This thread -- no doubt tiresome, cryptic or otherwise objectionable to
many onlookers -- seems to be moving towards a satisfactory conclusion.
I have various comments nevertheless:
1. I gather from what Roy says that a fixed version of -mean2- will
shortly appear on SSC, that acknowledges the use of my -sxpose- and also
attributes it to me. Contrary to any impression gathered otherwise,
achieving that was not the purpose of my contributions to this thread.
Indeed I had not seen Roy's -mean2.hlp- when I started, because it was
not publicly visible, and so I was entirely in ignorance that it did not
contain such full acknowledgment.
Setting that aside, as Roy, I and everybody else I know agree, proper
attribution of others' work is natural and necessary, so thanks to him
for correcting his earlier oversight, very surprising as it was, given
his repeated insistence on this list on respecting previous authors'
rights.
2. Any distinction between (a) borrowing all the code from program
-foobar- as a block and (b) inserting a single call to -foobar- is
neither here nor there as far as intellectual property or respect of
authors' rights is concerned. Either way, apart from small issues of
efficiency, the effect is, or should be, the same as far as users are
concerned. Equally, the degree of intellectual dependence on others'
work is the same as far as borrowers are concerned. (If the block is
changed, the question changes.)
3. Although I've now said so repeatedly, my intent was not to allege
that Roy abused my code. I just dissociated myself from an explicit
statement in a posting of his that used my program that such code was
for academic purposes only. That is (again) nothing to do with
plagiarism or acknowledgment.
4. I am at a loss to know why some people want detailed rules for
borrowing, something that does not really need detailed rules. Don't
steal. Give acknowledgments and references if appropriate. I don't see
that any more is required, other than to recommend looking at examples
(my published packages will do...).
An analogy springs to mind: I have several years' experience in
maintaining the Statalist FAQ and of finding, unsurprisingly, that often
people don't read it or pay attention to it, but nevertheless that very
poor behaviour is very rare, so Statalist works, regardless. Indeed Roy
himself once said in
<http://www.stata.com/statalist/archive/2007-07/msg00985.html> that he
hadn't actually read the entire FAQ (he may have caught up since). So,
let's turn this round. Roy or anybody else is entirely free to post,
unilaterally, detailed guidelines for borrowing that others might wish
to consider, but they should be warned that many people won't feel to
compelled to oblige, or even to read them at all. After all, the great
thing about standards is that there are so many to choose from.
Nick
[email protected]
Roy Wada (3 posts)
#3
> the recent discussion, Nick Cox objected when it appeared his own
> codes were being abused (by me). Although that was not the case, Nick
In case anyone wants to formalize this notion of intellecutal property
standards:
Duplicating someone's program for duplication's sake is generally
frowned upon, and is best avoided. If someone needs to borrow
extensive sections of codes for legitimate reasons, then it should be
acknowledged in the help file to be distributed along with the
resulting program. Both the author and the source program should be
mentioned by name in the help file. Within the program the borrowed
codes should be commented upon. If the borrowing is quite extensive,
then attempts should be made to contact the writer of the source.
Among other things, this practice prevents abusive situations where
someone takes someone's else codes or programs and republishes them
under their own name. In academia such work would be considered a
plagiarized work. Just how much can be taken without the need to cite
is a judgement call, but at certain point it becomes rather obvious
when the acknowledgement must be made. This is esecially the case when
the duplicator is being confused with the originator. Consistently
failing to give proper attribution is almost guaranteed to cause
problems later on, especially when the original writer is still around
and active.
#2
>> Package -mean2- will be available from SSC when (and if) its help
>> file is revised to provide appropriate attribution information, per
intellectual property standards proposed by Roy Wada.
Done.
I should have done this sooner, publicly taking someone's codes. In the
recent discussion, Nick Cox objected when it appeared his own codes were
being abused (by me). Although that was not the case, Nick was quite
right in exerting his authorship. I feel the same way about my own
codes, and actually about all the codes being hosted at ssc.
I know Kit is busy, but I am sure he is aware that several programs at
ssc does not currently meet this standard.
#1
> Package -mean2- will be available from SSC when (and if) its help file
> is revised to provide appropriate attribution information, per
intellectual property standards proposed by Roy Wada.
>
> Kit
I am guessing Kit is referring to sxpose. The current version of
-mean2- does not contain any codes from sxpose. The current help file
merely says sxpose from ssc is required.
But I will be very happy to do this, given that ssc has adopted the
higher standards. Let's hope that this standard is applied to all
programs hosted at ssc, especially the ones that do contain the codes
from other sources.
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
