[IPython-dev] New ipythonqt console questions/feedback

Fernando Perez fperez.net at gmail.com
Fri Oct 8 18:52:22 EDT 2010


On Fri, Oct 8, 2010 at 3:41 PM, MinRK <benjaminrk at gmail.com> wrote:
> It doesn't necessarily have to be done differently at startup, because you
> can 'destroy' a widget at any point, leaving the process alive.

Sorry, do you mean leaving the client process or the kernel process alive?

> I just updated my keepkernel branch with a couple things:
> 1) fixed error when resetting pykernel (it still reset, but printed an
> error)
> 2) fixed issue where closing a frontend, even secondary ones, always
> shutdown the kernel
> 3) shutdown_reply now goes out on the pub socket, so all clients are
> notified
>    3.a) this means that all clients can (and do) refresh the screen when a
> reset is called, just like the master frontend
> 4) kernel can stay alive after consoles are shutdown, and can be shutdown by
> any frontend at any point
>    4.a) this means that a shutdown request from any frontend can close all
> open frontends and the kernel, even if the kernel is detached, leaving no
> processes zombified.
>    4.b) 4.a required that a 'reset' element be added to
> shutdown_request/reply messages to identify the difference between a real
> shutdown message and stage 1 of a reset.
> http://github.com/minrk/ipython/commits/keepkernel/

Great!  Do you want this for inclusion now?  Initially you meant it
purely as a proof of concept, but at this point it's getting to be
useful functionality :)

I'll trade you a review for a review of my current pull request, that
fixes and cleans up ton of nasty execution logic:

http://github.com/ipython/ipython/pull/163

:)  The key question in yours is whether it leaves the Qt client
process zombified or not, I'm not quite clear on that point yet.

cheers,

f



More information about the IPython-dev mailing list