These are chat archives for ipython/ipython

15th
Nov 2017
Min RK
@minrk
Nov 15 2017 09:15
@parmentelat what's locale for this user? Is this in terminal IPython?
parmentelat
@parmentelat
Nov 15 2017 09:16
yes this is ipython in a terminal
Min RK
@minrk
Nov 15 2017 09:17
It could be related to the terminal locale
parmentelat
@parmentelat
Nov 15 2017 09:17
I don't know the locale for sure but we are in France..
Min RK
@minrk
Nov 15 2017 09:17
The user typed this by hand?
Or was it copied?
Do you have access to it now?
Running !locale in IPython should give a locale report
parmentelat
@parmentelat
Nov 15 2017 09:18
unfortunately no and we communicate through in a discussion forum on a mooc platform
so it's not exactly real time ..
Min RK
@minrk
Nov 15 2017 09:18
ok
parmentelat
@parmentelat
Nov 15 2017 09:19
I can ask him/her to run the locale if that's helpful, that'll come in eventually
Min RK
@minrk
Nov 15 2017 09:19
ok, thanks.
What version of Python as well?
parmentelat
@parmentelat
Nov 15 2017 09:20
we recommend because of all the f-strings in the course, but I'll ask that as well
recommend 3.6, that is
Min RK
@minrk
Nov 15 2017 09:21
My guess is that the root is an unusual encoding of the terminal itself
I'm not sure at which point that would be being mishandled, but I'd need to replicate it before tracking it down.
parmentelat
@parmentelat
Nov 15 2017 09:22
ok; I'll also suggest for the person to show up here directly, but only if he/she is already used to this kind of interaction
Min RK
@minrk
Nov 15 2017 09:22
sure
Joseph Weston
@jbweston
Nov 15 2017 11:32
Hi, I opened ipython/ipykernel#274 a couple of weeks ago, but now I'm not sure if I submitted it to the correct repo. ipython/ipython seems to get much more attention; should I resubmit the issue there?
Min RK
@minrk
Nov 15 2017 14:01
@jbweston that's a fine repo to open it on, we're just spread a little thin. Commenting now.
Does it require that you use both the asyncio stuff and the process pool?
Joseph Weston
@jbweston
Nov 15 2017 15:47
@minrk it does indeed require both. When using just a process pool executor everything is fine. It does indeed seem to be a weird interaction between the %gui asyncio eventloop integration and ProcessPoolExecutor. I tried digging around to find out how kernels get restarted to see if I could guess where some exception may be getting caught and not bubbling up and killing the kernel, but did not get very far
I will try latest master of jupyter_client and report back on the issue on GH
thanks!
Min RK
@minrk
Nov 15 2017 16:03
@jbweston thanks!