:30:11
uh, while in the microkernel,
:30:14
the, the operating system kernel is actually
:30:19
uh, just a collection of servers that
:30:23
do different things and then they have a common protocol
:30:26
for doing communication between themselves.
:30:29
[ So why is that... the GNU project that's had
so much lead-time, that's been doing this,
:30:36
Why...Why is it that he was able to kinda
come in at the tail end so to speak ]
:30:40
Well we actually started the
GNU Hurd not long before he started Linux.
:30:48
And it happened though we chose a design
that's a very advanced design
:30:53
in terms of the power gives you
:30:56
but also turns out to be very hard to debug.
:31:00
We decided to divide up the kernel
which traditionally had been one program,
:31:06
to divide it up into a lot of smaller programs
:31:09
that would send messages to
each other asynchronously to, to communicate.
:31:16
The problem is that, that style of programming
:31:21
has a great deal of potential for bugs,
:31:24
which are often very hard to
figure out because they depend on...
:31:27
does this program send this message
before or after this one sends that message...
:31:36
And the result was:
it took us years to get the thing to work.
:31:44
[ What is Linux's relationship to the GNU project? ]
:31:49
Well there's relationships to GNU
on kind of multiple levels.
:31:57
One is just the philosophical level of thinking that