Slashdot Mirror

because not all of us are ESR ... by choice

i mean, going from a nancy boy, frolicking with guns, to picking up chicks. it's a choice my man.

oh, and so to not be OT, here

Many Americans find anime a lot more appealing than a lot of stuff on TV because anime isn't as constrained by the American Standard that effects many of the popular programming.

That's definately part of the appeal. I think another part is that most of the crud (See Sturgeon's Law) never gets exported to the US. We see the same effect with British TV shows. Monty Python, Doctor Who, Red Dwarf, etc make it over here, but most of the bad stuff doesn't.

If you look back at mainstream American TV, there's been a fair amount of really good stuff (e.g. Babylon 5, original Star Trek, Twilight Zone, Buffy) and a lot of decent stuff (I won't list examples out of fear of controversy). We just don't notice because we're getting all the crud mixed in with it. Unfortuantly, throught the miracle of American Media Dominance, the rest of the industrialized world gets our crud unfiltered. Those countries where English is not the primary language get our crud badly dubbed on top of everything, but I digress.

Another factor is the way Anime is promoted in the US: almost entirely by word of mouth recomendations. By only watching stuff that our tasteful - if more adventurous - friends have pre-screened, we avoid whatever crud does manage to make its way over here.

Well if we, on the planet, constitute a group brain, then what about all the squirrels and iguanas and frogs and lions and wombats and single-celled organisms out there? Huh? Where do they fit in?

I don't think the group brain idea holds water--or spinal fluid. I think if we're a group brain, we've got an awfully bad case of... you name it: Multiple Personality Disorder, Bipolar Disorder, (and the biggie) Schizophrenea (sp?).

Nah, really though, I mostly agree with the other people who repsonded to your post. Sorry. Collective Unconscious idea has worn out it's welcome. Why not try memes instead?

I am not well-versed in the world of Linux, ( have my own allegiances but am being drawn to it more and more. Reading the article, it felt very clear to me that Linux will prevail (with a nod to William Faulkner's Nobel speech).

Consider a few quotes from the article:

The LinuxThreads implementation of the POSIX threads standard (pthreads), originally written by Xavier Leroy

A group at IBM and Intel, led by Bill Abt at IBM, released the first version of the New Generation POSIX Threads (NGPT) library in May 2001

On March 26-27, 2002, Compaq hosted a meeting to discuss the future replacement for the LinuxThreads library. In attendance were members of the NGPT team, some employees of (then distinct) Compaq and Hewlett-Packard, and representatives of the glibc team

On September 19, 2002, Ulrich Drepper and Ingo Molnar (also of Red Hat) released an alternative to NGPT called the Native POSIX Thread Library (NPTL)

Perhaps others have already pointed this out, but I am newly impressed with the universal nature of Linux. The power of an operating system that *everyone* is interested in improving, and has the opportunity to improve, is awesome. Yes, Microsoft has tremendous resources, and very earnest, good-willed, brilliant people. But to improve Microsoft's kernels, you have to work for Microsoft. That means switching the kid's schools, moving to Redmond, etc. etc. On the other hand, everyone from IBM to HP to some kid in, say, Finland, can add a good idea to Linux. When the kernel's threads implementation is a topic for conversation at conferences, with multiple independent teams coming up with their best ideas, Linux is sure to win in the long run.

I'm struck by the parallels to my own field of scientific research: Yes, the large multinational companies have made tremendous contributions in materials science, seminconductors, and biotech. They work on the "closed-source", or perhaps "BSD" model of development. But it is the "GPL"-like process of peer-reviewed, openly shared, and collaborative academic science that has truly prevailed.

See this URL for more details.

Sir, you've officially confirmed yourself a zipperhead

The ideal mix is an armed citizen. Ask ESR!

Perhaps I'm being a bit nitpicky, but wouldn't STFW be the correct response? If the switch is not documented in the manual, RTFM is incorrect.

Say what you like about Gates and Microsoft, but the fact remains that in dollar terms, he's done far more for worthy causes than the typical Open Source advocate

Sure, he is giving a lot of money to his foundation.

On the other hand, according to this source, he is worth more than US$ 60bn.

