These are chat archives for ipython/ipython

Oct 2014
Jason Grout
Oct 27 2014 14:29
A widget view should call its .remove() method to destroy itself.
(though I don't know if that actually happens when reexecuting a cell)
@takluyver, did you see our report of a problem with the callback-based view creation? Now, suddenly, order is not guaranteed for view creation, which violates some assumptions that were made in the box classes, for example
(i.e., if I have a list of children, now the order of their creation is not guaranteed, while before it was)
Jason Grout
Oct 27 2014 20:41
@jdfreder: are you around? Have you ever considered making the widget-subarea have css align-items: 'stretch'?
Jonathan Frederic
Oct 27 2014 21:00
Hey @jasongrout
Sorry I was at lunch
I'm here at UCB
I think @SylvainCorlay and I were talking about that earlier
and my response
was that that'd be fine (opinion), as long as interact was modified such that the containing box has a fixed width
Although wait, erm
you're not talking about removing the fixed widths of
the basic widgets
so I guess you wouldn't have to worry about that anyways
Jonathan Frederic
Oct 27 2014 21:35
@minrk add me!
Thomas Kluyver
Oct 27 2014 23:57
@jdfreder when you have a minute, PR #6763 could use a look. It's pretty straightforward.