Witam, IMHO the A&A editors are rather confused - here's their letter in response to our algebra paper http://de.arxiv.org/abs/astro-ph/0409533.
Why did they write such a confused letter? IMHO that sort of speculation is probably better not discussed publicly. The fact is that they sent me a confused letter, and IMHO there should be a reply.
Note that our paper does not contain any reference to software at all.
From aanda.paris at obspm.fr Fri Oct 1 17:28:03 2004 Date: Wed, 29 Sep 2004 13:50:39 +0200 From: aanda.paris at obspm.fr To: boud at astro.uni.torun.pl Subject: AA/2004/2055
29/09/2004
Dr Boudewijn F. Roukema boud at astro.uni.torun.pl
Our Ref. : AA/2004/2055
Dear Boud,
We are sorry that we cannot consider your paper for publication in A&A, since it is based on the HealPix software, that has been developped by Gorski et al, and is still actively developped and improved, and is widely distributed on the web (now on the ESO web site), and is widely used. It is distributed freely, with all the code sources (in fortran 90 or in C, etc..), and therefore your publication could be considered as a plagiat. The authors themselves could have done this paper, and if they have not judged it worth while, you have just to convince them to do so, or at least collaborate with them to do so.
A paper like this, with them as co-authors, would not have this problem.
But apparently they judge that the publication on the web as it is now is sufficient,
with best wishes,
The Editors
For the moment this is only a *proposal* of a reply to A&A. However, it is a public discussion with the aim of finding consensus using the formal consensus method. Having a secret discussion is pointless when we are doing public science and education.
Anyway, here's the proposal. Comments, suggestions, welcome.
Deadline: Fri 8 October 2004
(of course, if anyone needs a longer deadline, please say so).
pozdr boud
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%% PROPOSED REPLY TO A&A %%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%% DEADLINE fri 8 oct 2004 %%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Dear Editors,
On Wed, 29 Sep 2004 aanda.paris at obspm.fr wrote:
29/09/2004
Dr Boudewijn F. Roukema boud at astro.uni.torun.pl
Our Ref. : AA/2004/2055
Dear Boud,
We are sorry that we cannot consider your paper for publication in A&A, since it is based on the HealPix software, that has been
The second half of the above statement is false. Our paper contains algebra regarding a pixelisation system which is independent of any software implementation. It is not based on the HealPix software.
developped by Gorski et al, and is still actively developped and improved, and is widely distributed on the web (now on the ESO web site), and is widely used. It is distributed freely, with all the
This status of the Gorski et al software package is irrelevant regarding the scientific value of the algebra - "ideas, principles, algorithms" - behind the pixelisation system. This is independent of any software implementation.
Nevertheless, since you refer to that software, you might like to note that the Gorski et al software is *not* distributed under a free software licence (see http://www.gnu.org). It is formally correct that it is distributed at zero cost, but it is *not* "free software".
The Gorski et al software licence may not be redistributed without the permission of the authors. This violates software freedoms 2 and 3:
http://www.gnu.org/philosophy/free-sw.html
You are certainly aware that Astronomy & Astrophysics is a European journal, and many European governments are now committed to using free software. One of the most powerful elements of free software is the GNU General Public Licence, issued by the Free Software Foundation. I think it is relevant in this case to quote the GNU Manifesto, which you may not have read:
http://www.gnu.org/gnu/manifesto.html
; "Don't programmers deserve a reward for their creativity?" ; ; If anything deserves a reward, it is social contribution. Creativity ; can be a social contribution, but only in so far as society is free to ; use the results. If programmers deserve to be rewarded for creating ; innovative programs, by the same token they deserve to be punished if ; they restrict the use of these programs.
(in French: http://www.gnu.org/philosophy/free-sw.fr.html )
Gorski et al restrict the use of their programs.
Again, however, this is unrelated to our paper, which has nothing to do with software: it only regards algebra.
code sources (in fortran 90 or in C, etc..), and therefore your publication could be considered as a plagiat. The authors themselves
Our paper contains no reference to software, so it is difficult to understand your reference to "plagiat".
Please consider the COUNCIL DIRECTIVE of 14 May 1991 on the legal protection of computer programs (91/250/EEC) of THE COUNCIL OF THE EUROPEAN COMMUNITIES,
http://www.scaramanga.co.uk/archives/directive-91-250-EEC.html
: Whereas, for the avoidance of doubt, it has to be made clear that only : the expression of a computer program is protected and that ideas and : principles which underlie any element of a program, including those : which underlie its interfaces, are not protected by copyright under : this Directive; : : Whereas, in accordance with this principle of copyright, to the extent : that logic, algorithms and programming languages comprise ideas and : principles, those ideas and principles are not protected under this : Directive; : : : 2. Protection in accordance with this Directive shall apply to the : expression in any form of a computer program. Ideas and principles : which underlie any element of a computer program, including those : which underlie its interfaces, are not protected by copyright under : this Directive.
