Hi, > On 04 Jul 2015, at 10:29, Jan Darowski <[email protected]> wrote: > > I guess I need to implement Boyle's law as soon as possible and only > test against > the original fortran code...
just a brief comment: Indeed testing against other code is quite essential (not only since our software might be giving life threatening advice when wrong) but also since this is what users will judge us on: Do we match the plans of whatever software they were using before (and if those disagree we should have a good explanation). The Fortran program is of course the gold reference (as in fact it is the definition of the model) but other commonly used software as well (Davide offered at some point to run some profiles with the GUE planner tool). And we should not only compare total run times but the whole profile (and of course small deviations will not be able to be avoided). What I am a bit puzzled about are the „conservatism“ factors some other programs offer. I have no idea which parameters those affect but that is another thing to investigate. Regarding the license (which is quite unclear for the fortran code): I think Jan was sure not to use the fortran code directly. So there should not be any copyright model there but we might want to say something like „Eric Baker’s VPM-B model“ somewhere. Best Robert
signature.asc
Description: Message signed with OpenPGP using GPGMail
_______________________________________________ subsurface mailing list [email protected] http://lists.subsurface-divelog.org/cgi-bin/mailman/listinfo/subsurface
