This post will be the start of The X-Files, a blog series exploring the Go sub-repositories located at
golang.org/x/*. Chances are, you may not have heard them called that before or know where they came from. Introductions are in order…
A brief origin story
Go promises, with reasonable exceptions, that its compiler and standard library will remain forward-compatible for all minor releases of the Go 1 language specification. This is incredibly refreshing, especially if you’ve ever used a framework or language that seems to break the world on every patch release. While perhaps this makes Go boring, it permits developers to confidently use the language and stdlib, leaving the core team to focus on performance in the compiler, optimizations to the garbage collector, and other quality-of-life improvements.
But with this guarantee, the core team needed a place for experimentation, for tooling, for supplements to the standard library: the proverbial junk drawer of the Go language ecosystem. Thus, the sub-repositories were born. These projects live under the parent one, utilizing the same process and tools (like Gerrit), but without the strict compatibility requirements.
Wherefore art thou sub-repository?
The compatibility document describes subrepos as follows:
Code in sub-repositories of the main go tree […] may be developed under looser compatibility requirements. However, the sub-repositories will be tagged as appropriate to identify versions that are compatible with the Go 1 point releases. Go 1 and the Future of Go Programs
Well, that’s pretty vague … And, expectedly, the packages have accumulated a smorgasbord of libraries, drivers, tools, and experiments. In them, you can find quite a lot of interesting goodies - some well known, others more obscure. For instance,
golang.org/x/net delivered us the beloved
context.Context now as of 1.7 in the stdlib; meanwhile,
golang.org/x/text grows into an incredible text manipulation library that you’ve never heard of. There are a ton of useful utilities and glimpses behind-the-scenes into the processes of the core team.
So, what qualifies as a sub-repository? At the time of writing, there’s been an ongoing discussion on what the policy for just that should be. There is a lot of pressure to add more-and-more to these x-packages, which is just not a scalable possibility for the Go team to keep up with. On the flip-side, folks in the community want a space to locate de-facto implementations or interfaces of “foundational packages” (e.g., A/V or image codecs).
I imagine a compromise will be reached that relieves the core team from being buried in sub-repos while also elevating high-quality packages (hopefully under a unified import). Regardless of how that proposal shakes out, though, it’s definitely worth diving into these sub-repos.
What will be covered?
This series will focus on the top-level sub-repositories as reported by build.golang.org. Some of these packages are light enough to cover in a single post (like
golang.org/x/time/rate), while others will be split into multiple to better accomodate them. CLI tools and their corresponding library packages will be covered together. The order I do this will be primarily driven by interest, so if you’d like me to cover one in particular, let me know!
Without further ado, the series so far:
- Controlling Throughput with rate.Limiter
- Avoiding Concurrency Boilerplate With golang.org/x/sync
- With many more to come!