Re: Pthread assumption + starting real port

[Gernot Heiser]

On Tue, 14 Nov 2000 23:27:58 +0100, Farid Hajji <address@hidden> wrote,

farid.hajji> >   Another thing: why do you want to discard glibc so much? Is 
there a
farid.hajji> > better alternative?
farid.hajji> I don't want to _discard_ glibc right away. I just want to relax 
the
farid.hajji> dependency of the Hurd from glibc. There are numerous reasons to 
do so:
farid.hajji> 
farid.hajji> 1. glibc is much harder to port to another platform than most of 
other
farid.hajji>    GNU software. How would you port glibc to L4? Without Unix 
semantics
farid.hajji>    on the target platform, you'll need to redo the hurd sysdeps 
completely
farid.hajji>    from scratch and that's always a _lot_ of work. Then again, why 
not?
farid.hajji>    I'm not that fiercly opposed to glibc than it sounds.
farid.hajji> 
farid.hajji> 2. [ ... ]

We've been through that too. We looked at porting glibc and decided
that there's too much Unix in there. We tried OSkit, which looked
promising at first, but turned out a nightmare on non-Intel
(particularly 64-bit) systems. That was a painful and expensive
experience, we ended up dumping a lot of work.

We're now working on our own libc. SawMill is doing something similar,
and we hope to merge/share code. Doing a complete libc which is as
independent from the underlying OS as possible (obviously parts are
inherently tied to the OS) would be beneficial for lots of people.

Gernot


Gernot Heiser                ,--_|\   School of Computer Sci. & Engin.
Phone:  +61 2 9385 5156     /      \  The University of NSW
Fax:    +61 2 9385 5533     \_,--._*  UNSW SYDNEY NSW 2052, Australia
E-mail: address@hidden        v   http://www.cse.unsw.edu.au/~gernot