Sean Curry Apr 1, 2017
On Sat, Apr 1, 2017 at 12:19 PM, Bill Gray pluto@... [find_orb] <find_orb@yahoogroups.com> wrote:Hello all,
I've not posted an updated Windows Find_Orb since last year's
Pi Day, 2016 March 14. The software has marched on quite a bit
since then, but this was only noticeable to Linux and *BSD users.
I've now wrestled both 32-bit and 64-bit Windows versions into
place. (And then got the documentation wrestled into place to
match; that took a while, which is why the version date is
2017 April 1.) I strongly recommend updating to the new version.
You can get them at the usual place,
http://www.projectpluto.com/ find_orb.htm#download
or download the source code from the GitHub page linked to from
the above URL and build the program yourself.
There are a lot of "under the hood" improvements, only visible
in that Find_Orb will get orbits in some situations where it wouldn't
have before. This is most apparent when it has to try to link one
short arc with observations made years or decades earlier or later;
it will usually find an orbit linking everything, without requiring
the human to fiddle about looking for a linkage.
The handling of "extensions" such as setting uncertainties is very
slightly different, but in a way that allows one to set them in your
astrometric data as COM (comment) lines. The benefit of this is that
you can store sigmas routinely without causing MPC to complain to you :
http://www.projectpluto.com/ findhist.htm#extensions
I highly recommend doing this. In theory, the "new" astrometry
format (proposed in 2015) will handle this for us. But that doesn't
mean we have to throw out uncertainty positional and photometric
uncertainty information. (I would have harped on this idea, here
and on MPML. But I foolishly thought that the "new" format would be
implemented much more quickly than it actually has been.)
I cleaned up a few error/warning messages, so that when the
program misbehaves, you'll probably have a better idea as to why and
what to do about it. Sigmas sometimes were turned off when they had
been computed and ought to have been shown.
I also found several ways to make the initial orbit determination
faster and more accurate. (It would be closer to the truth to say
that I found some stupid things that I'd done in the code and removed
them. But saying that I sped it up and made it more accurate does
sound better, and it _is_ also completely true.)
-- Bill