DragonFlyBSD Is Almost To Linux 3.10 Era Intel Graphics Support

[phoronix]

Phoronix: DragonFlyBSD Is Almost To Linux 3.10 Era Intel Graphics Support

While FreeBSD, DragonFlyBSD, OpenBSD, and other BSD distributions have made much headway in the past year or two in porting the Linux DRM/KMS drivers to their kernels, the work still measurably lags behind the latest upstream Linux kernel code...

http://www.phoronix.com/scan.php?page=news_item&px=DragonFly-Linux-3.10-i915

[beast]

Nice, but why they don't port the NetBSD stuff which is on DRM 3.14?

[Luke_Wolf]

Nice, but why they don't port the NetBSD stuff which is on DRM 3.14?

Well one would hope that all of the BSDs are working together on this front given that they've all chosen the same path.

[enjolras]

BSD kernel, although having the same name, are internally very different in some area. Hence, porting from netBSD is not much easier than porting some linux. And it has some drawbacks to : when you port upstream code, you have upstream code with no changes. Hence, it's easier to follow updates from upstream. Following NetBSD could be an issue. What happens if NetBSD starts lagging behind ? You have to port the new changes from linux and deal with the changes netBSD did. That's what happened with the drm port from FreeBSD to dragonfly at some point and that's why dragonfly is following upstream.

But of course, these various ports serve as reference/documentation.

[techzilla]

well sure, you can hope...

Well one would hope that all of the BSDs are working together on this front given that they've all chosen the same path.

Hope without foundation is always irrational, it prevents you from actually addressing problems by dangling the possibility of improvement. So yes they theoretically "could" work together like the Linux community does, but their is no structural foundation for that required community to even exist. You could DIY such a community, preach volunteerism, and once again hope that the nature of humanity changes by mere example...

Or you could look at the cold hard facts, Linux works together by design, even when its the opposite of what said individuals would normally do.
This is because it has a structure in that working as a community is MORE beneficial, to the individuals, than going it alone. This is also a clear historical pattern with BSD *ish and other permissive licences. They can absolutely succeed as projects, no question about it as some in fact do, but what they don't do is create a Linux style community effort. This is for the most simple reasons, even though working together MAY be beneficial, it is often MORE beneficial for the interests to go alone.... and then they do, because that's what interests do, you get the results your structure serves best.

Now permissive can still form a community, but it must be a well funded community, whose interests actually align naturally. This is not common, and it's clear OS innovation is not one of those situations. SO does that mean we get no innovations? Not at all, we do in fact get innovations, they just never work together to create something near as polished as the Linux community. I felt this was necessary because of all the anti-GNU messages from people with economic interests, or wing nuts, telling you hope alone is somehow reasonable.

[beast]

Hope without foundation is always irrational, it prevents you from actually addressing problems by dangling the possibility of improvement. So yes they theoretically "could" work together like the Linux community does, but their is no structural foundation for that required community to even exist. You could DIY such a community, preach volunteerism, and once again hope that the nature of humanity changes by mere example...

Or you could look at the cold hard facts, Linux works together by design, even when its the opposite of what said individuals would normally do.
This is because it has a structure in that working as a community is MORE beneficial, to the individuals, than going it alone. This is also a clear historical pattern with BSD *ish and other permissive licences. They can absolutely succeed as projects, no question about it as some in fact do, but what they don't do is create a Linux style community effort. This is for the most simple reasons, even though working together MAY be beneficial, it is often MORE beneficial for the interests to go alone.... and then they do, because that's what interests do, you get the results your structure serves best.

Now permissive can still form a community, but it must be a well funded community, whose interests actually align naturally. This is not common, and it's clear OS innovation is not one of those situations. SO does that mean we get no innovations? Not at all, we do in fact get innovations, they just never work together to create something near as polished as the Linux community. I felt this was necessary because of all the anti-GNU messages from people with economic interests, or wing nuts, telling you hope alone is somehow reasonable.

You speak from a Linux Community, there is no such thing? There are over 500 distros, how many of them work on the kernel side, 5%, 10%? You have a unification in the userland because Red Hat do the job that nobody wants to do. And so on, Gnome, KDE, Mesa ... no projects can live without the power of money. The success of Linux isn't a result of Community members. IMHO BSDs will always behind Linux when not big companies get the power to control the development.