2018 Isn't The Year Of The GNU Hurd

We are now half-way through 2018 and the work on GNU Hurd and related components like GNU March have been very light.

Hurd.git has seen less than 40 commits year to date, Gnumach.git has seen less than two dozen so far this year, etc. It's now been a year and a half since the last GNU Hurd (v0.9) release.

While the GNU Hurd micro-kernel pre-dates the Linux kernel, it remains far behind. In fact, Hurd still doesn't have USB support, the 64-bit support has only just begun, and there isn't yet other modern hardware support features. Hurd pretty much is only semi-practical when running through QEMU.

It's unfortunate given the long history of the GNU Hurd and while besides the Linux kernel other efforts like Genode and Haiku and Redox OS have been able to accomplish much more.

Curious about the development stats for Hurd, I ran some GitStats as of this morning. Hurd is made up of 363,475 lines across 1,830 files and this was built up over 12,039 commits from 63 different authors. Here's a look at the Hurd commit activity over the years:

In 2017 there were 89 commits adding 9k lines of code and removing 4k lines of code. So far this year have just been 31 commits adding 1k lines of code and removing 561 lines of code. Quite the difference compared to the mid-90s when Hurd would see around 2,000 commits per calendar year or even years ago where there would be hundreds of commits per year. (The Linux kernel meanwhile can easily shift around hundreds of thousands of lines of code each cycle, happening multiple times per year, granted its a monolithic kernel and not the micro-kernel design of the Hurd.)

Last year were seven active authors to Hurd while year-to-date there has been five, also a decrease from a few years prior when there would be around a dozen contributors.

Those wanting to dig through more Hurd development statistics can find them here.