There is no computer program *expressed* in the submitted paper, though there are clearly ideas in the paper which *can* be expressed in a computer program. If our paper is correct, and this can only be independently decided by an independent referee, then the ideas in our paper must correspond to the non-free software package distributed at ESO.
However, this is the task of the referee to decide, not the Editors.
You might also like to note the point of view of Numerical Recipes:
http://www.library.cornell.edu/nr/bookfpdf/f0-1.pdf
; Copyright does not protect ideas, but only the expression of those ; ideas in a particular form. In the case of a computer program, the ideas ; consist of the program's methodology and algorithm, including the ; necessary sequence of steps adopted by the programmer. The expression of ; those ideas is the program source code (particularly any arbitrary or ; stylistic choices embodied in it), its derived object code, and any ; other derivative works. ; ; If you analyze the ideas contained in a program, and then express those ; ideas in your own completely different implementation, then that new ; program implementation belongs to you. That is what we have done for those ; programs in this book that are not entirely of our own devising.
Note that here, Numerical Recipes state that they have used the ideas and algorithms of computer programs devised by others - "not entirely of our own devising" - but have implemented them in their own style.
Yet let us return to our article: it does not concern software - it only concerns the solution to a system of coordinates, which the authors chose to publish in 1999 without including the full algebra of the system. Calculating the algebra requires a bit of work - that is what is presented in the paper so that other cosmologists can profit from this without having to rederive it from scratch.
could have done this paper, and if they have not judged it worth while, you have just to convince them to do so, or at least collaborate with them to do so. A paper like this, with them as co-authors, would not have this problem. But apparently they judge that the publication on the web as it is now is sufficient,
We did propose co-authorship to the authors of the software on 8 Sep 2004:
http://cosmo.torun.pl/pipermail/cosmo-torun/2004-September/000522.html
and we pointed out that if we did not hear any response by 22 Sep, we would presume that they were uninterested.
On 21 Sep, some of those authors chose to publish a paper similar to the one we had already sent them:
http://de.arxiv.org/abs/astro-ph/0409513
In this paper, BTW, in formulae (13) and (14), there is a clear inspiration from our own notation in Sections 2.2.1 and 2.2.2 of our paper, but without citation.
However, back to the subject.
Our paper is based on solving the unsolved definitions in Gorski et al 1999a, 1999b.
It is completely independent of any software. Moreover, the effect is to make it possible for cosmologists to implement software solutions for a pixel-coordinate system which monopolises cosmological observational data and presently is not implemented in free software.
The algebraic solution to this pixel-coordinate system has not been published in any astronomical journal.
It is difficult to understand how Astronomy & Astrophysics can refuse to publish original work which is useful and practical for cosmologists wishing to avoid the use of non-free software.
The opinions of the authors of non-free software are irrelevant for Astronomy & Astrophysics.
Any article published in Astronomy & Astrophysics must contain a bibliography.
What is the meaning of this bibliography?
In many cases when an article is cited, it means that the authors of the article have used some of the ideas and algorithms in the cited articles, and it is considered normal scientific practice to give credit to those authors of the cited articles, while nevertheless showing how those ideas can be improved and/or corrected, and possibly arguing that the ideas in the cited article are fundamentally wrong.
What does A&A normally do when an article is accompanied by a bibliography (i.e. nearly every article, IMHO), especially in the case where the authors of the submitted article and those of the cited article may have different points of view on the correctness or usefulness of those ideas or how to implement them in scientific research?
IMHO, it sends the article to an independent referee.
It does not tell the authors of the submitted article that use of ideas which are, in principle, implicitly present in another article could be considered "plagiat" and that consent of the authors of the cited article is necessary for publication of the submitted article.
At least, that is my understanding of how A&A works.
Software has no extra copyright protection to text.
But again, it should be made clear: if you read our submitted paper, you will see that it is independent of any software. Please read it.
Best regards Boud
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%% END OF PROPOSED REPLY TO A&A %%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%% DEADLINE fri 8 oct 2004 %%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
We are sorry that we cannot consider your paper for publication in A&A, since it is based on the HealPix software, that has been
^^^^^^^^^^^
The second half of the above statement is false. Our paper contains algebra regarding a pixelisation system which is independent of any software implementation. It is not based on the HealPix software.
