QDate range extended

[Jason Harris]

Hi,

On Wednesday 14 June 2006 Thiago Macieira wrote:
> -4712 is not arbitrary. And I thought you would be the first to realise
> what year that is: QDate allows you to go as far back as julian day 0.
>
> You cannot go even more backwards because QDate has an *unsigned* int as
> its only member. So the earliest it can see is julian day 0.
>
Yes I know it's JD=0...but the choice of unsigned int was still arbitrary, 
since there were, in fact, days before that.  I understand your argument 
about BIC in an inline function, but it doesn't make the choice for minimum 
date non-arbitrary.

> Are you sure all the internal computations you have in KStars are valid
> from the time of the dinosaurs until the Sun being no more?
>
No, of course not.  Some of the computations (especially the positions of 
comets, asteroids and Pluto) are totally meaningless after only a few 
centuries.  However, there are things that occur in the sky on the timescale 
of millenia; these are important concepts that I want people to be able to 
explore with KStars, such as the precession of the celestial pole (12,000 
year ago, Vega was the north star!), and the proper motion of stars (the user 
can see how distorted constellations get at a very remote date).  So, from 
the point of view of enabling these explorations, it's kind of weird to tell 
people they can only go forward in time to very remote dates, they can't go 
backward.

Anyway, that's all just to explain why I thought I needed the very remote 
dates.  I accept that it can't be changed, and since it's only needed for 
non-critical functions in one app, there's no point worrying about it.

I'll just keep my existing ExtDate* classes.  Actually, maybe I don't need 
to...since calendar dates become meaningless in the remote past anyway, I 
could switch over to a pure JD date when crossing some date threshhold (maybe 
1752), which would emphasize that calendar dates don't apply.  I could still 
keep the year, as a simple count of the number of earth-orbits.

> (In fact, does KStars account for the Sun's eventual expansion to a red
> giant?)

No, but I have toyed with the idea of writing a KDE app that runs numerical 
models of stellar interiors and evolution...

-- 
KStars: http://edu.kde.org/kstars
Community Forums: http://kstars.30doradus.org