And, according to this other source, our charitable friend Bill Gates makes about US$ 31 per second.

I don't think RMS, Linux, or ESR wealth or income will ever come close...

So, for Mr Bill Gates, giving US$ 1.2bn per year is... what? Giving away 1/50th of his total worth per year?? Now, that's pretty generous.

Don't misunderstand me: I truly thing it's generous. But you have to put this into perspective, especially when it comes to your comment about ESR. I personnaly think the article you referenced sipply means ESR is determined to enjoy his money... while we enjoy, for free, the software he created.

A few years ago that was Eight instead of Eighty. Apparently Moore's Law is affecting the editor holy wars as much as the machines!

No, it isn't. Neither ANSI nor ISO C allows main to return anything but an int.

On most systems, that will even link and run without complaint.

I shudder to think about what you consider acceptable C, if you have to qualify "link and run without complaint" with "even".

But your system isn't the C standard, and the fact that it runs there doesn't mean it is valid C. Most systems allow i = i++ += ++i as well. The result is undefined, and that means that anything is allowed to happen (including program crash, spontaneous massive existence failure, making demons run out of your nose, or simply set i to some guessable value).

The fact that most calling conventions are sane enough for it not to matter defining a function returning nothing when the run-time-system really expects one returning success or failure as an integer, doesn't mean that it is valid (or that it makes sense, i.e.: what will your C program return to the OS after running?)

Some systems with ancient preludes won't take it.

So what you are saying is that, even though it is explicitly not allowed by the standard, it also doesn't work at all on some systems?

Such systems generally won't like the perfectly valid int main(int argc, char **argv, char **env) convention either.

Surprise. It isn't perfectly valid C. I'm not even shure if it's POSIX (although extern char **environ is). But it is pretty common among unixes.

I suppose they're letting the magic smoke out prematurely.

(Lifted from the Jargon File)

Magic Smoke - n. A substance trapped inside IC packages that enables them to function (also called `blue smoke'; this is similar to the archaic `phlogiston' hypothesis about combustion). Its existence is demonstrated by what happens when a chip burns up -- the magic smoke gets let out, so it doesn't work any more. See smoke test, let the smoke out.

Usenetter Jay Maynard tells the following story: "Once, while hacking on a dedicated Z80 system, I was testing code by blowing EPROMs and plugging them in the system, then seeing what happened. One time, I plugged one in backwards. I only discovered that *after* I realized that Intel didn't put power-on lights under the quartz windows on the tops of their EPROMs -- the die was glowing white-hot. Amazingly, the EPROM worked fine after I erased it, filled it full of zeros, then erased it again. For all I know, it's still in service. Of course, this is because the magic smoke didn't get let out." Compare the original phrasing of Murphy's Law.

I don't know if you're trolling, but personally, I can't stand that "K". I don't mind Gnome's "G" everywhere, but I don't like KDE's "K" at all.

Visually, "K" is just an annoying, ugly letter, all kinds of sharp edges, and it doesn't brighten my day the way a nice "g" or "i" does. Just take a look at that "g", if you've got the right font, it's like a beautiful woman. You can't even get away from "K" in lowercase: "k" looks just like "K". It's like somebody getting kneed in the crotch, or something. (Kneed - begins with "K", ouch!).

When you say it out loud, it makes everything heavy and hard, like something from another language. Konsole. Konqueror. TheKompany. Though each day I am thankful I don't have to put up with a "Kalendar" or a "Klok", or, heaven forfend, a "Kalkulator". The other day, I found myself thinking of programming in kvikkalkul or plankalkül. Skary!

(What do non-English native speakers think of the "K"?)

Well, yes this sounds incredibly stupid, but I avoided KDE for a long time simply because of my strong anti-K stance. No marketing department would ever overuse "K" the way KDE has. Now that I've been using it since I installed RH7.3, I've gotten a little used to it, but man, I'd kill (kill - begins with K) for a nice soft "o", that would be so nice and comfy, maybe even a little funny, like Santa Claus laughing.

(Can you tell I've been programming all weekend?? You know, after you stare at letters for hours on end, they start to stare back....)

