[jwhat]

Hi SGI'ers, Unix'ers & GNU'ers,

in looking at SGI/BSD status, I took a tangent and decided to give GNU Hurd a go..

There is a Debian/Hurd distribution that allows you to do a virtual install on KVM.

This only runs on now old i440FX / BIOS VM which has emulated IDE HD/CDROM (obsolete HW).

In doing some historical digging I came across these quotes from Thomas Bushnell (initial Hurd architect/developer):

- RMS was a very strong believer -- wrongly, I think -- in a very greedy-algorithm approach to code reuse issues. My first choice was to take the BSD 4.4-Lite release and make a kernel. I knew the code, I knew how to do it. It is now perfectly obvious to me that this would have succeeded splendidly and the world would be a very different place today.
- RMS wanted to work together with people from Berkeley on such an effort. Some of them were interested, but some seem to have been deliberately dragging their feet: and the reason now seems to be that they had the goal of spinning off BSDI. A GNU based on 4.4-Lite would undercut BSDI.
- So RMS said to himself, "Mach is a working kernel, 4.4-Lite is only partial, we will go with Mach." It was a decision which I strongly opposed. But ultimately it was not my decision to make, and I made the best go I could at working with Mach and doing something new from that standpoint.
- This was all way before Linux; we're talking 1991 or so.

So here we are and Hurd is still plugging away to provide the last piece of the GNU puzzle.

As part of reading I come across a statement that GNU/Hurd will be looking to release 64 Bit Mach Kernel which will then have both 32 & 64 bit BSD servers sitting on top of them.

Kind of fun to read the history.

On my Debian/Hurd install I can only get CLI interface going as X11 does not start for some reason, so need to do more tinkering on the configuration.

Cheers from Oz,


jwhat/John

[robespierre]

Just another bozo with an axe to grind...
You know the saying. Success has many fathers, but failure is an orphan.

[vishnu]

Wow, groklaw, that place was huge during the SCO lawsuits but it's been dead in the water for almost 10 years now. Weird...

[jwhat]

HI Robespierre & vishnu,

I like "success has many fathers, but failure is an orphan" ;-)

But which "bozo" are you referring to, RMS ?

I have obviously led a sheltered life as I had never come across the "groklaw" site before.

Cheers from Oz,

jwhat/John

[vishnu]

I'm old enough to remember the epic flamewar between Andrew Tanenbaum and Linus Torvalds, when in the early days of Linux Tanenbaum criticized the Linux developers for creating a monolithic kernel:

https://en.wikipedia.org/wiki/Tanenbaum-Torvalds_debate

Which, obviously, the Hurd is not, but then again after 33 years of development the Hurd is still not useable, and nobody's really working on it anymore. At this point it's really just an abstract research project.

[robespierre]

It's overly generous to refer to Hurd as a research project, as it doesn't emanate from any defined programme of research. Based instead around egos of developers like Mr Bushnell, who in his own way does a fair job justifying the medieval stereotype of the venal, piggish mendicant friar.

There were working, successful Unix variants based on microkernels (including CMU Mach), including NeXTStep, OSF/1, and MkLinux. The difference is that they were developed by qualified researchers (in the case of NeXT, the legendary Avi Tevanian from CMU) and not blinkered ideologues.

[Raion]

I messed around with hurd years ago. I don't think that's enough for me to have an opinion on the technology behind it but as far as my view of the OS, it's this:

It's a lot of smart ideas with very poor execution. At this point I don't think there's any use case where GNU/Hurd is superior to GNU/Linux. It's not like say, IRIX where the appeal is similar to say tuning an old Datsun 510 with an L16, or a vintage Chevy with a 327 V8.

In my completely subjective opinion, Mach is a mistake. It's a generation one microkernel with all of the problems and drawbacks that are enumerated in the Andrew Tanenbaum v. Linus Torvalds discussion. This is not so true for later microkernels such as L3/L4. Those are far more capable and useful. The fact that Apple has dragged it along for the last 30 years (during its NeXT run even) is simply a case of throwing money at it. I'd argue that macOS and Windows actually have very similar structural flaws, both out of a desire to simplify configuration. What would you prefer to work with, the Registry, the Library or a bunch of plaintext config files in /etc and others prefixed directories? The last option may not lead to the most unified system or the most user-friendly for the lowest common denominator of computer user, but it's a far better option for most folks than trying to make a pseudo-Berkeley DB or an XML/Plist catalog with reckless abandon. Dumbing down computers has only led to more stupidity, it has not made people's quality of life better. It was far better when people had to actually learn how to use DOS or BASIC, because it kept ignorant people off of the web. Elitist, that may sound, but my tangent is over with. Ultimately I want what's best for the internet and sometimes that's not compatible with opinions of people at large, especially with how some hacker news guys masturbate over KDE or macOS aqua as the pinnacle of GUI design.

[jwhat]

Hi Vishnu,

yes I was around during the Linux / Microkernel debate and even through I understood nothing in detail, got sucked into the romance of the shiny new "Microkernel story".

But practically I just used FreeBSD and only converted to Linux (Ubuntu) when I found I was spending too much time trying to get things working on FreeBSD that ran without a hitch on Ubuntu.

So went with the flow .. as a family though we have all Apple in our household so I can get that Microkernal glow with my Macs ;-)

Raion is right about throwing money at it...

Any software project needs three basic things:
- the right people,
- the right sponsorship (ie funding) and
- manageable scope and clear sense of risks/feasibility

Cheers,


jwhat/John

[Raion]

Jwhat, FWIW I agree with you on FreeBSD being a pain. I guess for me as a certified RHEL admin I don't like taking my work home with me. Too much pain to deal with. GNU/Linux of 2022 is unrecognizable to a 2010 GNU/Linux, so I don't wanna deal with all of the new trash. No, this isn't even about stuff like systemd, but just the general way things have been set up. Hell, just 10 years ago Gentoo STILL basically required users to compile their own kernels (genkernel was buried in the docs, but now that's changed)

[jan-jaap]

Mr Bushnell, who in his own way does a fair job justifying the medieval stereotype of the venal, piggish mendicant friar.

Of course he actually is a Gregorian friar.