Hey all, I know we nixed the previous notebook progress bar because it was really bogging things down—but it would be handy to have one, and there are some ASCII-only implementations that probably won’t use as much memory. Any thoughts? I’d be happy to look into it myself. Best, John
On Friday, October 24, 2014, John ZuHone
Hey all,
I know we nixed the previous notebook progress bar because it was really bogging things down—but it would be handy to have one, and there are some ASCII-only implementations that probably won’t use as much memory.
That wasn't the issue, it was buggy and was leaving markup behind in the notebook after cleaning itself up. This led to things like output cells increasing in size with each progress bar used and ballooning JavaScript and HTML markup embedded in the notebook JSON document. It's also tricky because the maintainer of the progress bar we use isn't terribly responsive and wasn't interested in integrating a notebook-aware version last time I checked. That said, there is a progress bar widget in IPython 2.0 that might be useful. I also haven't looked into this in a year or so, it's possible the bugs that forced us to drop the notebook-aware progress bar have been fixed.
Any thoughts? I’d be happy to look into it myself.
Best,
John _______________________________________________ yt-dev mailing list yt-dev@lists.spacepope.org javascript:; http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
Hey,
Just a quick note to point out that astropy have just implemented this in
master I believe.
Stuart
On 24 Oct 2014 16:36, "Nathan Goldbaum"
On Friday, October 24, 2014, John ZuHone
wrote: Hey all,
I know we nixed the previous notebook progress bar because it was really bogging things down—but it would be handy to have one, and there are some ASCII-only implementations that probably won’t use as much memory.
That wasn't the issue, it was buggy and was leaving markup behind in the notebook after cleaning itself up. This led to things like output cells increasing in size with each progress bar used and ballooning JavaScript and HTML markup embedded in the notebook JSON document.
It's also tricky because the maintainer of the progress bar we use isn't terribly responsive and wasn't interested in integrating a notebook-aware version last time I checked.
That said, there is a progress bar widget in IPython 2.0 that might be useful.
I also haven't looked into this in a year or so, it's possible the bugs that forced us to drop the notebook-aware progress bar have been fixed.
Any thoughts? I’d be happy to look into it myself.
Best,
John _______________________________________________ yt-dev mailing list yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
_______________________________________________ yt-dev mailing list yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
There is one issue with that, which is that so far we have resisted making astropy a hard dependency. If we wanted to use their progressbar we'd have to do so (this could always be an on-demand import, I guess). I like astropy and use it a lot, but I agree that we should keep the number of hard dependencies to an absolute minimum. There's nothing stopping us from looking at that code and implementing it ourselves, I guess... John ZuHone Kavli Center for Astrophysics and Space Research Massachusetts Institute of Technology 77 Massachusetts Ave., 37-582G Cambridge, MA 02139 (w) 617-253-2354 (m) 781-708-5004 jzuhone@space.mit.edu jzuhone@gmail.com http://www.jzuhone.com
On Oct 24, 2014, at 12:38 PM, Stuart Mumford
wrote: Hey,
Just a quick note to point out that astropy have just implemented this in master I believe.
Stuart
On 24 Oct 2014 16:36, "Nathan Goldbaum"
wrote: On Friday, October 24, 2014, John ZuHone
wrote: Hey all, I know we nixed the previous notebook progress bar because it was really bogging things down—but it would be handy to have one, and there are some ASCII-only implementations that probably won’t use as much memory.
That wasn't the issue, it was buggy and was leaving markup behind in the notebook after cleaning itself up. This led to things like output cells increasing in size with each progress bar used and ballooning JavaScript and HTML markup embedded in the notebook JSON document.
It's also tricky because the maintainer of the progress bar we use isn't terribly responsive and wasn't interested in integrating a notebook-aware version last time I checked.
That said, there is a progress bar widget in IPython 2.0 that might be useful.
I also haven't looked into this in a year or so, it's possible the bugs that forced us to drop the notebook-aware progress bar have been fixed.
Any thoughts? I’d be happy to look into it myself.
Best,
John _______________________________________________ yt-dev mailing list yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
_______________________________________________ yt-dev mailing list yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
yt-dev mailing list yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
Yes, I think this is not sufficient to justify a hard dependency. But
perhaps a solution can be found. I'm surprised ipython itself doesn't have
one.
On Fri, Oct 24, 2014, 12:42 PM John ZuHone
There is one issue with that, which is that so far we have resisted making astropy a hard dependency. If we wanted to use their progressbar we'd have to do so (this could always be an on-demand import, I guess).
I like astropy and use it a lot, but I agree that we should keep the number of hard dependencies to an absolute minimum. There's nothing stopping us from looking at that code and implementing it ourselves, I guess...
John ZuHone Kavli Center for Astrophysics and Space Research Massachusetts Institute of Technology 77 Massachusetts Ave., 37-582G Cambridge, MA 02139 (w) 617-253-2354 (m) 781-708-5004 jzuhone@space.mit.edu jzuhone@gmail.com http://www.jzuhone.com
On Oct 24, 2014, at 12:38 PM, Stuart Mumford
wrote: Hey,
Just a quick note to point out that astropy have just implemented this in master I believe.
Stuart On 24 Oct 2014 16:36, "Nathan Goldbaum"
wrote: On Friday, October 24, 2014, John ZuHone
wrote: Hey all,
I know we nixed the previous notebook progress bar because it was really bogging things down—but it would be handy to have one, and there are some ASCII-only implementations that probably won’t use as much memory.
That wasn't the issue, it was buggy and was leaving markup behind in the notebook after cleaning itself up. This led to things like output cells increasing in size with each progress bar used and ballooning JavaScript and HTML markup embedded in the notebook JSON document.
It's also tricky because the maintainer of the progress bar we use isn't terribly responsive and wasn't interested in integrating a notebook-aware version last time I checked.
That said, there is a progress bar widget in IPython 2.0 that might be useful.
I also haven't looked into this in a year or so, it's possible the bugs that forced us to drop the notebook-aware progress bar have been fixed.
Any thoughts? I’d be happy to look into it myself.
Best,
John _______________________________________________ yt-dev mailing list yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
_______________________________________________ yt-dev mailing list yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
_______________________________________________
yt-dev mailing list yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
_______________________________________________ yt-dev mailing list yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
If the astropy code is modular enough, they could probably be persuaded to
release it on pypi and then depend on it. We could do the same.
John, I think the consensus is that this is a great thing that we'd love to
integrate into yt. If we can figure out a way to include it without
incurring the technical issues I raised above, I think this will be very
helpful.
On Fri, Oct 24, 2014 at 9:47 AM, Matthew Turk
Yes, I think this is not sufficient to justify a hard dependency. But perhaps a solution can be found. I'm surprised ipython itself doesn't have one.
On Fri, Oct 24, 2014, 12:42 PM John ZuHone
wrote: There is one issue with that, which is that so far we have resisted making astropy a hard dependency. If we wanted to use their progressbar we'd have to do so (this could always be an on-demand import, I guess).
I like astropy and use it a lot, but I agree that we should keep the number of hard dependencies to an absolute minimum. There's nothing stopping us from looking at that code and implementing it ourselves, I guess...
John ZuHone Kavli Center for Astrophysics and Space Research Massachusetts Institute of Technology 77 Massachusetts Ave., 37-582G Cambridge, MA 02139 (w) 617-253-2354 (m) 781-708-5004 jzuhone@space.mit.edu jzuhone@gmail.com http://www.jzuhone.com
On Oct 24, 2014, at 12:38 PM, Stuart Mumford
wrote: Hey,
Just a quick note to point out that astropy have just implemented this in master I believe.
Stuart On 24 Oct 2014 16:36, "Nathan Goldbaum"
wrote: On Friday, October 24, 2014, John ZuHone
wrote: Hey all,
I know we nixed the previous notebook progress bar because it was really bogging things down—but it would be handy to have one, and there are some ASCII-only implementations that probably won’t use as much memory.
That wasn't the issue, it was buggy and was leaving markup behind in the notebook after cleaning itself up. This led to things like output cells increasing in size with each progress bar used and ballooning JavaScript and HTML markup embedded in the notebook JSON document.
It's also tricky because the maintainer of the progress bar we use isn't terribly responsive and wasn't interested in integrating a notebook-aware version last time I checked.
That said, there is a progress bar widget in IPython 2.0 that might be useful.
I also haven't looked into this in a year or so, it's possible the bugs that forced us to drop the notebook-aware progress bar have been fixed.
Any thoughts? I’d be happy to look into it myself.
Best,
John _______________________________________________ yt-dev mailing list yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
_______________________________________________ yt-dev mailing list yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
_______________________________________________
yt-dev mailing list yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
_______________________________________________ yt-dev mailing list yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
_______________________________________________ yt-dev mailing list yt-dev@lists.spacepope.org http://lists.spacepope.org/listinfo.cgi/yt-dev-spacepope.org
participants (4)
-
John ZuHone
-
Matthew Turk
-
Nathan Goldbaum
-
Stuart Mumford