I think it would be a good idea for a system like that to have a
short, but very clear, unambiguous definition for each algorithm,
something that could then be used to compare against the
implementation in whatever languages are used. Having a set of worked
examples would also be useful.
But what else is needed? There are other systems; for example, in the
"utpala snon po" there are listed sets of 50 and 13 pebbles. There are
also other groupings, such as 24, in other texts, such as the Baidurya
dkar po. Perhaps Tenzang or Jampa could advise which are the most
important ones to cover - a sort of priority list that can then be
approached. In doing so, it would also be a good idea to inform the
rest of us which texts are considered definitive, so that the final
source code can include references. I started doing this with my
calendar code a few years ago, and think I should do a lot more.
Starting to do so early with the astrological code would be a good
idea.
There is also the astrological data from India, which often tends to
get overlooked, and the horary/electional listing of good and bad
days, mansions, etc, etc. There is a lot of material, and I think
guidance from those of you that teach all this would be beneficial.
Anyway, those are some thoughts for now. I am now back in London and
should get back to doing some coding myself during the next few days -
the Sherab Ling calculations are my next goal...
Edward.