• This package implements a fast Bayesian spam filter along the lines suggested
• 
  by Paul Graham in his article "A Plan For Spam".

Bogofilter has been out since august, and does this bayesian spam-stuff in C, which probably will run a bit faster than the perl or python versions just because of it's compiled-ness. I've never run it myself, but people on debian lists say it works better or not as good as spamassassin.

This isn't exactly the first bayesian mail filter out there. I've been using ESR's bogofilter for weeks now, and I must say it works better than I could have ever imagined. Bogofilter however is simply for sorting out spam, while it appears this filter can sort out other things. But honestly, I can setup some simple filters to separate personal emails from work emails, so I'm not entirely sure the extra stuff is that useful.

Yes. You missed the point. Back in IBM's heydey during the Iron Age of computing, metered CPU billing was the way most people got their access. This isn't a new idea at all -- it's a comparatively ancient idea being resurrected.

Everyone knows about BSD's failure and imminent demise. As we pore over the history of BSD, we'll uncover a story of fatal mistakes, poor priorities, and personal rivalry, and we'll learn what mistakes to avoid so as to save Linux from a similarly grisly fate.

Let's not be overly morbid and give BSD credit for its early successes. In the 1970s, Ken Thompson and Bill Joy both made significant contributions to the computing world on the BSD platform. In the 80s, DARPA saw BSD as the premiere open platform, and, after initial successes with the 4.1BSD product, gave the BSD company a 2 year contract.

These early triumphs would soon be forgotten in a series of internal conflicts that would mar BSD's progress. In 1992, AT&T filed suit against Berkeley Software, claiming that proprietary code agreements had been haphazardly violated. In the same year, BSD filed countersuit, reciprocating bad intentions and fueling internal rivalry. While AT&T and Berkeley Software lawyers battled in court, lead developers of various BSD distributions quarreled on Usenet. In 1995, Theo de Raadt, one of the founders of the NetBSD project, formed his own rival distribution, OpenBSD, as the result of a quarrel that he documents on his website. Mr. de Raadt's stubborn arrogance was later seen in his clash with Darren Reed, which resulted in the expulsion of IPF from the OpenBSD distribution.

As personal rivalries took precedence over a quality product, BSD's codebase became worse and worse. As we all know, incompatibilities between each BSD distribution make code sharing an arduous task. Research conducted at MIT found BSD's filesystem implementation to be "very poorly performing." Even BSD's acclaimed TCP/IP stack has lagged behind, according to this study.

Problems with BSD's codebase were compounded by fundamental flaws in the BSD design approach. As argued by Eric Raymond in his watershed essay, The Cathedral and the Bazaar, rapid, decentralized development models are inherently superior to slow, centralized ones in software development. BSD developers never heeded Mr. Raymond's lesson and insisted that centralized models lead to 'cleaner code.' Don't believe their hype - BSD's development model has significantly impaired its progress. Any achievements that BSD managed to make were nullified by the BSD license, which allows corporations and coders alike to reap profits without reciprocating the goodwill of open-source. Fortunately, Linux is not prone to this exploitation, as it is licensed under the GPL.

The failure of BSD culminated in the resignation of Jordan Hubbard and Michael Smith from the FreeBSD core team. They both believed that FreeBSD had long lost its earlier vitality. Like an empire in decline, BSD had become bureaucratic and stagnant. As Linux gains market share and as BSD sinks deeper into the mire of decay, their parting addresses will resound as fitting eulogies to BSD's demise.

A good place to start (without having to get a CS degree) would be reading How to Report Bugs Effectively, and of course How To Ask Questions The Smart Way.

eventually, according to Gate's Law, you'll need a Dual 3Ghz with thousands of megs in RAM just to run a word processor.

I recently wrote an BrainFuck interpreter in INTERCAL. I keep it here and it's an evil evil beast. :D

Several years back I also wrote a relatively full-featured BBS, complete with message boards, file areas and dropfile support for door games, in QBASIC, as well as a DataPac network scanner in same. :/

When I first saw it, I was amazed that it worked, but I would say it pushed the limits

Code example and discussion in the Jargon File

For a more detailed explation see here.

