working with python 2 but not with python 3

[gyth]

skyfield 1.8

#!/usr/bin/env python2
from skyfield.api import load
ts = load.timescale()
eph = load('de421.bsp')
sun = eph['sun']
earth = eph['earth']
t = ts.utc(2000, range(12 ))
lat, lon, distance = earth.at(t).observe(sun).ecliptic_latlon()
print('ok')

it work fine

but by changing the first line by this

#!/usr/bin/env python3

it throw this error

Traceback (most recent call last):
  File "/home/********/testforbug.py", line 8, in <module>
    lat, lon, distance = earth.at(t).observe(sun).ecliptic_latlon()
  File "/usr/local/lib/python3.5/dist-packages/skyfield/positionlib.py", line 248, in ecliptic_latlon
    d, lat, lon = to_polar(vector.au)
  File "/usr/local/lib/python3.5/dist-packages/skyfield/functions.py", line 59, in to_polar
    x, y, z = xyz
ValueError: too many values to unpack (expected 3)

am I missing something ?
I believe this script was doing ok before...

[gyth]

#207 and #208 have high chance to be the same issue.

test cases with skyfield 1.8 and numpy 1.15.1

from skyfield.api import load
ts = load.timescale()
eph = load('de421.bsp')
earth = eph['earth']
t = ts.utc(2000, range(12))

python 3

 lat, lon, distance = earth.at(t).ecliptic_latlon('date')
 lat, lon, distance = earth.at(t).ecliptic_latlon(t)

those two line of code work fine.

but

 lat, lon, distance = earth.at(t).ecliptic_latlon()

keep rising the same error message as before.

---

python 2

it is the revers:

lat, lon, distance = earth.at(t).ecliptic_latlon()

work.
but

lat, lon, distance = earth.at(t).ecliptic_latlon(t)
lat, lon, distance = earth.at(t).ecliptic_latlon('date')

rise this error
TypeError: ecliptic_latlon() takes exactly 1 argument (2 given)

Does anyone can reproduce my result ?
or did I do something wrong ?

[brandon-rhodes]

My guess is that you might have a different version of Skyfield installed under each version of Python, since that method didn't used to take an optional second argument; that would explain why you're not seeing the same error (which is a real one that needs to be fixed!) under Python 2, which isn't even accepting the new 2nd argument.

Could you try printing skyfield.VERSION in both Pythons?

[brandon-rhodes]

I've just released Skyfield 1.9 with this fix. Thanks for the report!

[gyth]

yes indeed ... my mistake: 2 different version off skyfield on my system:
pip == pip3 != pip2

thank for skyfield.VERSION, I'll use it in future bug report

[brandon-rhodes]

Hopefully you can now upgrade both Pythons and get working Skyfields :)