Here are the questions I asked during the hangouts session (paraphrased):
-------------------------------------------------------------
Regarding continuous integration:
Has looked into OBS? (open build server, https://build.opensuse.org/) It
can be installed on a local machine or server, supports automatically
creating and deleting fresh images with each build, and supposedly works
with osX as well as Linux although I haven't tested it (it does need a Mac
OsX VM).
-------------------------------------------------------------
Regarding bug handling:
It might be possible to do something with this with Google Code-in (
https://developers.google.com/open-source/gci/), although I am not 100%
sure this would be acceptable there.
Another possibility would be to allow volunteer triagers who may not be
developers but can at least handle basic stuff like finding duplicates and
following up with the reporters of old bugs.
-------------------------------------------------------------
Regarding embedding:
Perhaps there could be a generic "figure" widget for each backend. The
widget would automatically handle all the backend-specific stuff necessary
to create a figure object and display it in the widget (including resizing
and such). It would provide access to the low-level backend-specific parts
to make it possible to subclass it or change the details of how it works.
The figure windows would have this widget as their central widget, and
would probably access it using these low-level components.
However, for basic usage each widget would also have a "figure" attribute,
which would contain a single generic figure object. Basic users who just
want to embed a regular plot could access that figure object and use it in
the normal way.
These could probably all be accessed from a single module, although they
probably would all live in their own backend-specific modules.
It wouldn't allow people to use pyplot, but if this was the documented way
to do embedding and the documentation made it clear you needed to use the
OO interface when embedding I think it would greatly reduce the amount of
trouble people have.
------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60135991&iu=/4140/ostg.clktrk
_______________________________________________
Matplotlib-devel mailing list
Matplotlib-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
