William Stein

2008-03-08 17:38:42 UTC

Hi,

This email is about recent Sage download stats, and ends with an estimate

that there are currently perhaps about 5000 Sage users.

Here are the download numbers for Sage during the last 6 days from the three

sites (sagemath.org, sage.math, and modular.fas.harvard.edu) that I administer:

Linux Binary

26

OS X Binary

12

Source

85

VMware

78

TOTAL: 201 downloads

UPGRADES:

Here are the number of "sage -upgrades" during each of the last few weeks:

This last week: 135

Previous weeks: 337, 302, 163, 349, 307, 228, 338, 375, 336, 388, 355,

348, 398, 297, 363

CONCLUSIONS:

1. There are probably 100 downloads/week from all the other 4 mirror

sites combined (probably

a conservative estimate?), so that's about 300*4 = 1200 downloads per month.

2. There are also about 1200 upgrades per month.

3. Adding 1 and 2 we get about 2400 times Sage is installed or

upgraded per month.

4. In my experience with both research mathematicians and students (and my own

personal experience), people do not upgrade regularly unless forced to

or actively

working as developers on projects. Upgrading can break things they are doing,

and that can screw up their workflow, so they put it off until they

have some free

time (e.g., spring break, finals are over, whatever). Such free time

occurs maybe

once every 2 months on average. If n people use Sage and

upgrade/install once every

2 months on average, and there are 2400 upgrade/installs per month, then

there are probably about 2*2400 ~ 5000 sage users.

If anybody else has ideas about estimating the number of Sage users, please feel

free to share them.

---

If the above measure of number of users seems at all reasonable to people who

are much better at this sort of thing (statisticians? marketing droids?) then

we could use them as input to a 2008 goal to raise the number of Sage users

(as defined by a metric as above) to 10,000.

By the way, Maple, Mathematica, and Matlab all claim to have in excess of

1,000,000 regular users... That is definitely the size of use base

Sage aspires to

eventually have.

Estimate of number of Sage users by year:

Feb 2005: 3

Feb 2006: 100

Feb 2007: 500

Feb 2008: 5,000

Feb 2009: 10,000 (Goal)

Feb 2009: 25,000 (Goal)

Feb 2010: 100,000 (Goal)

Feb 2011: 1,000,000 (Goal)

Judging by my estimates for Maxima and other open source math software projects,

I think getting over 10,000 serious users for an open source math

project is already

a quite difficult thing to achieve, since it starts taking one outside

the realm of niche

market. Sage has definitely not achieved this (yet).

Going from 10,000 to 25,000 will be quite difficult as well. Going from 25,000

to 100,000 will be crazy hard since by then Mathematica/Maple/Matlab

etc. will have

taken serious notice and start fighting back (hopefully by innovating,

e.g., by geatly

improving their exact linear algebra, group theory, etc. functionality).

And also the range of people one

has appeal to is much larger. On the other hand, 25,000 people will be telling

their friends....

Going from 100,000 to 1 million users would be quite interesting.

There are massive

infrastructure and support issues, and the user base is the same as

Mathematica.

That means the daily traffic to sagemath.org could be about the same as during

the slashdotting, i.e., completely unsustainable without a lot of

reorganization. That

means there would be substantial demand for seminars and training

courses and other

forms of support. There's no way this will work unless we organize some sort

of international network of local people to help out with such

activities for Sage.

This is also the point where Sage would genuinely have reached the goal of

providing a "viable alternative to Maple, Magma, Mathematica, and Matlab" for

almost anybody.

Without a lot of hard work and effort by us, though, the user

base could easily shrink into the low hundreds again, so let's keep

working very very very hard.

--

William Stein

Associate Professor of Mathematics

University of Washington

http://wstein.org

--~--~---------~--~----~------------~-------~--~----~

To post to this group, send email to sage-***@googlegroups.com

To unsubscribe from this group, send email to sage-devel-***@googlegroups.com

For more options, visit this group at http://groups.google.com/group/sage-devel

URLs: http://www.sagemath.org

-~----------~----~----~----~------~----~------~--~---

This email is about recent Sage download stats, and ends with an estimate

that there are currently perhaps about 5000 Sage users.