^^^^^^!!!^^^^^^ [...]
My (AMr's) 0.03 PLN.
A&A, following Gorski et al., argue that you merely "disassembled" the HEALPix code and expressed it in algebraic form. _Nothing_ more than that. If this was true, they would be right: such a practice would be, IMHO, a plagiarism. In that case the fact that...
The algebraic solution to this pixel-coordinate system has not been published in any astronomical journal.
... and so you made a "translation" from the programming language(s) used in HEALPix to the algebraic formalism just to show how HEALPix works, does not help much.
Therefore you should try to convince A&A that your approach presented in your article is different. Or to be more precise: that, say, 80% of the intellectual content of AA/2004/2055 is common to that in HEALPix software but the remaining 20% is completely genuine.
Proving this, although may be difficult, is doable. To this end you should _insist_ that an independent referee analyses and compares the code and your algebra and judges to what extent your algebra is genuine. But....
Finding a referee that would be willing to wade through the HEALPix code to find out if, indeed, your algebra and the HEALPix code present different approaches would be extremely difficult, IMHO. So, what you have to do is to help A&A and the (would be) referee. To this end you might send them a detailed list of differences between the code and your algebra.
But again, it should be made clear: if you read our submitted paper, you will see that it is independent of any software. Please read it.
Sorry, Boud, but the point is they won't see it unless _you_ PROVE it. Your statement "it is independent of any software" is a typical "because I say so" statement and as such will hardly convince anyone, I'm afraid.
-- Andrzej
hi Andrzej, everyone,
On Mon, 4 Oct 2004, Andrzej Marecki wrote:
We are sorry that we cannot consider your paper for publication in A&A, since it is based on the HealPix software, that has been
^^^^^^^^^^^The second half of the above statement is false. Our paper contains algebra regarding a pixelisation system which is independent of any software implementation. It is not based on the HealPix software.
^^^^^^!!!^^^^^^[...]
My (AMr's) 0.03 PLN.
A highly valuable 0.03 PLN :).
A&A, following Gorski et al., argue that you merely "disassembled" the HEALPix code and expressed it in algebraic form. _Nothing_ more than that. If this was true, they would be right: such a practice would be, IMHO, a plagiarism. In that case the fact that...
In terms of publishing in a scientific journal, yes, i agree entirely, if the "if" were true which i can understand is A&A's worry.
The algebraic solution to this pixel-coordinate system has not been published in any astronomical journal.
... and so you made a "translation" from the programming language(s) used in HEALPix to the algebraic formalism just to show how HEALPix works, does not help much.
Equation (7), equations (24)-(34), and all the words which explain why equations have certain values, are not the sort of thing normally included in a computer program - unless it is very well documented.
All the same, playing devil's advocate, i accept that the devil may feel this way.
Therefore you should try to convince A&A that your approach presented in your article is different. Or to be more precise: that, say, 80% of the intellectual content of AA/2004/2055 is common to that in HEALPix software but the remaining 20% is completely genuine.
:(
My hope is (was?) to show to A&A exactly what is in the content of the article: it comes directly from Gorski et al 1999a/b, it is independent of whatever may or may not be in the code.
Proving this, although may be difficult, is doable. To this end you should _insist_ that an independent referee analyses and compares the code and your algebra and judges to what extent your algebra is genuine. But....
Something like this seems reasonable to me: surely A&A should be willing to ask an independent referee to look at this.
Finding a referee that would be willing to wade through the HEALPix code to find out if, indeed, your algebra and the HEALPix code present different approaches would be extremely difficult, IMHO. So, what you have to do is to help A&A and the (would be) referee. To this end you might send them a detailed list of differences between the code and your algebra.
Ouch :(. i'd prefer *not* to look at the HEALPix code.
But again, it should be made clear: if you read our submitted paper, you will see that it is independent of any software. Please read it.
Sorry, Boud, but the point is they won't see it unless _you_ PROVE it. Your statement "it is independent of any software" is a typical "because I say so" statement and as such will hardly convince anyone, I'm afraid.
I'm right! I'm right! I'm right! I'm right! Do you believe me now?
OK, i see your point.
Since what we need are independent referees, here's an idea: add to the comment on astro-ph something like:
"Seeking independent referees to compare content of paper with content of non-free software package HEALPix. Please post your independent analysis on this publicly archived mailing list: http://cosmo.torun.pl/mailman/listinfo/cosmo-torun"
But i'm not sure there's a lot of point: aanda would probably not be convinced by people who spontaneously post opinions. And not many people are motivated to be serious referees anyway since we get "brownie points" for published articles, not for how many articles we referee.
