Re: Donation of ARM CubieTruck to anyone interested in developing ARM-Debian-desktop
+++ Luc Verhaegen [2013-11-05 17:56 +0100]:
>
> I personally really dislike this upstream fetish.
>
> The effect is highly detrimental, as it mostly ignores the hard work
> needed on getting broken and bad code in difficult places fixed. The
> initial and perhaps easy bits make it upstream, and then after a while,
> a new platform comes along and interest is lost, and upstream has just
> some initial code included, which will only have limited use. At the
> same time, the difficult bits will have seen less love as well, and
> no-one will have won anything. And then, a few years down the road,
> upstream will do what they did with telechips.
You are quite right that distros waiting for upstreamed device support
causes problems. I've been there mysefl with Debian on Balloonboard,
which never got fully mainlined.
But it's not a 'fetish'. The issue is that there is no resource in
Debian (or other distros, mostly) to support piles of random vendor
kernel trees, one for each device. We tried to do this back in the early
days of arm and it didn't work - code mouldered and broke very quickly.
Nothing stops people supporting vendor kernel trees outside of debian,
and this often happens, but it's simply not practical within the distro.
We can and do support generic fixes in all our packages, including
kernels, but loads of device-specific kernel sources, or massive
incompatible patches from random different kernel versions is a really
big piece of work to maintain. Without a whole load of people turning up
and doing that work, this isn't going to change.
I have a sunxi-kernel device (EOMA-PC) and plan to help get Debian in
good shape on that generally. That will include using a sunxi kernel
tree jammed in for the time being, which help people get involved, but I
fully support the idea that that code needs to go upstream for the
distro itself to support it.
Wookey
--
Principal hats: Linaro, Emdebian, Wookware, Balloonboard, ARM
http://wookware.org/
Reply to: