A big release with new features and important bug fixes!

What's new:

• Context Managers. This context manager functionality is one of many great ideas found in processing.py that several members of the py5 community have been asking for. Ideas and feedback like this contribute much to py5, so please keep them coming! And many thanks to the processing.py team for coming up with this idea.

• py5bot now supports the SVG and PDF renderers

• Update to the latest Processing 4.0 beta 2 jars

• The save() and save_frame() methods can now save to a io.BytesIO object

• Version number now conforms to proper semantic versioning standards

Bug Fixes:

• #40

• #44

• Other small bugs I found and fixed along the way

What's Ahead:

• Rework noise functionality. See #38 for more information. The new version's noise functionality may not be backwards compatible.

• There are other items on my todo list but I want to make reworking the noise functionality a higher priority. I'll do a new release once this is complete.

The new py5 documentation website is now live! Go to https://py5.ixora.io to have a look.

The documentation is built with Jupyter Book. One of the reasons why I am so enthusiastic about this is that Jupyter Book is so well designed for creating Python documentation. It's very easy use and the end result is a much higher quality than what I was using before.

Now I need to get to work and document all the great features in py5 that nobody knows about!

I'm enthusiastic about this prototype for py5's documentation website: py5dev.ixora.io. It is built with Jupyter Book.

Once I got up to speed on Jupyter Book I was able to modify the Reference documentation to use the new system. That was probably the most tedious part. One challenge is that I am using restructuredtext for the documentation source content, so the reference documentation build process also outputs restructuredtext. That's fine because Jupyter Book supports that, but it is clear that markdown or Myst is easier to use. I will use Myst for new documentation files, not restructuredtext.

One of the great features of Jupyter Book is the ability to link the documentation to binder. Readers will be able to launch a notebook kernel right from the documentation. In addition, Jupyter Book supports Thebe, an amazing tool that I never imagined was even possible. Readers will also be able to execute py5 code through the documentation website itself, with the execution results provided by binder. This supports custom kernels like py5bot and embedded animations using the Sketch Portal. This connection with binder will be used extensively in the tutorials and howto sections.

Shortly after the 0.5a1 release I learned that the run_sketch utility did not work on Windows computers. It was an easy fix, and critical to support tabreturn aka Trisan Bunn's upcoming CC Fest presentation on Thonny and py5.

While doing additional testing I discovered that the run_sketch utility also did not work correctly on OSX. This was also critical, and unfortunately, not an easy fix. Frankly, it was a painful problem to solve. Nevertheless, I came up with a solution using pyobjc. Now run_sketch will work correctly on all platforms, and everyone who watches tabreturn's presentation on Sunday can give py5 and Thonny a try.

While exploring py5objc I discovered a new approach for running py5 on OSX that I believe I can use to address most the current limitations on OSX. I'll continue exploring this in future releases.

While doing all of this I also expanded py5's support for specifying colors using hexadecimal notation and web color notation. Now assignments to the pixels array can use both notations. In addition, py5 can support 3 character web color notations (e.g. "#822"). See #26 for more information.

What's Ahead:

• Rework noise functionality

• OSX feature parity

• Write Tutorials and How-tos

• Further develop the py5examples repo

• Unit tests

The 0.5a0 release was a bad release. I botched the py5bot kernel code in a poorly executed merge, and then didn't do sufficient testing to detect the problem. This release fixes those problems.

I have updated my release process to make sure this doesn't happen again. Also, I will prioritize developing a unit test framework for py5.

I managed to sneak in one new feature to this release: Proper support for specifying colors using hexadecimal notation and web color notion. See #26.

Also, the sketch_portal() method will by default throttle the frame rate to 30 frames per second. Previously the default was for no throttling. This change will improve the experience of users who use this for the first time on mybinder, where a Sketch running at the default Sketch speed of 60 frames per second will look jumpy in the Sketch portal when no throttling is used.

What's Ahead:

• Write Tutorials and How-tos

• Further develop the py5examples repo

• Unit tests, because clearly this project needs them

What's new:

• A second Jupyter Notebook Kernel: py5bot. There's also a new IPython magic called %%py5bot that is available within the py5 kernel.

• The size() method can be called from the setup() function intead of settings(). Before Sketch execution, py5 will identify code in setup() that belongs in settings() and will segment the code appropriately.

• Improved error reporting for IPython magics, py5bot, and all static mode sketches

• Sketch windows now appear in the taskbar or Dock with the py5 logo

Breaking changes:

• args has been renamed to pargs to avoid a name conflict with the conventional use of the variable args

• the use_thread default parameter in save() and save_frame() has been changed from False to True

Also:

• Massive improvements for OSX Users. The OSX Sketch exiting issues have been solved. The other OSX issues are better articulated in the Special Notes for Mac Users.

• Lots of little bug fixes

What's Ahead:

• Write Tutorials and How-tos

• Further develop the py5examples repo