The Hurd

[Jeroen Dekkers]

On Mon, Mar 18, 2002 at 07:38:57PM +0100, Eneko Lacunza wrote:
> Hi Jeroen, 
> 
> El lun, 18-03-2002 a las 18:44, Jeroen Dekkers escribi?:
> > > 	Anyway, I think there would have been another Linux 8), but it is of
> > > course arguable.
> > I don't think so. Linus wanted to write an operating system. Almost
> > everything was already there written by GNU. He only had to write a
> > kernel and a few other things. He misnamed this to "Linux" and didn't
> > credit GNU.
> [...]
> > This was the history lesson for today.
> 
> 	I knew all those things. The thing is that I only see that you somewhat
> "hate" Linus/Linux for some reason. I don't agree in any ways with you
> in that point.

I don't hate Linux, but I don't like it either. I just say my opinion,
most of the time based on facts and things I see. If something is
wrong you're welcome to correct me. This is still different from the
things at least some Linux developers do, attacking us on things which
aren't true.

Sometimes the Hurd is blamed for a lot of things it can't really do
much about. For example that it hasn't many developers. Explainging
why everybody is developing Linux instead of the Hurd isn't easy, most
of it is just historicial. That doesn't make me hating Linux. 

> 	Don't you see that the GNU system has been greatly helped by Linux?
> Linux helped it with more testing, more popularity, more developers...

I'm not sure that wouldn't have happened without Linux.
 
> 	Maybe, even if Linus noticed the Hurd, he would have started Linux
> anyway. It would have been legitimate, so please stop with this issue.

I'm not sure. He could have worked more closely with GNU if he had
noticed that they were already developping an operating system.

> > > > GNU/Hurd is a free system and in theory it can be much more reliable
> > > > than a system with a monolithic kenrel. (Mail microkernels vs
> > > > monolithic kernels just to me and not the list, please. We are already
> > > > offtopic enough)
> > > 	True. The problem is that even their creators don't recommend it for
> > > production purposes 8)
> > Yes, if we did a lot of people would have been disappointed and would
> > say the Hurd sucks. I still recommend developping things for it. When
> > the theories are becoming practice I will start recommending it for
> > active use. :)
> 
> 	I hope so. Don't you see why is now Linux so popular? And it even now
> disappoints lots of people... can you imagine in his beginingS?

Yes, I have already seen that.
 
> 	The fact is, that promises of a superior design are not enough. Things
> must be proved, then people, MAYBE would change.

Yes, but it's not easy to prove. However, we are on the way.
 
> > > 	This is not the point in this discussion. It's not Linux's fault that
> > > the Hurd a redesign and not a reimplementation.
> > No, but it made it easier for Linux to get developers and users in a
> > short time. The developers the Hurd needs for about 10 years. The Hurd
> > finally gets them.
> 
> 	And what? This is (maybe) a sad thing, but I think that Linux has done
> an excelent work.

Not really. And then I'm just talking about the code, i.e. design and
implementation. Not about what it made possible or other social things
or so.
 
> > I think it makes sense. The Hurd could have been much better than
> > Linux is at the moment because of the OO design. When you've more than
> > hunderd developers (I don't know the number due to the fractation of
> > Linux), OO helps a bit. Structure in the development also helps. Linux
> > lacks this both. And to refer back to the original discussion,
> > bitkeeper doesn't provide this.
> 
> 	Well, I don't think this is a good place to talk about programing
> paradigms. Anyway, I think that modularization is almost as good as OO,
> sometimes even better, and I think Linux has lot's of modularization.
> Maybe the problem is the monolitic design, but I don't see it creating
> such a problem... I think it even requires that de kernel is well
> coordinated.


A monolithic kernel doesn't have well-defined and stable
interfaces. Then the modules have to change everything. Of course you
can add it to this, but in the end you'll end up with a multi-server
system running in kernel space. :)
 
> > > 	Of course we can and we must develop new design and see if they are
> > > superior to the current ones. 8)
> > Are *you* going to help? :-)
> 
> 	That's a very good question 8)
> 
> 	I would want to do lots of things, but unfortunately a day only has 24
> hours. Currently I'm more interested in promoting and defending Free
> Software, that developing new innovative system designs. I think this
> also can help the Hurd and other innovations. 8)

I've got the same problems. Only the Hurd has a very high priority for
me. :-)

Jeroen Dekkers
-- 
Jabber supporter - http://www.jabber.org Jabber ID: jdekkers at jabber.org
Debian GNU supporter - http://www.debian.org http://www.gnu.org
IRC: jeroen at openprojects
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 232 bytes
Desc: not available
URL: <http://lists.fsfe.org/pipermail/discussion/attachments/20020318/ca2b03ba/attachment.sig>