-
Senior Member
registered user
Jim, last night I spent five hours composing a post that explained, in my impoverished German, to someone whose mouse broke when he updated Knoppix that it was because he hadn't installed Debian instead. I have too much invested in my opinion to change it now! 
Knoppix came about because Klaus Knopper wanted a portable work environment that wasn't a portable computer - even though they no longer weigh 24 pounds like the very first, the Osborne 1. http://oldcomputers.net/osborne.html
In turn, this led to these requirements (among others):- 1 - excellent hardware detection;
2 - a boatload of applications, which in turn required:
3 - compression.
And (3) allowed a good deal of feature creep in the form of feature duplication. Just for putting words into a file, Knoppix users can choose among Joe, KWrite, Kate, NEdit, vim, Xedit, gvim, zile, OpenOffic.org Writer, and Richard Stallman's own emacs. Free as in Freedom of Choice! 
Well, after all, Unix started life as a word processor. And this heritage shows in ULOs (Unix-Like Operating systems - I release the term under the BSD license.
)
But wait - ten's not enough! Where's wily? Never mind, I'll add it myself. And so I fire up synaptic, cross my fingers, and install wily. Hmmm ... the mousewheel doesn't work. Now where in the world is the configuration file?
The German user wants to know why there are five different config files for XFree86 in /etc/X11. I want to know why xorg fails to deliver a usable window environment on either of my home computers, with every UULO I've installed that depends on it - from Slackware to Solaris.
(UULO = Unix + ULO, also released under the BSD license.)
One user wants to install Knoppix but has only 2 GB of HD available. Another has a CD drive that can't read 700-MB disks. Still another is in Spain - but Knoppix loads /etc/apt/sources.list with mirrors in Germany.
Time for a math joke.
A mathematics professor is writing away at the blackboard and then turns to the students. He says, "It is intuitively obvious that...." and then runs out of the classroom to his office. Twenty minutes later he comes back. "Yes, it is intuitively obvious that..." and he goes on to finish his point.
Sometimes it's not very obvious, but optimizing a system for one set of conditions means makiing decisions that have to make the system sub-optimal for at least some other conditions. And this applies to UULOs in general and Knoppix in particular.
LiGNUx-on-a-disk is a great portable work environment and also a great portable demo environment and a great portable forensics environment - provided that it comes stuffed with what you want, of course. And provided that you can tweak what you get into an environment you want to use.
And Knoppix is good at this, too. A refinement of requirement (1) above is that the hardware be detected and the software be automatically configured to deal with it well.
Jim, this goes to your point about making the OS install flawlessly for the novice. Heck, even a 35-year veteran (ahem) doesn't want to waste time fiddling with details that can be - and in his view should be - handled by the computer.
Do I want to stand in front of my car every morning with a hand crank to get the engine started, or do I prefer to let the battery and electric motor do the job for me? Like it says on the sign over the backscratcher manufacturing plant, "To Itch His Own"
- but I'll pass.
Here's what a bunch of experts meeting to discuss "Hot Topics in Operating Systems" have written:
In 1999, the OS researchers at HotOS identified "futz" as the most important problem facing the community.
This is at http://www.eecs.harvard.edu/~syrah/nofutz/. The page goes on:
Even though you may have a fairly good idea of what you want the system to do or how you want it to behave, there usually isn't any way to express that directly. Instead, you must change the configuration — experimentally — until the observed behavior appears to match your expectations.
This is just the tip of the iceberg - they have lots more to say about futzing.
Unix came about in one of the world's premier research laboratories and was developed and used by people who futzed for a living. Thirty years later, having the ability and tolereance (or even love) for futzing remains a badge of honor for many system administrators.
Four years before the HotOS workshop I just quoted from, a crowd (notably including Ken Thompson) from that same research outfit (Bell Laboratories) published a paper saying they'd grown tired of their Unix and had all switched over to a new OS they developed called Plan 9. Here's a passage that just cracks me up:
First, the operating system they [personal workstations] run, UNIX, is itself an old timesharing system and has had trouble adapting to ideas born after it. Graphics and networking were added to UNIX well into its lifetime and remain poorly integrated and difficult to administer.
Isn't it ironic, then, that we owe our TCP/IP internet world to a DARPA grant for UCal-Berkely to implement it in BSD Unix? 
The Plan 9 paper is at http://www.cs.bell-labs.com/sys/doc/9.html. It goes on to say:
The distinction between operating system, library, and application is important to the operating system researcher but uninteresting to the user. What matters is clean functionality.
It sounds like they got the point about futzing. But notice that they're talking about the user, not the sysadmin.
I recently installed Plan 9 on my home computers and had to do what lazy-old-me considers major futzing. A big part of the problem is inadequate documentation. The man pages are great, of course, but there's no Ignoramus Installer's Guide that starts out, "First, check the box to make sure all the parts are there...." 
It turns out that what I thought was just one computer - my Pentium III, for example - is actually four different machines: a processor server, a file server, an authentication server, and a terminal.
See, one of the team's design goals was to re-centralize system-wide resources for economy and control. Users no longer had personal workstations, they had generic, interchangeable terminals that accessed the centrally-managed servers. At the same time, central services were split out onto different machines to control both the technology that each server used and the access given to different personnel.
Plan 9 has no superuser, not even on a home computer where the dogs have more interesting things to do than change your permissions while you're away - like chewing up old books.
So the goals of the team (Bell Labs, not my dogs) don't quite match mine. Ah well. Plan 9 is fun! Try wily - it's a clone of Plan 9's acme. (There's no roadrunner in the package. In fact, if you rearrange the letters of the all-in-one environment "acme," you get "emac"....)
Okay, back to the old fogey sitting at his Pentium III and cursing at the theft-proof shrinkwrap on his shiny new set of operating-system CDs. Once he takes a butcher knife to the package, will the software meet his needs?
I don't think UULOs are there yet. Many, many people have put in untold thousands of hours so far, and there will probably be a time when I'll say yes. I'm tempted to think that the drawbacks of its Unix origins, as pointed out by the Plan 9 group, would mean that it's impossible. But hey - Windows came from PC-DOS, and lots of people seem able to work with that, so who am I to say? 
Knoppix-on-CD sets a very high standard for other distros to meet in the area of futz-busting. But it loses too much ground when it's installed, even to the point of being a detriment to converting the very people who saw the CD and decided to give it a try.
The good news is that 2004 seems to have been the watershed year when LiGNUx in particular, among the ULOs, was finally, openly taken seriously. It's made a dent in the server market that can't be ignored, it has nearly monopolized the compute-speed supercomputer sector, and it has people from Microsoft and the SCO Group (the artist formerly known as Caldera) competing to see who can be the biggest fools in public.
Heck - Sun not only released Solaris 10 for free download, it targeted the LiGNUx core "i386" platform.
So yeah, the beachhead is secure and LiGNUx isn't going away. Three key groups of customers in U.S. business (plus governments in the rest of the world) know that it has to be included in any study of options.
One group of customers is the people who manage corporate computer systems. It's a good thing they weren't too hard to sell.
Another is the people who manage the money in a business. They see that Microsoft is pouring lots of resources into arguing over Total Cost of Ownership, so they know that the Bill & Steve show doesn't have a sure bet.
Finally, there are the people who manage people. Unfortunately, I've met far too many who can't actually manage people. But that's beside the point, really. If they're told that the new system will have the same functionality and enough interface similarity to make transition something that they can gripe about without actually worrying about loss of worker-bee productivity, they're okay with it.
And that leaves only the people who manage their own cubicle, be it in a corporation or in their living room. These folks are the final frontier for UULOs, and really, the toughest group of all.
For them, an OS is something personal. It's the foundation for all the other tools they use to achieve what they're trying to do. Some of their goals they can set for themselves, and others are imposed on them - but they're held accountable. So the OS just has to get out of their way and do its job.
Well shoot. I've taken more of your time than I meant to. But since you're here - have you tried Debian sarge?
Talk about coming a long way! Debian has made a huge impression on me lately. It was the first distro I installed way back when, and the experience was terrible.
With Debian's new installer (not the one in the stable release), things really go smoothly. And the hardware detection is almost as good as Knoppix - in my experience, anyway. I had to tweak exactly one setting to get screen resolution one level higher.
Progeny Debian - the new, free one - is also good. Didn't need any hardware tweaks, just pointed /etc/apt/sources.list to Debian sites instead of Progeny's. Give 'em a try!
-- Ed
Similar Threads
-
By brigi in forum Hardware & Booting
Replies: 4
Last Post: 01-26-2006, 05:02 AM
-
By Jolle in forum Hardware & Booting
Replies: 1
Last Post: 12-21-2005, 06:53 PM
-
By gonzo in forum General Support
Replies: 2
Last Post: 03-24-2005, 05:26 PM
-
By paoloskater in forum Networking
Replies: 0
Last Post: 01-24-2005, 03:26 PM
-
By knoppix-accessibility in forum Customising & Remastering
Replies: 3
Last Post: 12-13-2004, 01:26 AM
Posting Permissions
- You may not post new threads
- You may not post replies
- You may not post attachments
- You may not edit your posts
-
Forum Rules

Vintage 1998 iMac G3 M4984 Mac OS + Keyboard Working but Needs Some Repair Read
$295.00

VINTAGE Apple Mac OS X Computer Monitor
$250.00

Vintage Apple iMac G4 15" Lamp 700MHz 512MB Mac OS X 10.4.6 W/ Pro Speakers
$300.00

Vintage Apple PowerBook 15” G4 “Aluminum” (1.25GHz/2gb/150gb/Tiger) 2
$109.99

185--Sonnet Crescendo G3 PPCG3-375-5-K-06 G3 Add-On Card for Vintage Mac
$90.00

Apple Macintosh Plus 1MB M0001A Vintage Computer **TURNS ON **
$99.99

Apple Macintosh Plus 1MB M0001A Vintage Computer Boots Up / Floppy Disk Stuck
$99.99

Macintosh MacTerminal M0521 Version 1.0. Vintage Apple software 1984 Sealed
$129.99

Apple Power Macintosh PowerPC 7300/200 VTG w Keyboard Mouse Storage Drive Read
$199.99

Vintage Mac OS X v10.2.3 Install Disc 2 and Mac OS X Developer Tools Untested
$17.99