Wheels

Announcment: My book, from python import better, has been published. This post is based on one of the chapters from it.

When Python started out, one of the oft-touted benefits was "batteries included!". Gone were the days of searching for which XML parsing library was the best -- just use the one built-in to the standard library. However, the standard library can only hold so much special purpose stuff. Few now remember, but it used to have SGI Audio specific functionality.

These days, one of the biggest benefits of Python is the extensive third-party repository of stuff. This is the Python Package Index (PyPI), formerly known as the "Cheese Shop" after an obscure Monty Python skit. Of course, what else would be available from the Cheese shop than wheels of cheese? But a second pun was hiding behind the term "wheels": those are the things that need no reinvention!

The new PyPI warehouse launched, with new code hosting unbelievable amounts of content: around 140,000 packages at times of this post (unless I take too long in publishing it, and then who knows how big PyPI will be!)

Nobody can sift through 140K package descriptions, of course. A short-lived attempt to have "Stars" fell victim to allegations of ballot stuffing and moderation, and was quickly removed. Searching on key words would be useful, but searching without sorting rarely is -- and what would you sort on?

PyPI is not the place to find which libraries are useful. It is the place to find objective truths: which version is the latest, when was it released, what is a project's homepage, etc. Recommendations are best found elsewhere.

The first place I like to start is with the Awesome Python list. However, it is important to note that its contribution guidelines are just "submit a link" and there is no official way to remove a library from the list. Thus, the "awesome" in the name means "someone once thought it was awesome, and cared enough to add it". The list should be treated as mild suggestions. Before using a library, check release history, GitHub health, code quality and other metrics you might care about.

Another useful resource is Planet Python. It is a feed aggregator of various blogs. Many of the blog posts will feature either a recommendation of a particular library, a release announcement, or just discussion which involves using a third-party library. Along side the written word is the live performance -- PyVideo links to more Python talks than you can shake a stick at, aggregating talks from conferences around the world: again, many of the talks will feature discussion of a particular third-party library.

Last, but not least, the live, interactive, version of PyVideo: Python meetups and conferences. Those are where I discovered some of my favorite libraries.