My feeling now is that i should give a clear reply to A&A, but much, much shorter and showing that i understand their concern. If they are unwilling to send to a referee, following my response, so be it.
Let's see...
chwileczkę... boud
My hope is (was?) to show to A&A exactly what is in the content of the article: it comes directly from Gorski et al 1999a/b, it is independent of whatever may or may not be in the code.
Well, I personally *believe* you were ONLY inspired by Gorski et al 1999a/b, and has NOT used their code. However, since you *had HAD* the code before you submitted the paper to A&A, the HEALPix group - as we all know - claim that you were _not_ smart enough to invent the algebra (almost) from scratch but you _were_ smart enough to derive the algebraic equations from the code and so you stole their intellectual property.
Ouch :(. i'd prefer *not* to look at the HEALPix code.
Fine! So now you have to be smart again and show that... it is not possible to derive the algebra - as it stands in your article - from the HEALPix code. Either because the HEALPix code is too obscure or badly documented or your algebra is different or... I don't know.
Anyway, if you want to win this argument you have to show a proof - or better to say - an alibi that their claims (that you have "undone" their code back to the algebra) are void.
It may be hard to find a clever way to show such an alibi but you have no other way, I'm afraid. This is because when someone is accused she/he has committed a plagiarism, this is the alleged plagiarist duty to bring the evidence that no plagiarism took place.
a.
hi again,
On Mon, 4 Oct 2004, Andrzej Marecki wrote:
My hope is (was?) to show to A&A exactly what is in the content of the article: it comes directly from Gorski et al 1999a/b, it is independent of whatever may or may not be in the code.
Well, I personally *believe* you were ONLY inspired by Gorski et al 1999a/b, and has NOT used their code. However, since you *had HAD* the code before you submitted the paper to A&A, the HEALPix group - as we all know - claim
Well, the code had been present at CA-UMK, anyway - maybe there's still a copy somewhere. And usage of extracts for research/teaching is what's called "fair use" in USA copyright and "fair dealing" copyright in some other countries - see e.g. the wikipedia.
that you were _not_ smart enough to invent the algebra (almost) from scratch but you _were_ smart enough to derive the algebraic equations from the code and so you stole their intellectual property.
Ouch :(. i'd prefer *not* to look at the HEALPix code.
Fine! So now you have to be smart again and show that... it is not possible to derive the algebra - as it stands in your article - from the HEALPix code. Either because the HEALPix code is too obscure or badly documented or your algebra is different or... I don't know.
The whole point of the article is that it's self-contained. There is no statement which comes "from nowhere".
The algebra is about *why* certain expressions should be used - there is no particular need for these discussions and derivations in the code, since the code only needs to *use* the algebraic expressions, not prove to the user that they are correct. The work in the paper is to *show* which expressions should be used (or one choice of expressions to be used, anyway) and why.
Anyway, if you want to win this argument you have to show a proof - or better to say - an alibi that their claims (that you have "undone" their code back to the algebra) are void.
It may be hard to find a clever way to show such an alibi but you have no other way, I'm afraid. This is because when someone is accused she/he has committed a plagiarism, this is the alleged plagiarist duty to bring the evidence that no plagiarism took place.
On the contrary: "innocent until proven guilty."
Anyway, thoughts on my new proposed text are welcome: surely the fact that Gorski et al published in reponse to our paper, and that they published with a different style of algebra, is sufficient proof?
pozdr boud
OK, any thoughts on this version?
i've reset the deadline to one week, since the text is considerably changed, and it's clear that andrzej's devil's advocate point of view is difficult to respond to...
pozdr boud
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%% PROPOSED REPLY TO A&A %%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%% DEADLINE mon 11 oct 2004 %%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Dear Editors,
On Wed, 29 Sep 2004 aanda.paris at obspm.fr wrote:
29/09/2004
Dr Boudewijn F. Roukema boud at astro.uni.torun.pl
Our Ref. : AA/2004/2055
Dear Boud,
We are sorry that we cannot consider your paper for publication in A&A, since it is based on the HealPix software, that has been
I am extremely upset by the second half of the above statement: it is false. Such accusations should not be made lightly.
Our paper contains algebra regarding a system of pixelising the sphere which is independent of any software implementation. It is not based on the HealPix software. It is based on the cited paper: Gorski et al. (1999a/b), which presents the outline of a pixelisation system but not its full solution. The solution requires a small, but original and highly necessary (given its non-availability to astronomers) amount of algebra and elementary calculus.
