[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

re: Changes in formal naming for NetBSD porting effort(s)



   
   --L6iaP+gRLNZHKoI4
   Content-Type: text/plain; charset=us-ascii
   Content-Disposition: inline
   
   On Wed, Dec 17, 2003 at 12:44:07PM -0500, Nathan Hawkins wrote:
   > On Fri, Dec 12, 2003 at 02:58:42PM -0700, Joel Baker wrote:
   > > I really need to sit down and write a proposal / patches for NetBSD to
   > > support the 'vendor' sysctl tree, that can be checked usefully. Since that
   > > would be the canonical way of testing this (a 'debian' vendor could have a
   > > sub-field indicating which sort of port it was).
   > 
   > There is one possible problem with that. It occurred to that it is quite
   > possible now to run regular freebsd in a jail or chroot on a debian
   > freebsd box, and vice versa. I assume the same is true for the netbsd
   > port. If we use sysctl or uname to make the distinction, that jail or
   > chroot usage will be affected.
   
   This is, in fact, an issue with my chroot environment which I have to
   spend some amount of effort to work around.
   
   The nicer solution, IMO, would be user-mode-<foo>BSD, akin to
   user-mode-linux. Especially when combined with bind mounts, this has the
   potential to be very powerful, indeed.
   
   Unfortunately, I'm still not entirely clear on what all would need to be
   done, to accomplish this properly.


there's a user-mode freebsd coming or done or something.  no one that
i know of is working on it for netbsd yet ... 

... however, you may be able to use pkgsrc's "libkver" (at least, be
able to extend it easily) to pretend to be a different version of
netbsd.  people use libkver to build 1.6 packages on -current systems.
it uses the same methods (LD_PRELOAD) as debian's fakeroot...


.mrg.



Reply to: