First, I'd like to let everyone know that 0.11 is on schedule to be released on January 6th. We have an imminent core library code freeze, followed by some intensive sprints on acceptance testing, documentation, and example building leading up to the full release. Today we have a new Dev Build 0.11.0dev7 available (see below for installation notes) and there will likely be one or two more dev build/release candidates before the final release.
There are *many* exciting new features and fixes going into 0.11, and we will describe all of them in much more detail in release notes, announcements, blogs, etc. But for now I'd like to draw attention especially to the new Bokeh server. In addition to being much more performant and usable, the new Bokeh server is much simpler to create apps for. This is highlighted by the fact that it has been really easy to write several new apps, with a minimal amount of code, in a very short time! You can see several nice examples here:
We look forward to adding even more before the release (and if you make any great examples we always love new contributions). To run any of these examples, navigate to that directory, and run "bokeh serve" on any file or directory there e.g.
bokeh serve --show selection_histogram.py
bokeh serve --show stocks/
This will open the app in a new browser tab.
Bokeh also has a new proper client library for interacting with the server. It's possible to use this client library to interact with the Bokeh server using the existing "output_server" function. This method of using the Bokeh server is more suited to exploratory or local, individual usage, but can still be quite useful. You can see examples of this usage of the server here:
Instructions for running those examples are in the README there.
The new server is new (maybe stating the obvious!) and although it has much more extensive unit-testing than the old server, it is an area where our automated CI testing is still catching up, and still requires extensive manual testing. All of the examples above have performed very well across all platforms in our testing so far, but additional feedback from more users and more testing is always valuable and appreciated, so please try things out and let us know any issues or questions!
Dev Build Details
The dev build is available on Anaconda Cloud at:
and if you want to try it out, you can follow the installation instructions here:
And the docs for this devel build are available at:
Please, test it and report back!
Bryan Van de Ven