Here are the download numbers for Sage during the last 6 days from the three

sites (sagemath.org, sage.math, and modular.fas.harvard.edu) that I administer:

Linux Binary

26

OS X Binary

12

Source

85

VMware

78

TOTAL: 201 downloads

UPGRADES:

Here are the number of "sage -upgrades" during each of the last few weeks:

This last week: 135

Previous weeks: 337, 302, 163, 349, 307, 228, 338, 375, 336, 388, 355,

348, 398, 297, 363

CONCLUSIONS:

1. There are probably 100 downloads/week from all the other 4 mirror

sites combined (probably

a conservative estimate?), so that's about 300*4 = 1200 downloads per month.

2. There are also about 1200 upgrades per month.

3. Adding 1 and 2 we get about 2400 times Sage is installed or

upgraded per month.

4. In my experience with both research mathematicians and students (and my own

personal experience), people do not upgrade regularly unless forced to

or actively

working as developers on projects. Upgrading can break things they are doing,

and that can screw up their workflow, so they put it off until they

have some free

time (e.g., spring break, finals are over, whatever). Such free time

occurs maybe

once every 2 months on average. If n people use Sage and

upgrade/install once every

2 months on average, and there are 2400 upgrade/installs per month, then

there are probably about 2*2400 ~ 5000 sage users.

If anybody else has ideas about estimating the number of Sage users, please feel

free to share them.

---

If the above measure of number of users seems at all reasonable to people who

are much better at this sort of thing (statisticians? marketing droids?) then

we could use them as input to a 2008 goal to raise the number of Sage users

(as defined by a metric as above) to 10,000.

By the way, Maple, Mathematica, and Matlab all claim to have in excess of

1,000,000 regular users... That is definitely the size of use base

Sage aspires to

eventually have.

Estimate of number of Sage users by year:

Feb 2005: 3

Feb 2006: 100

Feb 2007: 500

Feb 2008: 5,000

Feb 2009: 10,000 (Goal)

Feb 2009: 25,000 (Goal)

Feb 2010: 100,000 (Goal)

Feb 2011: 1,000,000 (Goal)

Judging by my estimates for Maxima and other open source math software projects,

I think getting over 10,000 serious users for an open source math

project is already

a quite difficult thing to achieve, since it starts taking one outside

the realm of niche

market. Sage has definitely not achieved this (yet).

Going from 10,000 to 25,000 will be quite difficult as well. Going from 25,000

to 100,000 will be crazy hard since by then Mathematica/Maple/Matlab

etc. will have

taken serious notice and start fighting back (hopefully by innovating,

e.g., by geatly

improving their exact linear algebra, group theory, etc. functionality).

And also the range of people one

has appeal to is much larger. On the other hand, 25,000 people will be telling

their friends....

Going from 100,000 to 1 million users would be quite interesting.

There are massive

infrastructure and support issues, and the user base is the same as

Mathematica.

That means the daily traffic to sagemath.org could be about the same as during

the slashdotting, i.e., completely unsustainable without a lot of

reorganization. That

means there would be substantial demand for seminars and training

courses and other

forms of support. There's no way this will work unless we organize some sort

of international network of local people to help out with such

activities for Sage.

This is also the point where Sage would genuinely have reached the goal of

providing a "viable alternative to Maple, Magma, Mathematica, and Matlab" for

almost anybody.

Without a lot of hard work and effort by us, though, the user

base could easily shrink into the low hundreds again, so let's keep

working very very very hard.

--

William Stein

Associate Professor of Mathematics

University of Washington

http://wstein.org

--~--~---------~--~----~------------~-------~--~----~

To post to this group, send email to sage-***@googlegroups.com

To unsubscribe from this group, send email to sage-devel-***@googlegroups.com

For more options, visit this group at http://groups.google.com/group/sage-devel

URLs: http://www.sagemath.org

-~----------~----~----~----~------~----~------~--~---