[Remind-Fans] PATCHES for moon illumination and moonrise/set times

[Stephen Morgan]

Hello again,

Actually, perhaps a better solution is as follows. Stellarium is a big,
active, GPL planetarium software project, and their code includes an
implementation of ELP 2000-82b. It wouldn't necessarily relieve you of any
due diligence you feel necessary, but I would at least have more confidence
in the code being well-maintained and free from bugs. I could rewrite the
function to use their implementation instead; how does that sound?

Their code can be downloaded from
https://sourceforge.net/p/stellarium/news/2016/12/stellarium-0151/
and is located in the archive at
stellarium-0.15-1/src/core/planetsephems/elp82b.{c,h}.

By the way, as a matter of etiquette, I didn't feel I should be spamming
the list with 3.0MB files, but telling people to download a 150MB archive
just to extract two files inside is also problematic. Let me know if I
should send the file myself when I've finished.

Best regards,
Stephen Morgan

On 7 February 2017 at 10:49, Stephen Morgan <morgan.stephen99 at gmail.com>
wrote:

> Hello everyone,
>
> I've written a couple of patches to add some moon-related functionality to
> remind. The first is a simple patch to create the moonillum function, which
> gets the percentage of the moon's surface which is illuminated at a given
> time.
>
> The second patch, which is a bit more troublesome, allows remind to get
> the times of moonrise and moonset on a given date, accurate to within 3
> minutes of the United States Naval Observatory moonrise/set tables (RMS
> variation 0.7 minutes, at least for the dates I've checked).
>
> The only problem is that the implementation I've used adds a significant
> amount of code, mostly copied verbatim from this implementation of ELP
> 2000-82B (https://github.com/variar/elp2000-82b). This adds 2.7MB of
> code, and increases the size of the resulting binary from ~500kB to ~3.1MB.
> This problem could be solved by using a less accurate algorithm which is
> sufficient for these purposes (using ELP 2000-82B is rather like
> sandblasting a soup cracker), but that would require a bit more effort to
> find and code than I'm perhaps willing to put in.
>
> Alternately, a preprocessor flag could be added exclude the bulky code if
> you don't care about the moonrise/set times.
>
> To save on space in this email I've omitted the ELP 2000-82B code from the
> patch. Just copy it directly into a subdirectory elp2000-82b in the src
> directory.
>
> I hope others find these useful!
>
> Best regards,
> Stephen Morgan
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.roaringpenguin.com/pipermail/remind-fans/attachments/20170208/91772323/attachment.html>