[IPython-User] Tips to speed up notebook?

G B g.c.b.at.work@gmail....
Tue Oct 16 12:32:39 CDT 2012


In short: holy cow!  Much, much faster!

Ok, so some background to cover my embarrassment at not having tried this again before asking publicly.   I'd been running my notebook back and forth between Firefox and Safari all through the first weeks of my work with it.  I'd say for the first 15-20 pages.  Mostly I was expecting Firefox to handle the notebook better (open source projects tend to favor Firefox), but Safari generates nicer printed output.  I hadn't seen an appreciable performance difference while using either, though Firefox seems to launch and load more slowly.  Eventually I decided to stop messing with Firefox and stick with Safari.  I should have tried it again when things got so horribly slow with Safari— I'm not sure when they diverged, whether it was the size of the notebook or a Mountain Lion update.

It looks like Firefox is smart enough not to re-render things that aren't visible, which is probably why scrolling can be a bit inconsistent, while Safari must be updating the whole canvas with every small modification.

[As far as I know, Firefox handles it's updates appropriately— through the browser itself, on startup (rather than a continuous background process that can change anything).]

So I'm torn between being terribly embarrassed to have wasted everyones time on this, and wishing I'd embarrassed myself two weeks earlier to save a mountain of lost productivity.

Unfortunately I can't pass my notebook to Apple, but I'm going to submit a bug on this.  Is there a large, public notebook somewhere I can point them to as an example?



On Oct 16, 2012, at 1:33 AM, Thomas Kluyver <takowl@gmail.com> wrote:

> If you're not happy about Chrome, have you considered Firefox? I know
> it also runs automatic updates, but I consider Mozilla quite
> trustworthy.
> _______________________________________________
> IPython-User mailing list
> IPython-User@scipy.org
> http://mail.scipy.org/mailman/listinfo/ipython-user



More information about the IPython-User mailing list