Nevertheless, it is understandable that you are worried that the derivation in our paper might be just an extract of the non-free-software (see below) HEALPix software package. In principle, some part of the solution in our paper probably is, in some indirect sense, implicitly present in the HEALPix package. However, in the context in which free software is receiving increasing support, even from governments in Western Europe, it is unreasonable to expect astronomers to search through non-free code in order to find the necessary algebra. Surely it is more useful for them to see the derivation directly so that they do not have to rederive it from Gorski et al. (1999a/b).
A similar situation occurs when, say, an article A is submitted to A&A which has content which solves an unsolved problem in an article B cited in the bibliography, but it is well-known that the authors of the cited article B are aware of the full solution to their problem; and it is also known that many astronomers C communicate directly with B, in private, but other astronomers D need to independently rederive the solution if they do not wish to enter into a direct working relationship with authors B.
Wouldn't it in this case be reasonable to send the article to an independent referee, independent of all astronomers known to be close colleagues of any of the authors A or B, while asking the referee to judge whether or not the content is independent of the non-freely distributed work of authors B?
Please note:
developped by Gorski et al, and is still actively developped and improved, and is widely distributed on the web (now on the ESO web site), and is widely used. It is distributed freely, with all the
The Gorski et al software is *not* distributed under a free software licence (see http://www.gnu.org). It is formally correct that it is distributed at zero cost, but it is *not* "free software"; it may not be redistributed without the permission of its authors. This violates software freedoms 2 and 3:
http://www.gnu.org/philosophy/free-sw.html (en) http://www.gnu.org/philosophy/free-sw.fr.html (fr)
All four software freedoms (0, 1, 2, 3) are part of the fundamental basis of the GNU/Linux free software system.
could have done this paper, and if they have not judged it worth while, you have just to convince them to do so, or at least collaborate with them to do so. A paper like this, with them as co-authors, would not have this problem. But apparently they judge that the publication on the web as it is now is sufficient,
Surely the authors of a non-free software package related to a cited paper are not in a neutral position regarding algebra derived from that cited paper; they may (either correctly or wrongly) believe that any related work is necessarily unoriginal.
If a paper is submitted, which, for example, is stated to be based on Bloggs et al. (1995n) and it is suspected that the submitted paper contains similar content to Bloggs et al. (1996d), then is it reasonable to insist that the submitting authors collaborate with Bloggs et al. (1996d), even if Bloggs et al. "could have done" the submitted paper and "have not judged it worth while"?
Or would it not be more reasonable to send it to a referee independent of both the submitting authors and Bloggs et al. ?
Finally, you might also want to note that Gorski et al. themselves have implicitly *approved* the scientific value of publication of our paper.
On 8 Sep 2004, we proposed co-authorship to the HEALPix software authors (Krzysztof Gorski <Krzysztof.M.Gorski at jpl.nasa.gov>, Eric Hivon <efh at ipac.caltech.edu>, Frode Hansen <Frode.Hansen at roma2.infn.it>, "Benjamin D. Wandelt" <bwandelt at uiuc.edu>)
http://cosmo.torun.pl/pipermail/cosmo-torun/2004-September/000522.html
and we pointed out that if we did not hear any response by 22 Sep, we would presume that they were uninterested. We had a constructive discussion with Frode, and he chose not to be a co-author, at least pending possible future discussion. The others did not reply directly, but they replied one day before the 22 Sep deadline, by publishing a paper similar to the one we had already sent them:
http://de.arxiv.org/abs/astro-ph/0409513
In this paper, BTW, in formulae (13) and (14), there is an inspiration from our own notation in Sections 2.2.1 and 2.2.2 of our paper, but without citation.
Their derivation is clearly independent work from ours (while they were clearly motivated and inspired by our work, they were clearly capable of doing it independently); our work is clearly independent from theirs. However, the fact of their publishing (electronically, at least) in response to receiving our paper suggests that they consider the content valuable to the scientific community and *not* present in their non-free software package.
Their paper 0409513 is presented in ApJ referee format; would it be reasonable that one independent derivation is published in ApJ but a second independent derivation is *not* published in A&A?
Best regards Boud
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%% END OF PROPOSED REPLY TO A&A %%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%% DEADLINE mon 11 oct 2004 %%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
On Mon, 4 Oct 2004, Andrzej Marecki wrote:
Here's the devil's advocate speaking again...
[...]
I am extremely upset by the second half of the above statement: it is false. Such accusations should not be made lightly.
[...]
IMVHO, this wording is too strong.
IMHO (don't know about V, i try but i'm not always very successful...), you're right.
How about back to what it was before (i think)? Just:
"The second half of the above statement is false."
It's still strong, but their allegation is strong as well.
pozdr boud
-
Andrzej Marecki -
Boud Roukema