Can't post the code, due to Lameness filter.

Suits Generate Bogons! so beware!

Suits Generate Bogons! so beware!

Suits Generate Bogons! so beware!

grok /grok/, var. /grohk/ vt.

[common; from the novel "Stranger in a Strange Land", by Robert A. Heinlein, where it is a Martian word meaning literally `to drink' and metaphorically `to be one with']

The emphatic form is `grok in fullness'.

1. To understand. Connotes intimate and exhaustive knowledge. When you claim to `grok' some knowledge or technique, you are asserting that you have not merely learned it in a detached instrumental way but that it has become part of you, part of your identity. For example, to say that you "know" LISP is simply to assert that you can code in it if necessary - but to say you "grok" LISP is to claim that you have deeply entered the world-view and spirit of the language, with the implication that it has transformed your view of programming. Contrast zen , which is similar supernal understanding experienced as a single brief flash. See also glark .

2. Used of programs, may connote merely sufficient understanding. "Almost all C compilers grok the void type these days."

--

What can ya say? Im a Karma whore...

grok /grok/, var. /grohk/ vt.

[common; from the novel "Stranger in a Strange Land", by Robert A. Heinlein, where it is a Martian word meaning literally `to drink' and metaphorically `to be one with']

The emphatic form is `grok in fullness'.

1. To understand. Connotes intimate and exhaustive knowledge. When you claim to `grok' some knowledge or technique, you are asserting that you have not merely learned it in a detached instrumental way but that it has become part of you, part of your identity. For example, to say that you "know" LISP is simply to assert that you can code in it if necessary - but to say you "grok" LISP is to claim that you have deeply entered the world-view and spirit of the language, with the implication that it has transformed your view of programming. Contrast zen , which is similar supernal understanding experienced as a single brief flash. See also glark .

2. Used of programs, may connote merely sufficient understanding. "Almost all C compilers grok the void type these days."

--

What can ya say? Im a Karma whore...

grok /grok/, var. /grohk/ vt.

[common; from the novel "Stranger in a Strange Land", by Robert A. Heinlein, where it is a Martian word meaning literally `to drink' and metaphorically `to be one with']

The emphatic form is `grok in fullness'.

1. To understand. Connotes intimate and exhaustive knowledge. When you claim to `grok' some knowledge or technique, you are asserting that you have not merely learned it in a detached instrumental way but that it has become part of you, part of your identity. For example, to say that you "know" LISP is simply to assert that you can code in it if necessary - but to say you "grok" LISP is to claim that you have deeply entered the world-view and spirit of the language, with the implication that it has transformed your view of programming. Contrast zen , which is similar supernal understanding experienced as a single brief flash. See also glark .

2. Used of programs, may connote merely sufficient understanding. "Almost all C compilers grok the void type these days."

--

What can ya say? Im a Karma whore...

grok /grok/, var. /grohk/ vt.

[common; from the novel "Stranger in a Strange Land", by Robert A. Heinlein, where it is a Martian word meaning literally `to drink' and metaphorically `to be one with']

The emphatic form is `grok in fullness'.

1. To understand. Connotes intimate and exhaustive knowledge. When you claim to `grok' some knowledge or technique, you are asserting that you have not merely learned it in a detached instrumental way but that it has become part of you, part of your identity. For example, to say that you "know" LISP is simply to assert that you can code in it if necessary - but to say you "grok" LISP is to claim that you have deeply entered the world-view and spirit of the language, with the implication that it has transformed your view of programming. Contrast zen , which is similar supernal understanding experienced as a single brief flash. See also glark .

2. Used of programs, may connote merely sufficient understanding. "Almost all C compilers grok the void type these days."

--

What can ya say? Im a Karma whore...

But there's nothing to download right now.

What does it actually do right now? Sounds like vaporware to me! To treat it as anything but pie-in-the-sky fantasy is a great disservice to all the things that actually exist right now...

According to Steve van Dulken, who oversees the patent archive at the British Library, 'For every 100 applications lodged, I'd say that 10 are a bit whacky.' And, of course, the other 90 are crap :)

It's arguably the first RPG in the universe

Nah. ADVENT beats it by a decade.

