Sage Survey 2009

[Harald Schilly]

Hello everybody.

Last fall I made a survey about Sage. This year I want to repeat it
with more questions that fit better the responses from last time.
Everyone (non-users, users, developers) is invited to take it. After
about two weeks, I'll publish the results and we can discuss it. I'll
also analyze it and distill the most important things in a summary.
Later on, I'll keep the form open for new submissions so that we'll
get continuous feedback (e.g. each month a new form).

https://spreadsheets.google.com/viewform?formkey=dERRZldlYTNObUY4QmptNUVsRDNhZGc6MA

Harald

[Simon King]

Hi Harald,

I just went to the page and found a misprint: "In order to archieve
our mission" should be "achieve".

Cheers,
Simon

[Harald Schilly]

On Nov 18, 1:39 pm, Simon King <simon.k...@nuigalway.ie> wrote:
> ... "achieve".

Thanks, fixed now!

[Kwankyu Lee]

As East Asian countries, please include Korea along side with China
and Japan.

Thank you!

[David Kirkby]

2009/11/18 Harald Schilly <harald....@gmail.com>:

Given both Solaris and OpenSolaris are supposed to be supported
operating sytems, I'm surprrised they are not there, given Windows and
linux both are - I can't recall if OS X was, but if not, that needs to
be there too

I'm on a laptop at a rail station, so have not had a look at much of
it, but that hit me straight away.

Dave

[Harald Schilly]

They are meant to be included by the dots. It's just for a rough
estimate where you are, nothing specific.

[Harald Schilly]

On Nov 18, 3:10 pm, David Kirkby <david.kir...@onetel.net> wrote:
> Solaris and OpenSolaris ...

Well, I tried to minimize the number of questions. The operating
systems are the top three from the website&download statistics. It's
also ok to add Solaris in the box below.

H

[Kwankyu Lee]

Hi Harald,

On Nov 19, 12:59 am, Harald Schilly <harald.schi...@gmail.com> wrote:
> They are meant to be included by the dots. It's just for a rough
> estimate where you are, nothing specific.

I know. :-)

This survey is nice. Thank you for your work.


Kwankyu

[Peter Jeremy]

On 2009-Nov-18 13:35:11 +0100, Harald Schilly <harald....@gmail.com> wrote:
>Last fall I made a survey about Sage. This year I want to repeat it
>with more questions that fit better the responses from last time.
>Everyone (non-users, users, developers) is invited to take it.

My initial comment is that making virtually every question mandatory
is likely to turn people off or make them pick random answers to
questions they don't want to answer.

--
Peter Jeremy

[Harald Schilly]

On Nov 20, 8:04 am, Peter Jeremy <peterjer...@acm.org> wrote:
> My initial comment is that making virtually every question mandatory
> is likely to turn people off or make them pick random answers to
> questions they don't want to answer.

The last year I thought this too, but turns out most of the answers
were not random at all. Right now there are 126 submissions, and
really a lot of text to read about in various topics. I assume that if
you take the time to write some sentences in each text box, you do not
select random answers.
Of course, there are also fake submissions, but they are easy to spot
and of course nobody is forced to answer anything. It's just a process
to take some sort of snapshot of the community, distill it, and feed
it back to the community - self reflection and so on ...

H

[William Stein]

On Fri, Nov 20, 2009 at 3:07 AM, Harald Schilly
<harald....@gmail.com> wrote:
> On Nov 20, 8:04 am, Peter Jeremy <peterjer...@acm.org> wrote:
>> My initial comment is that making virtually every question mandatory
>> is likely to turn people off or make them pick random answers to
>> questions they don't want to answer.
>
> The last year I thought this too, but turns out most of the answers
> were not random at all. Right now there are 126 submissions, and

Yep -- I was just looking at them and was pleased to see that there
are so many responses, especially given the length of the survey.

> really a lot of text to read about in various topics. I assume that if
> you take the time to write some sentences in each text box, you do not
> select random answers.
> Of course, there are also fake submissions, but they are easy to spot
> and of course nobody is forced to answer anything.
> It's just a process
> to take some sort of snapshot of the community, distill it, and feed
> it back to the community - self reflection and so on ...

How many people responded last year?

William

[Harald Schilly]

On Nov 20, 12:25 pm, William Stein <wst...@gmail.com> wrote:
> How many people responded last year?
>

Last year 186 valid answers in nearly 2 weeks. We are now at 156
answers in 4 days.

H

[rjf]

Fake submissions?? Huh? By people who want to sell you fake Rolex
watches?

My objection is that the obvious question -- do you "know" maxima is
not
asked. (Similarly Axiom, I guess). But see sci.math.symbolic for
further discussion
of statistics.

On Nov 20, 3:07 am, Harald Schilly <harald.schi...@gmail.com> wrote:
> On Nov 20, 8:04 am, Peter Jeremy <peterjer...@acm.org> wrote:
>
> > My initial comment is that making virtually every question mandatory
> > is likely to turn people off

I agree.

> or make them pick random answers to
> > questions they don't want to answer.


>
> The last year I thought this too, but turns out most of the answers
> were not random at all.

No, I just had to go back repeatedly and say no, I don't know about
this or that.
Annoying.

Right now there are 126 submissions, and
> really a lot of text to read about in various topics. I assume that if
> you take the time to write some sentences in each text box, you do not
> select random answers.

Not random in the sense of an appropriate distribution. Just random in
the
sense of who cares how much I know about linux or os x.

As the old joke goes,

How's your wife?

Compared to what?

[Harald Schilly]

Hi, I've closed the survey. We have 233 submissions after 1 week.

