Re: The future of Mach
On Thu, 19 Aug 1999 21:59:24 -0000, the world broke into rejoicing as
Pablo Baena <firstname.lastname@example.org> said:
> On Sat, Aug 21, 1999 at 01:32:12AM +0200, David Lázaro wrote:
> > I remmeber reading something about deprecating GNU Mach in the future.
> > Is this true? Could somebody give a hint about what the next official
> > microkernel will be?
> > As a side note maybe somebody can explain me why Mach is not wanted
> > as the future microkernel in which to run the Hurd. Tech rationale and
> > such... Maybe you have some ace up in your sleave? :)
> > Good work, though, I thought the Hurd will have to walk a long walk
> > before running in my Laptop but it is running nice (the laptop gets
> > warmer than with Linux but it runs nonetheless).
> As far as I know, people is working on using mach since a long long time,
> (7 years?), so I'm afraid that we will have to wait again a long time
> before we had Hurd up and running with another kernel.
It certainly makes sense to make Hurd work reasonably well atop *some*
microkernel before rearchitecting to run atop something else.
But once there's enough structure there, the "main architects" (Thomas
Bushnell comes to mind for one) have had thoughts of moving to something
like L4 or Fiasco, and I suspect that there's not much in the way of
*capricious* dependancies on Mach.
> By the way, what about exo kernels? I saw there's a lot of information
> edited at the MIT, but I couldn't compile the system yet. It really
> seems to be a good thing.
Installing a new experimental OS seems to always be problematic.
Xos looks pretty interesting. It's not clear to what extent it would be
"symbiotic" with Hurd, but one of the interesting side-effects I forsee
out of the Debian/Hurd effort is that of making Debian less dependent on
OS details, which thereby may make it easier to add further OSes to the
"Win32 sucks so hard it could pull matter out of a Black Hole."
-- Pohl Longsine