Subject: Re: For Approval: NASA Open Source Agreement Version 1.1
From: Richard Schilling <rschi@rsmba.biz>
Date: Fri, 13 Feb 2004 11:57:59 -0800

On 2004.02.13 07:38 Ian Lance Taylor wrote:
[snip]

> > I believe that is a misguided concept in open source licensing that
> > some hold to.  Tracking the use of a product does not make a license
> > non-open source.  Open Source licensing deals with accessibility and
> > cost, but tracking, per se, is not even relevant to that
> > characteristic.  In fact, tracking the uses of open source is a
> *key*
> > marketing tool and the only way we can judge if an investment of
> time
> > into open source is paying off, is it not?
> 
> First let me say that I understand that NASA's proposed license
> doesn't require tracking, it merely encourages it.
> 
> I, and others, think that a tracking requirement would not be
> appropriate in an open source license.
> 
> 1) Tracking presumably requires reporting back to some organization.
>    What happens if that organization disappears?  Does it then become
>    impossible to distribute the code?  If it does, the code would
>    clearly no longer be open source.

Detailed tracking is done anyway through web logs.  And, companies 
dissappear all the time - no big deal there.  It's their data so they 
can discard it if they like.  I would rather know that more details 
about the product's use are being tracked than not.  When a company 
tracks the usage of their product they have an easier time gaining 
support from onlookers, which is good for the product.

I want to write to my congress people and make a case for NASA spending 
a lot of money on open source development.  It's more compelling to do 
that if I can point out where the product is being used.

Perhaps you're nervous because of the hype that licensing lawsuits get 
in the press?  The rest of the NASA license guarantees free, unhampered 
use so tracking does not present a discrimination issue.

Tracking information is meant to be held private, so it wouldn't be 
appropriate to release that inforamtion to the public anyway.  Tracking 
information is absoutely key to a developer's ability to guage the 
success of their product.  For example, if you have reliable numbers to 
compare your downloads with, say sales figures from a comperable 
well-known product (e.g. MS Office), then you can promote that product 
more effectively.


> 
> 2) It is generally considered to be desirable to permit open source
>    software to be used anonymously, such as by a dissident under your
>    least favorite form of government.  Arguably preventing the
>    possibility of anonymous use violates OSD #5.

Once a product under GNU is initially downloaded, the person can 
distribute it anonymously.  OSD #5 simply states non-descrimination - 
which means that you won't restrict the availability of your product to 
a particular group/caste/class/industry, etc...   I maintain that as 
long as non-restricted access is granted non-descrimination is complied 
with.  Besides, non-descrimination typically is not meant to ensure 
that you cannot know where your product is used.


> 
> 3) While free software is not identical to open source software, they
>    are generally congruous.  The FSF specifically forbids tracking:
>        http://www.fsf.org/philosophy/free-sw.html:

I think what I was pointing out before was that the standards for open 
source definitions are arbitrary and varied.  I would like to stick to 
opensource.org's criteria on this list.




> 
>        You should also have the freedom to make modifications and use
>        them privately in your own work or play, without even
>        mentioning that they exist. If you do publish your changes, you
>        should not be required to notify anyone in particular, or in
>        any particular way.

GNU specifically states that changes will be submitted back to the 
authors.  That requirement, if enforced, provides much more stringent 
tracking than what NASA proposes.  But, I also point out that this 
requirement in GNU is not enforced at all, generally, making that 
provision useless, and makes the language in the GNU license an empty 
requirement on its face.  For the GNU license it means the requirement 
might not be enforced by a judge even though the requirements are 
there.  Why put such a requirement in the license if it's not going to 
be enforced?

This is a key difference between NASA's license and the existing ones - 
enforcement of the requirement.  If NASA requires registration, then 
fine, but I expect NASA to make an effort to actually enforce the 
requirement.  And, the efforts to do that are clearly stated in the 
license itself, making the license more consistent between intent and 
implementation than many other licenses (on that point).

Richard


> 
> Ian
> --
> license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3
> 
> 
--
license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3