On Nov 18, 1:35 pm, Harald Schilly <harald.schi...@gmail.com> wrote:
> Later on, I'll keep the form open for new submissions so that we'll
> get continuous feedback (e.g. each month a new form).

I've put a small link to a new form (cleaned list of responses) on the
website. I'll modify and enhance the questions and only a few will be
mandatory. Feedback welcome!

@rfj: I've added a "Maxima, Sympy, ..." row in the software grid to
include open source computer algebra systems. I didn't want to add
that row in the middle of the survey (after about 150 responses of
233) to avoid distorted statistics.

Who wants to help me analyzing the responses?

There were also concerns about publishing the data. I want to share
everything, because this is a great opportunity for everyone to read
what others think about the project. For example, if I have a feeling
for this or that and I read it from somebody else, it helps that idea.
It also fosters exchanging different views, between developers and
users, and so on. I'm not done reading the data, but as far as i can
see there shouldn't be any concerns. I see this like a virtual
gathering of all Sage interested people in a huge room and everyone is
now able to read each others mind.
@Simon, about the location / geographical areas: They are very
unspecific (usa, europe+russia, eastern asia ... there are millions of
people!), but I can delete that column if it is really a problem. The
survey wasn't asking about the country and nowhere is this
information...

H

[Martin Albrecht]

> @Simon, about the location / geographical areas: They are very
> unspecific (usa, europe+russia, eastern asia ... there are millions of
> people!), but I can delete that column if it is really a problem. The
> survey wasn't asking about the country and nowhere is this
> information...

I bet you could still easily identify my answer just like I could identify
yours. My impression is that the biggest leakage would happen within the
regular contributors group (which isn't millions anymore) While I am not too
concerned about my personal data in this survey (since I expected it would be
shared) others might think differently.

Martin

--
name: Martin Albrecht
_pgp: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x8EF0DC99
_otr: 47F43D1A 5D68C36F 468BAEBA 640E8856 D7951CCF
_www: http://www.informatik.uni-bremen.de/~malb
_jab: martinr...@jabber.ccc.de

[Robert Bradshaw]

On Nov 25, 2009, at 10:43 AM, Harald Schilly wrote:

> Hi, I've closed the survey. We have 233 submissions after 1 week.
>
> On Nov 18, 1:35 pm, Harald Schilly <harald.schi...@gmail.com> wrote:
>> Later on, I'll keep the form open for new submissions so that we'll
>> get continuous feedback (e.g. each month a new form).
>
> I've put a small link to a new form (cleaned list of responses) on the
> website. I'll modify and enhance the questions and only a few will be
> mandatory. Feedback welcome!
>
> @rfj: I've added a "Maxima, Sympy, ..." row in the software grid to
> include open source computer algebra systems. I didn't want to add
> that row in the middle of the survey (after about 150 responses of
> 233) to avoid distorted statistics.
>
> Who wants to help me analyzing the responses?
>
> There were also concerns about publishing the data. I want to share
> everything, because this is a great opportunity for everyone to read
> what others think about the project. For example, if I have a feeling
> for this or that and I read it from somebody else, it helps that idea.
> It also fosters exchanging different views, between developers and
> users, and so on. I'm not done reading the data, but as far as i can
> see there shouldn't be any concerns. I see this like a virtual
> gathering of all Sage interested people in a huge room and everyone is
> now able to read each others mind.

Sharing aggregate data, I think, is fine (e.g. totals for all the
multiple choice questions). Also, it is probably safe to give all the
answers for any specific question, but I would not make available a
list of answers by person.

> @Simon, about the location / geographical areas: They are very
> unspecific (usa, europe+russia, eastern asia ... there are millions of
> people!), but I can delete that column if it is really a problem. The
> survey wasn't asking about the country and nowhere is this
> information...

There's a difference between "250 people from europe answered the
survey" and "the person who said this, this, and that lives in ease
asia."

- Robert

[Dr. David Kirkby]

Harald Schilly wrote:

> There were also concerns about publishing the data. I want to share
> everything, because this is a great opportunity for everyone to read
> what others think about the project. For example, if I have a feeling
> for this or that and I read it from somebody else, it helps that idea.
> It also fosters exchanging different views, between developers and
> users, and so on. I'm not done reading the data, but as far as i can
> see there shouldn't be any concerns. I see this like a virtual
> gathering of all Sage interested people in a huge room and everyone is
> now able to read each others mind.
> @Simon, about the location / geographical areas: They are very
> unspecific (usa, europe+russia, eastern asia ... there are millions of
> people!), but I can delete that column if it is really a problem. The
> survey wasn't asking about the country and nowhere is this
> information...

I'm sure looking at the survey, you must be able to pick out my entry. I do not
recall saying anything that I've not at least implied before publicly. I was
well aware that I could probably be identified when I wrote the comments.

If people can be identified by

* An unusual choice of operating system
* Exact location - I see nothing wrong with unspecific locations myself.
* Detailed research interests

then it would probably not be difficult for others to work out who the comments
came from. I doubt I could personally determine anyone from their research
interests, but I'm sure others could.

It might be better for the *next* survey to have two comments areas.

* Confidential - shared with you and William only.

* Public - someone does not care if the comments are made public, even if they
may be identified from them.

[Simon King]

Hi Harald!

Where can one read the survey results? I think you did not provide a
link.

On 26 Nov., 00:06, "Dr. David Kirkby" <david.kir...@onetel.net> wrote:
[...]

> It might be better for the *next* survey to have two comments areas.
>
> * Confidential - shared with you and William only.
>
> * Public - someone does not care if the comments are made public, even if they
> may be identified from them.

Sounds good to me.

Cheers,
Simon