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

Re: hurd does NOT need /hurd

On Mon, May 20, 2002 at 12:19:12PM -0700, Thomas Bushnell, BSG wrote:
> Sure, but the point is that it's a dynamic environment.  
> The "environment" in question is the presence of an extra port,
> initialized in a special way.  Think perhaps of something like a file
> descriptor, not in the actual file descriptor table, but which is
> inherited automatically the way file descriptors are.
> Then, for the translator to work, it has to be able to communicate in
> a certain way on that port.  "settrans -a" is prepared to do that, and
> that's what's happening in Marcus's example.

Right, so in the subprocess, the bootstrap port (as Marcus described) has
been created and gets inherited down the process tree, eventually to
/hurd/ftpfs.  (Right?)

So, a program within a settrans -a (/bin/sh, in this case) can start a
single translator, like any other program.  (Right?  That looks like
what's happening there--all of the extra work has already been done by

So, if settrans added /hurd to the PATH when executing the subcommand,
it would seem to have the intended effect, at least for active

(Sorry, I'm trying to figure out where my understanding is wrong.  Hopefully
I'll get access to a Hurd box at some point and can experiment myself, but I
don't have a spare machine right now.)

Glenn Maynard

To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Reply to: