Concise links this week.
- DiscoverBSD’s roundup for 2014/06/02.
- Code review culture meets FreeBSD.
- Michael Dexter’s BSDCan 2014 trip report.
- A RetroBSD license audit. (from #nycbug)
- Ass ember.
- FreeBSD GNATS is gone; now it’s Bugzilla. Nobody sheds a tear.
- NetBSD runs on BeBoxes? I didn’t realize.
- FreeBSD now has ‘stock’ network drivers.
- It’s always nice when people relicense.
- I like crosspollination, too.
- The default font path in pkgsrc has changed.
You know, I notice we merge a lot of patches from FreeBSD, I mean, they are a much bigger project with far more developers, and I suppose given our origins, they’re somewhat of a sister OS, but I wasn’t aware that any of the other BSDs took changes that we made, that’s impressive considering our relative size. is our ancestry with FreeBSD the reason why our merging changes from them is more common than our merging changes from OpenBSD or NetBSD, or is that more due to developers on our team being current/former developers of FreeBSD, or simply due to FreeBSD’s popularity?
Having DragonFly changes show up in FreeBSD has been more common in recent years – common enough I don’t always link to it.
I don’t think merging changes with them is far and away more common. Our version of pf is ported straight from OpenBSD, as is our dhclient. We borrowed some code from NetBSD for a number of things a while back. Internationalization, maybe?
You could argue we’re reducing the number of FreeBSD-linked things – ftigeot has been moving the kms/i915 code closer to what’s in Linux, so that instead of chasing another BSD’s version of it, we will (hopefully) be able to chase the version much of the world develops for.
Isn’t it harder to match things made for a Linux kernel than a BSD kernel?
Yes and no, it depends on the subsystems.
Catching up with Linux can be somewhat difficult due to the need to re-implement some of the APIs with a BSD License.
But for the drm code, the main issue is that it had been abandoned for 10 years on FreeBSD/DragonFly. The last developer left around 2004 and now we need to reduce an enormous technical debt.