[Ksummit-discuss] [CORE TOPIC] GPL defense issues

[Julia Lawall]

On Sat, 27 Aug 2016, Bradley M. Kuhn wrote:

> Greg KH wrote at 08:43 (PDT):
> > [CCFinder is a] much different tool than what James was referring to.  The
> > work that the LF has been doing is not based on CCFinder, and (in my
> > opinion) is much more powerful and provides much more information about
> > attribution of work and the proof of where all code changes came from.
>
> It's really tough to know what this new tool is based on, because it's not
> Open Source and Free Software (not yet anyway).  It's a tool that's been
> presented in talks, but the data and code behind it remains hidden.
>
> Both I and Karen have talked extensively on multiple occasions with dmg,
> the tool's author.  He verified for me explictly that CCFinderX remains
> indeed state-of-the-art (and still really hard to install ;).  He
> indicated that his system, while it may or may not using CCFinderX's code
> directly [0], has technology that's roughly the same.  Anyone who has worked
> with CCFinderX and reviewed dmg's talks can see the obvious similarity.
>
> The key new contribution it sounds like dmg has made is data curation.
> From his verbal description, it sounds like dmg has done a really excellent
> job reconstructing the history of contributions to Linux that don't appear
> directly in Git.  I'm excited to (eventually) see his data!
>
> In April, I of course asked dmg to make the tool and the data available
> publicly under an Open Source and Free Software license, and he said he will
> do it eventually, but for the moment only his University team and the Linux
> Foundation (as funders of the work) have access to it, and they aren't
> willing to share their software and knowledge at the moment with others.
>
> [0] dmg was cagey on all this because he's an academic and has not yet had
>     the work published academically, and he lives under a publish-or-perish
>     regime.  I'm sympathetic; he's between a rock and hard place.  Indeed,
>     I've been looking lately into how unfriendly CS academia still remains to
>     Open Source and Free Software that's developed in public -- and it's
>     still as bad as when I was in graduate school.  I had so hoped it'd
>     gotten better.

As an academic, I don't see the issue.  I've never rejected a paper
because the tool was already open source.  On the contrary, reviewers
increasingly complain that a tool is not available, and so they can't see
how it works.  I don't know who dmg is or in what community he wants to
publish, but the more typical reason for not making a tool available is
that the author is not comfortable with the current state, and doesn't
have time to put it into a state that he is proud of, or doesn't currently
have the resources to be able to manage outside contributions.  I guess
that the same happens in non-academic contexts as well.  Maybe this person
is concerned that if the tool is available, others will be able to collect
the data this person wants to collect.  But typically it is only the tool
developer who will initially have the time, motivation, vision, etc
necessary to collect and interpret the data in an interesting way.

julia

> --
> Bradley M. Kuhn
> President & Distinguished Technologist of Software Freedom Conservancy
> ========================================================================
> Become a Conservancy Supporter today: https://sfconservancy.org/supporter
> _______________________________________________
> Ksummit-discuss mailing list
> Ksummit-discuss at lists.linuxfoundation.org
> https://lists.linuxfoundation.org/mailman/listinfo/ksummit-discuss
>