[okfn-discuss] about open access but not open access

[Marc Joffe]

William

Thanks for your well-considered comments.  Microsoft tools generate open
formats such as CSV and XML, so the concern listed in your second paragraph
is not applicable to my effort.  I can use Excel to gather data and then
migrate that data to MySQL without in any way compromising the open data
nature of my work. I certainly agree that if a purportedly open project uses
a proprietary format thereby compelling users to buy software to see the
results, that the project is not truly open.

I want to explain to the list why I feel so strongly about this and why you
may get the impression that I am making a moral argument.  I used to work
with financial modeling tools and data sets at a rating agency.  As you
know, rating agencies and financial models were blamed for the deep
recession that started in 2008. I believe that if the models and data had
been more open, they could have been subject to a wider peer review and thus
could have become more effective.

In 2011, I came across the book Macrowikinomics, which talked about
applications of wiki technology to various industries including finance. The
book spent a few pages favorably discussing a new firm called Open Models
Valuation Company (http://www.omvco.com), which was going to open up the
world of financial modeling. I was excited to learn about this innovation
and immediately contacted the company. Eventually a representative got back
to me.  During our chat, he advised me that the company had not published
any open source models and did not intend to do so.   

This seems to me a case of false advertising, and thus worthy of criticism.
Open software and data can address many of society's problems - including
bad credit models.  Organizations that talk the talk of openness, and then
fail to walk the walk should be called out by this community.

Cheers,
Marc

-----Original Message-----
From: William Waites [mailto:ww at styx.org] 
Sent: Saturday, May 04, 2013 1:09 PM
To: okfn-discuss at lists.okfn.org; marc at publicsectorcredit.org
Subject: Re: [okfn-discuss] about open access but not open access

On Sat, 4 May 2013 07:20:52 -0700, "Marc Joffe"
<marc at publicsectorcredit.org> said:

    > What tools one uses to
    > assemble the open data is not important (as long as the tools'
    > licenses do not restrict distribution of the work product).

Marc, I don't have a strong opinion about whether one should only write open
access things about open access. I think the moral argument is a bit silly
-- the justification should be that if you want your work to matter, to be
built upon, it has to be accessible. In scientific and quasi-scientific
disciplines where reproducability of results matters, the whole process must
be open otherwise if nobody can try the same experiment and come up with the
same results, it might all have been made up. No need for moral arguments
(which tend to get people's backs up anyways).

Now, the tools do matter. Because it means that without using those tools,
which I may not have access to (I don't),  I can't check your work. This
means I have to trust that you didn't make any errors in creating the data,
and if I do find errors I can't easily fix them in a way that maintains the
integrity of the process. This is very important. Otherwise the open data
comes out of a black box and we have no way of knowing what it means or to
what extent it can be relied upon.

-w