Subject: Re: Open Source Decision Models
From: Somik Raha <somikr@gmail.com>
Date: Fri, 5 Mar 2010 09:19:43 -0800
Fri, 5 Mar 2010 09:19:43 -0800
Kevin, Chuck, thank you for your responses. I am familiar with the GPL and
other open-source code licenses and I do plan to distribute the models. To
clarify, a lot of these models are wired up in what many people will not
agree to as "code." Would you call a worksheet with* equations* in several
cells "code"? Assume for simplicity that all of the equations and supporting
macros have been written by me. Also, all of the model can be ported to Open
Office, but I'd still consider the model (equations+macro code) as the
product of distribution, which does not fit the conventional understanding
of "code."

Chuck - thank you for the pointers about academic patents.

*Secondly, at least here in the USA, pure facts / data are not protectable,
but  creative arrangements of the data and the code written for your macros
likely are.  This said, almost all programs consist of a combination of code
and data, and existing licenses should be applicable to your case without
needing to create a new license.
*

I like the term "creative arrangements of data" - what I'm trying to call a
model is technically a worksheet with such creative arrangements, that make
it easy for a decision-maker to input data, and then computes what the best
decision should be for a particular context that the model applies to.

I would like to be OSI-compatible, so with Creative-Commons, I might go with
the ShareAlike version (and not add any other restrictions), or perhaps even
dedicate it to the public domain. But I'm thinking that there might be many
like me who have done a lot of creative work on spreadsheets, that is
actually quite helpful for management decisions and goes far beyond data -
this is about capturing a thought process.

So the key question is, is there any precedent on treating "worksheet
equations+macros+arrangement of cells" as code? If not, what if we define a
"spreadsheet model" as such, and create an OSI-compatible license for such
models?

Thanks in advance for your reflections.

Regards,
Somik

On Fri, Mar 5, 2010 at 9:18 AM, Somik Raha <somikr@gmail.com> wrote:

> Kevin, Chuck, thank you for your responses. I am familiar with the GPL and
> other open-source code licenses and I do plan to distribute the models. To
> clarify, a lot of these models are wired up in what many people will not
> agree to as "code." Would you call a worksheet with* equations* in several
> cells "code"? Assume for simplicity that all of the equations and supporting
> macros have been written by me. Also, all of the model can be ported to Open
> Office, but I'd still consider the model (equations+macro code) as the
> product of distribution, which does not fit the conventional understanding
> of "code."
>
> Chuck - thank you for the pointers about academic patents.
>
> *Secondly, at least here in the USA, pure facts / data are not
> protectable, but  creative arrangements of the data and the code written for
> your macros likely are.  This said, almost all programs consist of a
> combination of code and data, and existing licenses should be applicable to
> your case without needing to create a new license.
> *
>
> I like the term "creative arrangements of data" - what I'm trying to call a
> model is technically a worksheet with such creative arrangements, that make
> it easy for a decision-maker to input data, and then computes what the best
> decision should be for a particular context that the model applies to.
>
> I would like to be OSI-compatible, so with Creative-Commons, I might go
> with the ShareAlike version (and not add any other restrictions), or perhaps
> even dedicate it to the public domain. But I'm thinking that there might be
> many like me who have done a lot of creative work on spreadsheets, that is
> actually quite helpful for management decisions and goes far beyond data -
> this is about capturing a thought process.
>
> So the key question is, is there any precedent on treating "worksheet
> equations+macros+arrangement of cells" as code? If not, what if we define a
> "spreadsheet model" as such, and create an OSI-compatible license for such
> models?
>
> Thanks in advance for your reflections.
>
> Regards,
> Somik
>
> On Thu, Mar 4, 2010 at 4:46 PM, Chuck Swiger <chuck@codefab.com> wrote:
>
>> Hi, Somik--
>>
>>
>> Somik Raha wrote:
>>
>>> Hi Folks,
>>>
>>> I'm a Phd student in the area of decision analysis and have been making
>>> decision models for public safety that I'd like to put out there for
>>> administrators working on such issues, to help clarify their thinking. Most
>>> of this work is in the form of Excel models. I want to go the open-source
>>> route - but I don't know of any license that can be put on excel models. The
>>> closest seems to be a modified form of the LGPL for Excel macros. The
>>> intellectual property that I want to protect is in the wiring of the
>>> spreadsheet and sometimes in the macros.
>>>
>>> Would love to hear suggestions. Here are the two alternatives I've come
>>> up with so far:
>>> 1) Come up with a new open-source license for the content of spreadsheet
>>> files in general
>>> 2) Come up with a new open-source license for Decision Models (this
>>> includes the idea of the model, as represented in powerpoint slides,
>>> spreadsheet models or code).
>>>
>>
>> First, you should review your student handbook and discuss this with your
>> PhD advisor to be sure you are the one to make the decision.  You might
>> discover that your university claims ownership of IP developed by students
>> using their facilities.  (Ask first, and get it in writing beforehand to
>> avoid surprises.)
>>
>> Secondly, at least here in the USA, pure facts / data are not protectable,
>> but  creative arrangements of the data and the code written for your macros
>> likely are.  This said, almost all programs consist of a combination of code
>> and data, and existing licenses should be applicable to your case without
>> needing to create a new license.
>>
>> Which licenses you might want to consider depends on whether you want to
>> make this freely available without restrictions (in which case, simple
>> permissive licenses like BSD/MIT are good, the Academic Free License is also
>> a possibility) or if you want to copyleft the licensed material under the
>> GPL/LGPL; the creative commons licenses are another possibility, except that
>> they mostly contain more restrictions like "no derivative works" or
>> "non-commercial" which prevent them from being OSI open source.
>>
>> If your circumstances are truly unique, and no existing license seems to
>> suit your needs, then you might consider writing your own license, but
>> please avoid doing that unless you really need to.  :-)
>>
>> Regards,
>> --
>> -Chuck
>>
>
>
>
> --
> Somik Raha
> Doctoral Candidate
> Decision and Risk Analysis
> Dept. of Management Science & Engg.
> Stanford University
> (650) 450-8246
> somik@stanford.edu
> http://www.stanford.edu/~somik/
>
> ************************************************
> The essence of the moral reformer and of the creator in the field of art
> lies in not adjusting to the predominant norms or tastes, but unfurling the
> flag of what "ought to be" over and above people's preferences.
> -- Risieri Frondizi, in "What is Value?", Pg. 30 (1971)
> ************************************************
>



-- 
Somik Raha
Doctoral Candidate
Decision and Risk Analysis
Dept. of Management Science & Engg.
Stanford University
(650) 450-8246
somik@stanford.edu
http://www.stanford.edu/~somik/

************************************************
The essence of the moral reformer and of the creator in the field of art
lies in not adjusting to the predominant norms or tastes, but unfurling the
flag of what "ought to be" over and above people's preferences.
-- Risieri Frondizi, in "What is Value?", Pg. 30 (1971)
************************************************


Kevin, Chuck, thank you for your responses. I am familiar with the GPL and other open-source code licenses and I do plan to distribute the models. To clarify, a lot of these models are wired up in what many people will not agree to as "code." Would you call a worksheet with equations in several cells "code"? Assume for simplicity that all of the equations and supporting macros have been written by me. Also, all of the model can be ported to Open Office, but I'd still consider the model (equations+macro code) as the product of distribution, which does not fit the conventional understanding of "code."

Chuck - thank you for the pointers about academic patents.

Secondly, at least here in the USA, pure facts / data are not protectable, but  creative arrangements of the data and the code written for your macros likely are.  This said, almost all programs consist of a combination of code and data, and existing licenses should be applicable to your case without needing to create a new license.

I like the term "creative arrangements of data" - what I'm trying to call a model is technically a worksheet with such creative arrangements, that make it easy for a decision-maker to input data, and then computes what the best decision should be for a particular context that the model applies to. 

I would like to be OSI-compatible, so with Creative-Commons, I might go with the ShareAlike version (and not add any other restrictions), or perhaps even dedicate it to the public domain. But I'm thinking that there might be many like me who have done a lot of creative work on spreadsheets, that is actually quite helpful for management decisions and goes far beyond data - this is about capturing a thought process. 

So the key question is, is there any precedent on treating "worksheet equations+macros+arrangement of cells" as code? If not, what if we define a "spreadsheet model" as such, and create an OSI-compatible license for such models?

Thanks in advance for your reflections.