This reminds me of a humorous story:

The Magic Switch

I can't remember the author (GLS) but if you google you'll probably find a more original version.

The good thing about the knowlege industry is that not only can you telecommute. It's the fact that you can share the knowlege and be richer for it.

Beyond just giving them the source, you've got to make sure that you make every effort to make the recipients of your aid part of the team in the cathedral or at least feeling like they're part of the bazaar.

There's nothing worse than sending in aid that makes the person wind up with this big shiny thing that they don't have the resources to maintain or expand on.

So yeah. Clean water first. Food second. No war third. Good medicine, industrial infrastructure, a reliable democratic and open government... and then technology that the developing country can really feel that they own, rather than that they adopted because they found it or someone gave it to them.

I hereby invoke Godwin's Law on this article.

I just finished Kindergarten, where we learned our ABC's, and I've invented an alphabet that can be read 7 times as fast, with only one third the effort. Why didn't some PhD come up with this before?

There are alphabets that are better than Latin in some respect. Take Tengwar for instance. The script is designed along sound phonological principles (no pun intended): voiced consonants, fricatives, and nasals have a predictable change in shape from the basic voiceless stops (p, t, c, q). It's been adapted to at least English, Sindarin, Quenya, Polish, Lojban, Esperanto, and Toki Pona. On the other hand, some people have expressed increased dyslexia due to use of Tengwar.

I just finished Kindergarten, where we learned our ABC's, and I've invented an alphabet that can be read 7 times as fast, with only one third the effort. Why didn't some PhD come up with this before?

There are alphabets that are better than Latin in some respect. Take Tengwar for instance. The script is designed along sound phonological principles (no pun intended): voiced consonants, fricatives, and nasals have a predictable change in shape from the basic voiceless stops (p, t, c, q). It's been adapted to at least English, Sindarin, Quenya, Polish, Lojban, Esperanto, and Toki Pona. On the other hand, some people have expressed increased dyslexia due to use of Tengwar.

The thing that makes MP3 support so simple is the fact there are manufacturers will have chips that can decode MP3 streams for very little cost. These chips (available from Philips, Texas Instruments, and other big-name chip manufacturers) also feature an on-chip USB controller. Functionally rich - low profile - low power usage - low implementation costs. The current array of MP3 players probably share more guts than one would think.

The system-on-a-chip design eliminates the need to bundle a more expensive embedded CPU and support a software decoder via firmware. While this is indeed a better option in the long run (due to flexibility and upgrades), it could mean the difference of $1.50 per chip versus $12.00 per CPU/flash combination. Sell 10,000 units and you can see where the savings appears. Of course, take into consideration the elimination of having to implement decoding a stream using multiple chips and interfaces, the savings are exponential. Companies are about the bottom dollar - minimum cost, maximum profit.

The fact that MP3 has been so popular over the years (the first foot in the door, so to speak), it only makes sense to cater to the 95+% of those people out there that still have collections made of solely MP3s. A smart company would take into consideration emerging technologies, but nonetheless still focus on the popular demand - the one to make the most profit.

If you visit Xiph's OGG Vorbis hardware support page you can help get the word out that you want OGG support. I can only assume that with the release of Tremor and Xiph's pledge to give free engineer to time to companies that a company would be foolish not to take advantage of free development. If you take the Slashdot Effect into consideration, eventually the emails and phone calls for OGG support in future devices will be heard, quickening the availability of such devices.

Hitler was very determined and aggressive

you have now lost the argument.

Godwin's Law in the Jargon File

Everyone knows about BSD's failure and imminent demise. As we pore over the history of BSD, we'll uncover a story of fatal mistakes, poor priorities, and personal rivalry, and we'll learn what mistakes to avoid so as to save Linux from a similarly grisly fate.

Let's not be overly morbid and give BSD credit for its early successes. In the 1970s, Ken Thompson and Bill Joy both made significant contributions to the computing world on the BSD platform. In the 80s, DARPA saw BSD as the premiere open platform, and, after initial successes with the 4.1BSD product, gave the BSD company a 2 year contract.

