In python 3.6 version there is 'weight' attribute in Node class which has no effect in solution.

I like the fact that how stupid I feel after seeing other people's solution after solving the kata. One liner against my bare minimum required logn.

Half of the time spent was on understanding the description. I do not propose any suggestion to improve it because sometimes trying to figure out a little bit of unknown in this website was something we appreciated.

Ive just completed the task in C++ and nothing seems broken at least to me.

I would tag this puzzle instead of fundamentals tho.

nothing, obviously.

It was there for the solutions like this (with optimal transaction count but not minimum transaction amount):
2->1 30
3->2 20
4->2 25

One solution would be
1->2 30
2->3 45
3->4 25

Yeah, people paying exactly whay they owed is already satisfying this condition. However, the reverse is probably not true.

Both are approved now!

Done that! I may not be able to approve two translations in a row. I will try again later. Thank you for making clearer description as well.

I have taken the precaution of informing solvers if they come up with a better answer than the reference solution, asking them to raise an Issue with applicable information supplied.

Good idea!

Oh thanks for interest in the kata! I have no idea how javascript and haskell tests should be written; hence, if someone else accepts it I would be glad.

Moreover, I was thinking to add more fixed edge cases of the problem and make my solution clearer. I guess this might be the time. Also, random tests in python may not test some edge cases. In future, I will try to change that too.

Many inspiring solutions in python.

what?

In description it says "See http://en.wikipedia.org/wiki/SI_prefix for a full list of prefixes" and "For this exercise we just want units bigger than a meter, from meters up to yottameters, excluding decameters and hectometers."

It was clear to me.

It is all up to the author of the Kata. If he/she wants us to round we do :)