From owner-chemistry %-% at %-% ccl.net Mon Oct 17 13:47:01 2011 From: "Christopher Cramer cramer**umn.edu" To: CCL Subject: CCL: Science code manifesto Message-Id: <-45664-111017122328-29372-5s0h/w8IAthsSxOVBjp/GQ*|*server.ccl.net> X-Original-From: Christopher Cramer Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=us-ascii Date: Mon, 17 Oct 2011 18:22:30 +0200 Mime-Version: 1.0 (Apple Message framework v1084) Sent to CCL by: Christopher Cramer [cramer|umn.edu] As I grow older, I carve notches into a stick each time Fortran vs. C++ and open-source vs. purchased executables comes up on CCL. It's getting to where I'll soon need a new stick. Personally, I think there is a middle ground in this doctrinaire argument (the latter one -- I simply ignore the Fortran vs. C++ one). Let us accept that all scientific reports accepted by reputable disseminators of information should be reproducible, as a matter of principle. But, let's also be serious -- after you read a fascinating paper on some conclusions derived from a complicated femtosecond spectroscopy experiment, are you going to spend $2M to build your own seriously decked out laser table in order to verify that the critical Figure 6 of the paper is not fabricated? Are you going to insist that the author shut his or her lab down for a week while you come and demand your sacred right to reproduce the experiment on his or her existing table? At some level, if only for the sake of efficiency, one must have some trust in one's colleagues' scientific integrity -- and, in those rare instances where someone DOES fabricate results, if they are actually INTERESTING results, then some other group somewhere WILL put together the analogous laser and publish competing results and the community will sort it all out the way it did with cold fusion, polywater, plastic fantastic, etc. As for computational chemistry, my opinion would be that all theories/models MUST be reported in the literature in a sufficiently detailed manner that someone skilled in the art (to borrow Jim Kress' phrase taken from IP law) COULD implement them into code (assuming the goal is to report future results from that new theory/model). That implementation might be nowhere near as fast as a proprietary one, but it allows results to be checked. If results FAIL to agree with a proprietary code, that's certainly a legitimate thing to note in a publication and that will likely inspire a third party to try it -- or the for-profit organization to revisit its code -- and the community will sort it out. Advocacy for all code being open-source is a values-based activity -- and I would not want to imply that I denigrate the philosophy underlying such advocacy. I will say, however, that the success of the scientific endeavor does not depend so critically upon it that its value is self-evident. Chris -- Christopher J. Cramer Elmore H. Northey Professor University of Minnesota Department of Chemistry 207 Pleasant St. SE Minneapolis, MN 55455-0431 -------------------------- Phone: (612) 624-0859 || FAX: (612) 626-7541 Mobile: (952) 297-2575 email: cramer/./umn.edu jabber: cramer/./jabber.umn.edu http://pollux.chem.umn.edu (website includes information about the textbook "Essentials of Computational Chemistry: Theories and Models, 2nd Edition")