gfx-next update: drm-4.8-rc2 tagged in drm-next

[Matthew Macy]

 ---- On Mon, 26 Sep 2016 22:07:37 -0700 Kevin Oberman <rkoberman at gmail.com> wrote ---- 
 > OK. It's time to cut this crap out! This is why people talk about FreeBSD 
 > development being a hostile place. This sort of thing is why! If you want 
 > to discourage competent people from working on FreeBSD, this is an 
 > excellent way. 
 >  
 > For years the FreeBSD support for X has lagged way behind. This was true 
 > even before things like KMS made it much harder for for FreeBSD to track 
 > development. The FreeBSD Foundation has had to hire people to move FreeBSD 
 > to just a three or four years. Or, in computer years, about a two or three 
 > decades. 
 >  
 > Now someone comes along and puts in the amazing amount of work into moving 
 > FreeBSD's X to a point that is very near the state of the art and move the 
 > infrastructure it sits on to one that is sufficiently maintainable so that 
 > FreeBSD can hope to stay there. 
 >  
 > The people who did this are not people who have a history of major FreeBSD 

I actually have my fingerprints all over the kernel. Most of the handful major FreeBSD companies all make substantial use of my code. Unfortunately, much of my code doesn't make it back for cultural reasons (unfortunately the committer model doesn't scale as well to extremely large projects and evolutionarily is not as robust - even if it does make it more fun as a fraternity).

Thanks for your support, but let's not get too excited if somebody gets his account hacked or just plain gets lonely in his mother's basement and needs attention. I've privately asked that he be removed from the lists.

So long as 99% of people are duly respectful of the fact that this has been a great deal of work and that there is more work that lies ahead we'll be fine. The last two KMS updates gestated a year out of tree before going in. This update started in late April so if it were to be fully stabilized and all be in by November that would be quite an accomplishment. On top of that this is actually much bigger in many respects in that it bridges more than 3 1/2 years of kernel releases and shifts all the localized complexity out of the drivers themselves. In theory at least, the shifting of complexity will make it much easier for myself and others to track whatever happens to be the latest Linux release and thus may allow FreeBSD to maintain support for the latest hardware on an ongoing basis. Of course it will require continued work to support. It remains to be seen if the community can muster up that effort in one form or another on a sustained basis.


 > contributions and have a real desire to help FreeBSD as well as its 
 > derivatives. They do the work. They work with FreeBSD on the freebsd-x11@ 
 > users to get testing done, and, when it looks good, submit it to 
 > Phabricator for review, and ask for review. NextBSD was never mentioned and 
 > suddenly someone jumps up and yells "Go away. We don't serve your kind 
 > here!". Just because he works on NextBSD. Tell you what... move it to 
 > freebsd-advacocy@ where this sort of thing belongs or just go away. Let 
 > Matthew get this done. If anyone is banned from the list, it's Andrei. (And 
 > I am NOT advocating that.)