[Kstars-devel] kdeedu/kstars/kstars

[Jason Harris]

CVS commit by harris: 

Two fixes related to rise/set/transit times:

(1) The rise/set/transit times are displayed in three places: the 
popup menu, the Details dialog, and in the WUT tool.  In the WUT, 
if the transit or set time is before the rise time, then the 
time is recomputed for the following day.  I added this 
provision to the popup menu and the Details dialog as well, so 
that the set and transit times are now consistent.

(2) Determining the rise/set/transit times is an iterative
process that recomputes the position of the object for the Julian 
Day corresponding to the first-guess rise/set/transit time, and 
then determines the rise/set/transit time for that revised 
position.  The function 
SkyObject::getNewCoords(jd,jd0,geo) is supposed to then restore 
the object's position to the current JD, under the assumption 
that the first parameter ("jd") is the current Julian Day value.  
However, this assumption was violated in the WUT, resulting in the 
restored position being incorrect.  This was especially a problem 
for the Moon; its restored position was off by over 10 degrees. 

I have replaced SkyObject::getNewCoords() with 
SkyObject::computeCoordsForJD(jd, geo).  This function simply stores
the object's original position in a local variable and resets it at 
the end, rather than recomputing it.

Fix (1) is backported to KDE_3_1_BRANCH and debian_branch; Fix (2) 
is not, because there is no WUT in those branches, so the code 
(while fragile) is not broken.

TODO: the rise/set/transit times can still be in disagreement by +/- 1 
minute between the three locations; this still needs investigating.

CCMAIL: kstars-devel at kde.org


  M +14 -3     detaildialog.cpp   1.25
  M +8 -0      kspopupmenu.cpp   1.12
  M +18 -22    skyobject.cpp   1.35
  M +4 -5      skyobject.h   1.31
  M +6 -4      wutdialog.cpp   1.16