These are chat archives for ipython/ipython

10th
Dec 2014
Kyle Kelley
@rgbkrk
Dec 10 2014 02:37
Well well, Thomas and I are going to PyCon! Tutorial accepted.
Brian E. Granger
@ellisonbg
Dec 10 2014 03:40
Congrats!
Scott Sanderson
@ssanderson
Dec 10 2014 04:37
nice! are you just tutorialing about general IPython usage?
or something more specific?
Kyle Kelley
@rgbkrk
Dec 10 2014 04:40
Generally, how to become a power user of IPython while getting a smattering of IPython architecture.
Scott Sanderson
@ssanderson
Dec 10 2014 04:41
cool
Matthias Bussonnier
@Carreau
Dec 10 2014 14:33
@Hack-The-Paradise so
CodeMirror.requireMode('octave', function(){})
IPython.notebook.get_selected_cell().code_mirror.setOption('mode','octave')
works for me on 2.3.1
Capture d’écran 2014-12-10 à 15.32.55.png
Hack-The-Paradise
@Hack-The-Paradise
Dec 10 2014 14:42
i had a problem with pymatbridge
now i fixed that and i rerun the extension
and i see the manual highlighting
you also load the extension before this or not?
(i load the extension explicitly)
(you already answered to that, my bad, but it's strange cause i cannot see if i don't)
Matthias Bussonnier
@Carreau
Dec 10 2014 14:48
No I don't load anything.
I don't even have the extension installed.
nor matlab
Hack-The-Paradise
@Hack-The-Paradise
Dec 10 2014 14:49
ok, tried on incognito and worked
hm,,,
Matthias Bussonnier
@Carreau
Dec 10 2014 14:50
you should deactivate caching of your browser when dev tools are openned.
Hack-The-Paradise
@Hack-The-Paradise
Dec 10 2014 14:51
thy will be done
now i will try to find a consistent (automatic) solution
Dave Hirschfeld
@dhirschfeld
Dec 10 2014 15:29
Thanks @rgbkrk/@Carreau - setting frame-ancestors does the trick.
Kyle Kelley
@rgbkrk
Dec 10 2014 16:50
Awesome
Can I ask what your embedding is for @dhirschfeld, to help us with direction and documentation?
Scott Sanderson
@ssanderson
Dec 10 2014 18:34
Is there an existing supported way to get tracebacks to display when extension loading fails?
I opened up ipython/ipython#7158 with a fix that worked for me, but I'm wondering if there's already a supported way of doing it
Thomas Kluyver
@takluyver
Dec 10 2014 18:37
yes, you can pass exc_info=True to the logging call
Min RK
@minrk
Dec 10 2014 18:37
@ssanderson you could add exc_info=True
Thomas Kluyver
@takluyver
Dec 10 2014 18:37
...jinx
Jonathan Frederic
@jdfreder
Dec 10 2014 18:38
usually that's followed by "you owe me a soda"
Scott Sanderson
@ssanderson
Dec 10 2014 18:39
ah, welp...that's a much cleaner solution than what I did
Thomas Kluyver
@takluyver
Dec 10 2014 18:40
yep - want to update your PR to do that?
@jdfreder only if he speaks before someone says his name, I think.
Scott Sanderson
@ssanderson
Dec 10 2014 18:43
@takluyver just pushed that change instead
Thomas Kluyver
@takluyver
Dec 10 2014 18:43
thanks
Scott Sanderson
@ssanderson
Dec 10 2014 18:44
although looking at it now, I think I actually added the arg to the wrong log statement
Jonathan Frederic
@jdfreder
Dec 10 2014 18:44
@takluyver ahh I forgot about that twist
I wonder if typing in gitter counts..
Thomas Kluyver
@takluyver
Dec 10 2014 18:48
@ssanderson yep, I just thought the same thing
Scott Sanderson
@ssanderson
Dec 10 2014 18:48
that's what I get for pushing without actually testing, since I had already fixed the issue that prompted me to make the original PR :)
Thomas Kluyver
@takluyver
Dec 10 2014 18:49
heh
I think there are often bugs in exception handling code because of that
Scott Sanderson
@ssanderson
Dec 10 2014 18:57
Should be the correct patch now
and I deleted the calls to showtraceback per your comment
Thomas Kluyver
@takluyver
Dec 10 2014 18:59
great, thanks
it's looking good to me.
Scott Sanderson
@ssanderson
Dec 10 2014 19:33
is there a way to tell IPCompleter to not complete names from the hidden namespace?
guessing not from looking at the code
Thomas Kluyver
@takluyver
Dec 10 2014 19:34
no, I don't think so. The completion machinery needs an overhaul, but it's definitely not going to happen before 3.0
Scott Sanderson
@ssanderson
Dec 10 2014 20:42

is anyone else seeing In [9]: from zipline.data import load_from_yahoo

woops

bad paste
TypeError: show_in_pager() got an unexpected keyword argument 'start'
I get that on pinfo calls on latest master
Thomas Kluyver
@takluyver
Dec 10 2014 20:45
could well be, that code was changed recently
Scott Sanderson
@ssanderson
Dec 10 2014 20:45
looks like c0ca9867
Jason Grout
@jasongrout
Dec 10 2014 20:46
easy promises fix looking for reviewers: ipython/ipython#7164 :)
Scott Sanderson
@ssanderson
Dec 10 2014 20:47
issue filed ipython/ipython#7165
Jonathan Frederic
@jdfreder
Dec 10 2014 20:48
@jasongrout see ipython/ipython#7161
ipython/ipython#7161
Kyle Kelley
@rgbkrk
Dec 10 2014 20:49
Oddly, could have merged #7164
Which I guess would have been a nothing commit
Jonathan Frederic
@jdfreder
Dec 10 2014 20:49
Yeah it's strange that it lets you do that
Kyle Kelley
@rgbkrk
Dec 10 2014 20:49
Guess it doesn't show the merge commit
because it was the exact same
Jonathan Frederic
@jdfreder
Dec 10 2014 20:49
I was puzzled looking at the diff
Kyle Kelley
@rgbkrk
Dec 10 2014 20:49
yeah
they both came from a different parent I think though
Jonathan Frederic
@jdfreder
Dec 10 2014 20:50
ahh
Kyle Kelley
@rgbkrk
Dec 10 2014 20:51
4b34cbb vs. 6daf9b3
odd stuff
Scott Sanderson
@ssanderson
Dec 10 2014 20:57
Think I found a fix for #7164. There's a dummy hook that can be overridden by users, but it has the wrong function signature. PR opened at ipython/ipython#7166
Jason Grout
@jasongrout
Dec 10 2014 22:35
@jdfreder: I'll give you an async racehorse here too for #7161: :horse_racing:
Jonathan Frederic
@jdfreder
Dec 10 2014 22:36
sweet!