These early triumphs would soon be forgotten in a series of internal conflicts that would mar BSD's progress. In 1992, AT&T filed suit against Berkeley Software, claiming that proprietary code agreements had been haphazardly violated. In the same year, BSD filed countersuit, reciprocating bad intentions and fueling internal rivalry. While AT&T and Berkeley Software lawyers battled in court, lead developers of various BSD distributions quarreled on Usenet. In 1995, Theo de Raadt, one of the founders of the NetBSD project, formed his own rival distribution, OpenBSD, as the result of a quarrel that he documents on his website. Mr. de Raadt's stubborn arrogance was later seen in his clash with Darren Reed, which resulted in the expulsion of IPF from the OpenBSD distribution.

As personal rivalries took precedence over a quality product, BSD's codebase became worse and worse. As we all know, incompatibilities between each BSD distribution make code sharing an arduous task. Research conducted at MIT found BSD's filesystem implementation to be "very poorly performing." Even BSD's acclaimed TCP/IP stack has lagged behind, according to this study.

Problems with BSD's codebase were compounded by fundamental flaws in the BSD design approach. As argued by Eric Raymond in his watershed essay, The Cathedral and the Bazaar, rapid, decentralized development models are inherently superior to slow, centralized ones in software development. BSD developers never heeded Mr. Raymond's lesson and insisted that centralized models lead to 'cleaner code.' Don't believe their hype - BSD's development model has significantly impaired its progress. Any achievements that BSD managed to make were nullified by the BSD license, which allows corporations and coders alike to reap profits without reciprocating the goodwill of open-source. Fortunately, Linux is not prone to this exploitation, as it is licensed under the GPL.

The failure of BSD culminated in the resignation of Jordan Hubbard and Michael Smith from the FreeBSD core team. They both believed that FreeBSD had long lost its earlier vitality. Like an empire in decline, BSD had become bureaucratic and stagnant. As Linux gains market share and as BSD sinks deeper into the mire of decay, their parting addresses will resound as fitting eulogies to BSD's demise.

Godwin's Law

Try to 'urge' those out of my possession.

rubber-hose cryptanalysis n.

[sci.crypt newsgroup] The technique of breaking a code or cipher by finding someone who has the key and applying a rubber hose vigorously and repeatedly to the soles of that luckless person's feet until the key is discovered. Shorthand for any method of coercion: the originator of the term drily noted that it "can take a surprisingly short time and is quite computationally inexpensive" relative to other cryptanalysis methods. Compare social engineering, brute force.

During the episode of Scrapheap Challenge where the N.E.R.D.S. built Frobette (see also frob) they handed you and Robert a copy of The New Hacker's Dictionary. Did you keep it?

How did you get to be so insanely cool?

I don't mean to sound sexist, but I hope you definitely plan on having kids, even though it's exceedingly painful to give birth and exceedingly difficult to raise children. It's important for genes like yours to stay in circulation. Really we need them, badly! I mean, look at George W. Bush! Aaaagh!

Although, now that I think about it, it's also possible to help out your fellow humans in lots of other ways, and you've certainly done a lot of that.

Thank you for existing. Your presence has enriched all of our lives. I hope yours is going well.

During the episode of Scrapheap Challenge where the N.E.R.D.S. built Frobette (see also frob) they handed you and Robert a copy of The New Hacker's Dictionary. Did you keep it?

How did you get to be so insanely cool?

I don't mean to sound sexist, but I hope you definitely plan on having kids, even though it's exceedingly painful to give birth and exceedingly difficult to raise children. It's important for genes like yours to stay in circulation. Really we need them, badly! I mean, look at George W. Bush! Aaaagh!

Although, now that I think about it, it's also possible to help out your fellow humans in lots of other ways, and you've certainly done a lot of that.

Thank you for existing. Your presence has enriched all of our lives. I hope yours is going well.

An Unbiased Review of Debian 3.0

This is a critical review of Debian 3.0, but I want to say right from the start that I'm not trying to bait anyone. However I feel that reviewers often root for Debian as the open-source underdog, and give it marks which it doesn't deserve. If RedHat 8.0 came out with installation software like Debian 3.0 it would be savaged. I think it's time for an honest review, to spur the Debian developers into making the best possible distribution. I really want Debian to succeed. I want to use it daily, and recommend it to my friends. But I can't do that right now and I think it's important people understand why.

Installation
My first experience of Linux came with a boxed version of SuSE 6.0, back in the middle of 1999 when Linux was starting to get noticed in a big way. The entire thing was a text-mode affair, powered by the venerable YaST version 1. I spent days just poring through the manual, trying to wrap my head around fdisk, and hoping it would all turn out okay. It did, and I never looked back. Six months later a version of RedHat (five point something or the other I think) was shipped with a magazine I bought, and I gave it a whirl. This too was backed with a text-based installer, but it was a lot easier to use than YaST. I didn't even bother with the documentation, I just slipped it in the CD drive and winged it. Shortly thereafter I tried the first version of Mandrake, which had pretty much the exact same installation process..

The point of all this reminiscing is to show that I'm not a complete neophyte (though I'm nowhere near being a guru for that matter). Since then I've tried the RedHat and Mandrake graphical installs, and while RedHat is the one I like best, Mandrake has been the distribution I've stuck with solely because of drakconf and it's associated tools, which make configuring a Linux system a breeze. However lately I've been aspiring to ascend to guru status, or at the very least PFY, so I gave Debian a whirl. I have to admit I was disappointed both with the installation procedure and the finished system. In all my time with Linux, Debian's is the worst installer I've ever had to use.

Setup
There is a lot wrong with it, but mainly the fact is that it's an awfully stupid piece of software. And I don't mean stupid as in bad, I mean as in not clever. It expects the user to know everything. So, for example, even though XFree86 has fully documented the branded names that each driver supports, Debian simply supplies a list of the driver names themselves. People with, say, a GeForce card packaged by Creative will have a hard time picking the nv driver. However they should be glad that they have a choice at all - a lot of screens only give highly technical examples and refer the users to documentation that hasn't even been installed yet! For example why couldn't a list of keyboards, e.g. Irish Keyboard, US Keyboard, Sun US Keyboard etc. be given instead of expecting the user to type in xfree86, pc105, ie with uk as alternative.

This is simple fundamental stuff, the kind of thing most other distros had sorted out back in '99 when everything was via textmode and the Linux GUI was new and exciting. However, in this day and age, I would expect far more from a distribution. There should be no need for me to enter in the same locale based settings over and over again. Once I'd selected Europe->Western->Dublin as the timezone, the system should have realised that the appropriate locale was en_IE@euro, that the keyboard should be set up with proper Euro support (it doesn't seem to be, AltGr is mapped as Alt so I can't easily print bars, the Euro symbol, or accents for stuff I write in Irish), that the Euro packages should be installed by default (they weren't) and a whole raft of other tiny stuff like KDE and Gnome localisation. Certainly people should be presented with the chance to confirm these options, but it should be a simple matter of hitting Enter most of the way. If they want to change the default, they should first be presented with a list of preconfigured settings for, e.g. keyboards, out of which they can then opt into the sort of technical xfree86, pc105, etc. settings.

This willfull stupidity of the installer extends to other aspects of the setup also - with so many kernels available, Debian should pick the most appropriate one to use for my system. It's not that hard to open up /proc/cpuinfo. Instead I was confronted with a maze of kernels once I got to the software selection stage, installed 2.4.18, and then belatedly realised that only 2.4.16 had the ALSA drivers I wanted. Why not offer two defaults in the final base install screen Kernel-2.2.20-$arch and Kernel-2.4.16-$arch (where $arch is the probed value of the most suitable CPU) with a third option to select the kernel yourself. And for the record, I have no idea what the point of the modules page was - was I meant to manually install each and every module?!

Package Selection
This brings me nicely along to package selection. Tasksel wasn't too bad, though I'd expect more options. For example, instead of X11 have X11, Typical Desktop (Gnome & KDE) and Esoteric Desktop (WindowMaker and Enlightenment) and so on. I was mystified to see I could select Fortran and Tcl/Tk support, but not Perl, PHP, or Java - some of the most popular languages today. However nothing, not in all my 22 years on this Earth, could prepare me for the horrors of dselect. Sweet merciful divine!

