*From*: Michel Petitjean <petitjean.chiral__gmail.com>*Subject*: CCL: Science code manifesto*Date*: Wed, 19 Oct 2011 18:45:42 -0400

Sent to CCL by: Michel Petitjean [petitjean.chiral[#]gmail.com] My contribution to the thread: Most reviewers are even unable to check most basic math formulas, so I hardly imagine them reading thousands of source lines including library calls (my own ones are in fortran, and commented in French !), thus I agree with George Fitzgerald. Despite that I give for free some of my sources, I hardly imagine giving the result of many years of my work to somebody without any counterpart (what should think my employer, who pays me ?). Publishing clearly the algorithms and giving a free access to the binaries should suffice to check. And even in this situation, it can be a hard work: checking the result sometimes need to compute it with an other algorithm (when existing !!), and it is much work. Once programmed, checking that the result is the same may be more difficult than generating the result itself. Some examples illustrating various difficulties for checking the results: Example 1: generate a d-dimensional Voronoi diagram. Example 2: calculate the minimal radius cylinder enclosing N points. Example 3: calculate analytically the vdw surface of N intersecting spheres. People ready to do that are welcome to contact me. And please be sure that I would be delighted to learn about comparisons, at least for ex. 2 and 3. All my best, Michel Petitjean MTi, INSERM UMR-S 973, University Paris 7 35 rue Helene Brion, 75205 Paris Cedex 13, France. Phone: +331 5727 8434; Fax: +331 5727 8372 E-mail: petitjean.chiral**gmail.com (preferred), michel.petitjean**univ-paris-diderot.fr http://petitjeanmichel.free.fr/itoweb.petitjean.freeware.html 2011/10/19 Alexander Bagaturyants bagaturyants{=}gmail.com <owner-chemistry**ccl.net>: > > Sent to CCL by: "Alexander Bagaturyants" [bagaturyants-#-gmail.com] > Dear CCLers, > I believe that the current discussion becomes counterproductive. > Of course, the author may decide whether to give an access to one's code or not. It is the author's inherent right. > Actually, the code itself is necessary if (when) somebody wants to use it somehow in his/her future work. >> From the pure scientific point of view, we need only the results of calculations and their verification. > However, some authors are so inadequate that they even protest against any comparison or benchmarking of their codes and the results obtained. > I believe that the scientific community must strongly oppose such a position. It has nothing to do with an honest attitude to science at all. > Any code must be open to any comparison, verification and benchmarking, and the results of such a comparison, verification or benchmarking must be fully available to scientific community. >