> It's all fine and good making technical decisions based on drawings
> and handwaving and philosophizing, but at some point someone has to do
> the code.

HI Adrian,

. What I eluded too is not a small project. It is something that would need 
proper discussion and agreement, since it would be pervasive and touch 
critical parts of the OS, such as the init system, system config databases , 
and add proper services management facility. It would also benefit 
from a new form of  kernel IPC. It would need consensus from FreeBSD board or 
whatever to have any chance of even starting up. Nobody in his 
sane mind would start it otherwise. Most likely he would work in vain,
And when consensus that something HAS to be done will exist, and from empty 
discussion you would have a implementation plan, when maybe the FreeBSD 
foundation would get involved and sponsor such a important project to see it 
done to the end.

And there are efforts today to go down the path I mentioned, NextBSD is the 
incarnation of such an effort. And while they offer code and they do make 
progress I do not seeing anyone in FreeBSD beeing too eager to commit that code 
:P (Im not saying that you should adapt launchd and add another comapt layer 
for FreeBSD for mach ). I for one like what Solaris does. What  Im saying that 
such work would never be possible directly in FreeBSD, because lack of 
consensus that anything serious should be done, apart from patching on sides.

 I am saying that gathering consensus that something has to be done must exist 
before any code is written . Else you wont get much.
freebsd-current@freebsd.org mailing list
To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"

Reply via email to