Regards,
Somik

On Fri, Mar 5, 2010 at 9:18 AM, Somik Raha <somikr@gmail.com> wrote:
Kevin, Chuck, thank you for your responses. I am familiar with the GPL and other open-source code licenses and I do plan to distribute the models. To clarify, a lot of these models are wired up in what many people will not agree to as "code." Would you call a worksheet with equations in several cells "code"? Assume for simplicity that all of the equations and supporting macros have been written by me. Also, all of the model can be ported to Open Office, but I'd still consider the model (equations+macro code) as the product of distribution, which does not fit the conventional understanding of "code."

Chuck - thank you for the pointers about academic patents.

Secondly, at least here in the USA, pure facts / data are not protectable, but  creative arrangements of the data and the code written for your macros likely are.  This said, almost all programs consist of a combination of code and data, and existing licenses should be applicable to your case without needing to create a new license.

I like the term "creative arrangements of data" - what I'm trying to call a model is technically a worksheet with such creative arrangements, that make it easy for a decision-maker to input data, and then computes what the best decision should be for a particular context that the model applies to. 

I would like to be OSI-compatible, so with Creative-Commons, I might go with the ShareAlike version (and not add any other restrictions), or perhaps even dedicate it to the public domain. But I'm thinking that there might be many like me who have done a lot of creative work on spreadsheets, that is actually quite helpful for management decisions and goes far beyond data - this is about capturing a thought process. 

So the key question is, is there any precedent on treating "worksheet equations+macros+arrangement of cells" as code? If not, what if we define a "spreadsheet model" as such, and create an OSI-compatible license for such models?

Thanks in advance for your reflections.

Regards,
Somik

On Thu, Mar 4, 2010 at 4:46 PM, Chuck Swiger <chuck@codefab.com> wrote:
Hi, Somik--


Somik Raha wrote:
Hi Folks,

I'm a Phd student in the area of decision analysis and have been making decision models for public safety that I'd like to put out there for administrators working on such issues, to help clarify their thinking. Most of this work is in the form of Excel models. I want to go the open-source route - but I don't know of any license that can be put on excel models. The closest seems to be a modified form of the LGPL for Excel macros. The intellectual property that I want to protect is in the wiring of the spreadsheet and sometimes in the macros.

Would love to hear suggestions. Here are the two alternatives I've come up with so far:
1) Come up with a new open-source license for the content of spreadsheet files in general
2) Come up with a new open-source license for Decision Models (this includes the idea of the model, as represented in powerpoint slides, spreadsheet models or code).

First, you should review your student handbook and discuss this with your PhD advisor to be sure you are the one to make the decision.  You might discover that your university claims ownership of IP developed by students using their facilities.  (Ask first, and get it in writing beforehand to avoid surprises.)

Secondly, at least here in the USA, pure facts / data are not protectable, but  creative arrangements of the data and the code written for your macros likely are.  This said, almost all programs consist of a combination of code and data, and existing licenses should be applicable to your case without needing to create a new license.

Which licenses you might want to consider depends on whether you want to make this freely available without restrictions (in which case, simple permissive licenses like BSD/MIT are good, the Academic Free License is also a possibility) or if you want to copyleft the licensed material under the GPL/LGPL; the creative commons licenses are another possibility, except that they mostly contain more restrictions like "no derivative works" or "non-commercial" which prevent them from being OSI open source.

If your circumstances are truly unique, and no existing license seems to suit your needs, then you might consider writing your own license, but please avoid doing that unless you really need to.  :-)

Regards,
--
-Chuck



--
Somik Raha
Doctoral Candidate
Decision and Risk Analysis
Dept. of Management Science & Engg.
Stanford University
(650) 450-8246
somik@stanford.edu
http://www.stanford.edu/~somik/

************************************************
The essence of the moral reformer and of the creator in the field of art lies in not adjusting to the predominant norms or tastes, but unfurling the flag of what "ought to be" over and above people's preferences.
-- Risieri Frondizi, in "What is Value?", Pg. 30 (1971)
************************************************



--
Somik Raha
Doctoral Candidate
Decision and Risk Analysis
Dept. of Management Science & Engg.
Stanford University
(650) 450-8246
somik@stanford.edu
http://www.stanford.edu/~somik/

************************************************
The essence of the moral reformer and of the creator in the field of art lies in not adjusting to the predominant norms or tastes, but unfurling the flag of what "ought to be" over and above people's preferences.
-- Risieri Frondizi, in "What is Value?", Pg. 30 (1971)
************************************************