contributed code - copyright

[Bill Hargrove]

Only when the reprints arrived this morning did I realize (with some
horror and outrage) that it was our paper, "Effects of fire on
landscape heterogeneity in Yellowstone National Park, Wyoming", to
which Bill Baker obliquely referred in his recent general complaint
about failure of persons using contributed code to properly cite his
work.

As Dr. Baker points out in his message, our Table 2 does contain a
number of landscape metrics.  What Dr. Baker fails to note, however, is
the final sentence of the heading of that same table, which reads
"Analyses were conducted using SPAN (Turner 1990)."  The bibliography
gives the complete citation of a reference describing in detail the
SPAN program.

I want all grassu readers to know that we did not, and do not, use the
r.le package of contributed codes for this or any other project.  This
is not because of their quality or characteristics; rather, it is
simply because we already have many other tools, like SPAN, that we
have developed in-house, which do the same kinds of things.  Because we
wrote these software tools ourselves, we have a better understanding of
their behavior, and more trust in the results that they produce.

Our paper describes work which begins with the classification and
interpretation of remote imagery of the Yellowstone fires.  This is the
context within which GRASS tools were used.  For landscape analysis, we
typically use programs like SPAN, RULE, TURTLE, and FRAGSTATS.  Those
with access to a forms-capable WWW browser like MOSAIC or netscape, may
wish to visit our on-line version of TURTLE, at URL:

http://egis.ornl.gov/

Visitors to the on-line turtle will find that it too calculates the same
general landscape metrics which we report in our paper.

Dr. Baker is mistaken in his assumption that, because we cite the GRASS
manual, and because we report some rather basic landscape metrics, we
have used his r.le package and therefore failed to give credit as
required.  Indeed, many of the basic landscape metrics that Dr. Baker's
r.le suite of programs calculate were developed by landscape ecologists
here at Oak Ridge.

In any case, Dr. Baker (and others) may rest assured that, should we
actually *use* their contributions in our future work, they *will* be
given appropriate credit.


--------------------------------------------------------------
  ___                   
  /  )      /  /      /  /                          
 /--<   o  /  /      /--/ __  __  __  __  _      _       
/___/__<__/__/__    /  /_(_<_/ <_<_/_/ <_(_)_\/_(/_     
                                  /  
                               <_/
--------------------------------------------------------------
Bill Hargrove
Oak Ridge National Laboratory
Environmental Sciences Division
P.O. Box 2008, M.S. 6038
Oak Ridge, TN  37831-6038
(615) 574-1902
(615) 576-8646 (fax)
hnw at mtqgrass.esd.ornl.gov
--------------------------------------------------------------
------ From BAKERWL at uwyo.edu

>I've contributed some code to grass (the r.le programs) not in an
>official release, but via the grass/incoming directory on the moon.
>There is a request in one of the README files that if the code is used
>that a published article we wrote about the programs be cited.  Recently,
>an article was published in the Journal of Vegetation Science (Vol. 5,
>Issue 5, p. 731-742) that uses GRASS to calculate a number of indices
of landscape structure (their Table 2).  So far as I am aware, these
>indices cannot be calculated in GRASS without using the r.le programs.
>The authors, however, do not cite our article as requested, but just
>cite GRASS in general.  
>
>Short of scientific etiquette, there may be nothing wrong with this,
>and I may even be wrong about interpreting this as anything but an
>oversight.  But, it does raise a concern in my mind that others who
>contribute code may share.  Some code is contributed without any 
>request for citation as there is nothing to cite, other code is
>associated with published work or documentation that could be cited.
>Do any code contributors have thoughts about how we can achieve some
>of the minor rewards of developing and contributing code (i.e., it
>gets used), while still having a little protection against use without
>citation?  Maybe this is only important to those of us in academia
>and elsewhere where citation is a measure of our value.  Perhaps
>I'm the only one concerned about this...

>       Bill Baker
>       Dept. of Geography
>       Univ. of Wyoming
>       Laramie, WY 82071