Dear Colleagues, I did several proofs as regards the fractional part of the JD calculated by the old MuniDOS. The only problem was the one discussed by Dr. Kato [vsnet-alert 3979], related to the mid exposure (you had to subtract, from the JD given by the software, a quantity, in minutes, equal to the mid-exposure, with this one in seconds: for example, if your exposure was of 40 seconds, you had to subtract 20 minutes, of course expressed in days). As regards the error of 36524 days (so close to the lenght of a year * 100!), I simply got only the fractional part of the calculated JD, adding the right value later (or, simply, at the beginning of the header of my report). I still have to try wih the new version of MuniDOS. Best regards, Gianluca Taichi Kato wrote: > > Re: [vsnet-alert 3984] On Munidos error > > Probably we had better move to vsnet-chat before the bug is completely > understood. > > > Actually the result was out by 36524 days, which is a recognisable number. > > But on checking the fractional part I see that was out by +0.00059 days, > > which is about 51 seconds and doesn't seem related to anything. > > There seems to be at least another unrecogniozed problem in Munidos :-( > > Is this error constant or randomly variable? I wonder whether JD > calculation is internally done using double precision floating points. > > Regards, > Taichi Kato -- *********************************************************************** * Gianluca Masi "Two things fill the soul * * Via Madonna de Loco, 47 with awe and reverence [...]: * * 03023 Ceccano (FR) the starry sky above me and * * ITALY the moral law within me" * * e-mail: gianmasi@fr.flashnet.it * * web: http://vsnet.eurolink.it/comets/ I. Kant * * Minor Planet Center Obs. Code: 470 - Ceccano * * Center for Backyard Astrophysics - Italy * ***********************************************************************