Community commentary: Encouraging open code in public procurement policies

[Ben Beeson]

Aloha,

	I can't resist this one. My $0.02 follows.  

	When it comes to research, the key thing is the ability to verify the 
results.  Any research paper worth the trouble to write (and read for 
that matter) ought to outline pretty explicitly how the researcher(s) 
arrived at their result.  In fact, if after reading that paper, you do 
not feel that you could replicate that study given sufficient resources, 
then the research has failed to express the discovery of new knowledge.  
This is because new knowledge isn't considered knowledge until it is 
either independently verified or explained so well that others accept it 
at face value.  Usually problems with explaining the instrument and 
method are vetted out during peer review, but not always. Those problems 
not fixed during peer review are  likely to be revisited later in 'the 
public eye' unless the papers are revised to explain the process better.  

	I agree wholeheartedly that public science should be very open.  That is 
the only way to really expose the findings of research to the kind of 
scrutiny it ultimately needs to succeed.  One recent good example of the 
type of scrutiny I refer to here is the selection process for the new 
American Encryption Standard. Some may argue that not enough time was 
offered for public scrutiny of this important tool, but the point is that 
many learned scholars had and still have  the opportunity to examine and 
test the proposed algorithms, something that will ultimately lead to 
increased public trust of the result of the selection process.

	More and more, I believe that public things must be "public."  The tax 
payers should have the opportunity to see what they have bought. I don't 
know if we can ever get there from here,  but I think that all public 
institutions should open their sources also. 

Respectfully,

Ben             


>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<

On 2/10/02, 10:54:48 AM, "Rod Gammon" <AEG-Inc at hawaii.rr.com> wrote 
regarding [luau] Re: Community commentary: Encouraging open code in public 
procurement policies:


> Aloha-

> My take, in reply to the comment at newsforge on requiring open source
> (subsequent to Warren's posted article):

> I am convinced that public science must be open source, this is the 
basic,
> modern principle of accountability or verifiable results. How convinced? 
My
> own scientific software is GPL (nlp stuff for Mandarin, try it you'll 
like
> it.) So at the least NSF and universities should require open source from
> their researchers. As researchers (those of us here) we should demand our
> colleagues use open systems, just as we demand that they declare all 
their
> other parameters in a study. In the case of universities that don't 
require
> open and help patent such software- I urge each of you to join me in
> demanding a rebate check on our share of tax dollars enjoyed by those
> institutions. Why pay someone to develop something they will charge us 
for
> later? Any free market capitalists in the house tonight? Now this is
> actually sort of in place already. Having worked on educational software
> under federal grant, many projects are granted to design an open 
framework
> (e.g., testing apparatus) and then various units (particular tests) can 
be
> later developed proprietary. It seems a standard model- open 
infrastructure,
> proprietary content. But it is not widespread, to my knowledge. Ideas on 
how
> to make it widespread?

> -rg


> ---
> You are currently subscribed to luau as: beesond001 at hawaii.rr.com
> To unsubscribe send a blank email to $subst('Email.Unsub')