www.informatimago.com/
There is no worse tyranny than to force a man to pay for what he doesn't
want merely because you think it would be good for him.--Robert Heinlein
http://www.theadvocates.org/

Re: How Many Emacs Users?

[Kin Cho]

Based on my own experience, I'd say less than 10% of the unix users
use emacs.

-kin

Pascal Bourguignon <spam@thalassa.informatimago.com> writes:

> Tim McNamara <timmcn@bitstream.net> writes:
> > David Kastrup <dak@gnu.org> writes:
> > 
> > > Brad Collins <brad@studiojungle.net> writes:
> > >
> > >> I'm writing a investment proposal and it would be helpful to
> > >> include the number of Emacs users (Xemacs etc) are out there.
> > >> 
> > >> Are we talking in the tens, hundreds, a million or more?
> > 
> > How would there be any way to know, unless GNU or FSF keep download
> > statistics?  And even that woul dnot be reliable, given that there are
> > many precompiled binary distributions that GNU/FSF now nothing about.
> > There's no sales receipts for Emacs...
> 
> About  half the  unix  users, I'd  say.   The other  half (minus  some
> epsilon) would be  vi users.  Well, remove MacOSX  users from the unix
> user count...
> 
> Between 2 and 4 million I'd say.
> -- 
> __Pascal_Bourguignon__                     http://www.informatimago.com/
> There is no worse tyranny than to force a man to pay for what he doesn't
> want merely because you think it would be good for him.--Robert Heinlein
> http://www.theadvocates.org/

Re: How Many Emacs Users?

[Brad Collins]

> Between 2 and 4 million I'd say.

Thanks for the replys -- back channeled as well...

This is not a life or death thing.  And it's obvious there is no way
of determining how many people who use any flavour of Unix or
Gnu/Linux could be using, so I think it would be better to use a
simular approach that is used by congressional offices and say that
for every person subscribed to a Emacs mailing list there are X
number users who aren't.

It's a bit like Cicadas.  You sure as hell know they are there,
making that infernal racket, and you'd think bugs that size would not
be terribly adept at hiding, but finding the damn things is more
trouble than its worth.

b/

--
Brad Collins
Chenla Labs
Bangkok, Thailand

Re: How Many Emacs Users?

[David Kastrup]

Brad Collins <brad@studiojungle.net> writes:

> > Between 2 and 4 million I'd say.
> 
> Thanks for the replys -- back channeled as well...
> 
> This is not a life or death thing.  And it's obvious there is no way
> of determining how many people who use any flavour of Unix or
> Gnu/Linux could be using, so I think it would be better to use a
> simular approach that is used by congressional offices and say that
> for every person subscribed to a Emacs mailing list there are X
> number users who aren't.

No way, Jose.  Emacs mailing lists are for the people actually
interested in details.  But Emacs is one of the main editor
variants.  A _lot_ of people use it that would never want to even get
acquainted with it more closely.

At a time where the average Emacs user compiled and installed his
Emacs by hand, the mailing list numbers might have still had some
relation to the actual user numbers.

But since Emacs pretty much by now is a standard GNU component and
one of the default "big" editors for Linux, the silent masses are too
dominant to be discounted.

Would you estimate the number of Microsoft Office users from the
subscribers to some mailing list?

If you did, they may have already lost the desktop war.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: How Many Emacs Users?

[Per Abrahamsen]

David Kastrup <dak@gnu.org> writes:

> At a time where the average Emacs user compiled and installed his
> Emacs by hand, the mailing list numbers might have still had some
> relation to the actual user numbers.

Back then the average Emacs user was student at a university, where
the sysadmin compiled and installed Emacs on the big, central shared
computer.  Or if they were progressive, on the big, central shared NFS
server.

Your other estimate was better, judging by various pools, among those
Unix users who care what OS they are running, that is today mostly
GNU/Linux enthusiasts, I guess it is 50% vi(m) users, and 40% Emacs
users, and 11% other[1].  

What is hard is to estimate how large a fraction the non-enthusiast
fraction of the GNU users constitute these days.  The people who use a
GNU/Linux + StarOffice + Mozilla setup because of the freedom or price
benefits over a Microsoft or Apple box with similar tools, but don't
really care about the Unix personality.  These are unlikely to
participate in pools.  And I suspect they constitute most of the
growth in the GNU desktop market share.

Anyway, I take a shoot and say half the users are currently in that
silent "other" category, which give an estimate of one million Emacs
users. The real number is probably within one or two orders of
magnitude of that.

Footnotes: 
[1]  Numbers may not add up to 100% because of rounding, despite being
made up for the occasion. 

Re: How Many Emacs Users?

[Tim McNamara]

Pascal Bourguignon <spam@thalassa.informatimago.com> writes:

> Tim McNamara <timmcn@bitstream.net> writes:
>> David Kastrup <dak@gnu.org> writes:
>> 
>> > Brad Collins <brad@studiojungle.net> writes:
>> >
>> >> I'm writing a investment proposal and it would be helpful to
>> >> include the number of Emacs users (Xemacs etc) are out there.
>> >> 
>> >> Are we talking in the tens, hundreds, a million or more?
>> 
>> How would there be any way to know, unless GNU or FSF keep download
>> statistics?  And even that woul dnot be reliable, given that there
>> are many precompiled binary distributions that GNU/FSF now nothing
>> about.  There's no sales receipts for Emacs...
>
> About half the unix users, I'd say.  The other half (minus some
> epsilon) would be vi users.  Well, remove MacOSX users from the unix
> user count...

Not all (Emacs 21.3.50.1 on powerpc-apple-darwin 7.2.0 a.k.a. OS X
10.3.2).  OS X comes with Emacs as part of the standard distribution,
not that most OS X users have even the slightest awareness of that
fact.

> Between 2 and 4 million I'd say.

That's pretty cool.

Re: How Many Emacs Users?

[Joe Corneli]

Answer: One million, four hundred and fourty three thousand, three
hundred and thirty three.


Demonstration.

4,330,000 = magnitude of google(emacs).

1. Assume that each page has exactly one author.

2. Note that the pages need not be authored by distinct individuals.

3. Also, not every page will have been authored by an emacs user.

4. However, not every emacs user will have authored a web page that
talks about emacs.


Weight (3) and (4).  Assume that half of the pages were written by
non-Emacs users.  Furthermore, assume that for every 3 non-users who
wrote a webpage, there will be 1 user who didn't write a webpage.

Weight (2).  Assume every person who has written a webpage has
actually written 2 webpages on average.

Weight (1).  This is too complicated to deal with properly, so the
assumption stands.


By (2) we have (/ 4330000 2) = 2165000 = number of distinct authors.

By (3) we have (/ 2165000 2) = 1082500 = number of pages written by
emacs users (or nonusers).

By (4) we have (/ 1082500 3) = 360833 = number of emacs users who
didn't write pages.

Answer is (+ 1082500 360833) = 1443333


Scholium.

If you don't like this number, you can adjust the weights according
to your own personal value system.  I actually think it is likely to
overestimate by a factor of about 2.5 to 3 (mostly because of
underweighting assumption 3). Numerology aside, my personal view is
that the figure should be between half a million and a million.