Firstly the developers should check out Eugenia's comments on osnews.com about the new Yast2 package manager, as many of the same things apply. In the end it all boils down to the old KISS clich, keep it simple! Instead of giving a load of choices for dependency resolution with half a million optional packages thrown in, just give n + 1 choices, one for each of the n package/package-combinations that fixes the dependency, and one to install without resolving it. Similarly with conflict resolution it should be remove selected, remove conflicting or ignore.

Worse yet are the help screens that pop up at every opportunity, yet which don't actually explain everything (like the meaning of those EIOM headers at the top of the screen). At the end of the day, it should be fairly obvious what's going on. Leave complex package selection tools for the post install, at this stage people just want to get the damn thing working. It drove me nuts having to pass through that stupid help screen every time a dependency arose.

What's worst of all is that if, for example, dselect fails to download a package from the Internet, it prompts the user with a basic text mode question asking them if they want to cancel. I assumed this meant just cancel that particular package. It didn't, and I found myself dumped into the console on a base system. I knew enough to extricate myself, but this is hardly something the average newbie is going to be able to cope with.

The Installation Overall
I want to make sure people realise I'm not trying to advocate a graphical installer. It would be a good move ahead, and should be available for Debian 4.0, but all the stuff I've mentioned here could be easily implemented in a text-mode installer written using ncurses. In fact, I would recommend a Model-View-Controller approach, with the Model, the bit that does all the actual work, being packed into a library, and two Views being created with, say, ncurses and Qt, each of which uses the Model library to do what's needed.

Debian's installer does have some redeeming features. For one thing it is rock solid. With several versions of Mandrake I have had proble ms setting up the mouse and getting the package selector to install all the selected packages. This didn't happen in Debian. Downloading updates from the web during the install is also a great idea (though I was a little aghast to find my 56K modem facing into 100M of updates). The provision of non-free sites is a great help, given the conflict between Debian's all-free stance and the wants of the average user.

The crucial factor is that the installer should be made as intelligent as possible, and to hide the actual de tails behind Advanced buttons. Guess as much as possible from initial locale data. Use branded names instead of driver names for hardware, be it keyboards, mice, graphics cards or soundcards. I hadn't mentioned this but Debian should aim to have sound working as a default in every new installation, prompting users for their soundcard make from a list in a similar in fashion to the XFree one. In this day and age, every OS should have sound support. By all means, let one of the brands on the list be No Soundcard, but offer to install and configure it at any rate.

Dselect needs to be totally re-designed. I can appreciate its power, but it's far to complex and hard to use. Aim to replicate the way things work in graphical GUIs - have drop down lists and checkboxes which can be ticked to install items, even if said boxes are represented by [ ] and [X]. There is a case to be made for complex package installation software, but half way through an OS install isn't really the place.

The Configured System
Having finally got everything installed, I was, I confess, pretty disappointed with the results. Bugs started appearing. Firstly, when selecting the Irish locale in KDE 2.2.2, I found KDE trying to tell me that the Irish currency was the pound, something which hasn't been the case since the Euro was introduced in 2000, two and a half years ago. Then kwrite decided it wouldn't display documents it opened and konqueror decided all pages should be 2000 pixels wide, even though the window was about 800.

Sound didn't work, and consequently the KDE bootup screen stalled for ages at the window manager stage while arts slowly died, then popped up a No Sound message box. None of the PPP connection tools wor ked when not used by root. None of the hard disk partitions were configured (even though they had been recognised by the piece of code that set up LILO). My CDRW at /dev/hdd wasn't set up, not even as a plain CD-ROM. The menus were all over the place. The fonts in GTK apps were hideously big. XftConfig wasn't set up to disable antialiasing for standard size fonts, nor were the workarounds for symbol and console fonts (mentioned here) included. Another bug.

It was a mess.

Firstly the menus. In Enlightenment and Gnome you have a special Debian menu included with the rest in the app launchers. These menus contain everything. Thus, when you're looking for a program, you just go to the Debian menu and it's all gravy. However the Debian menu wan't included in KDE, instead there were a load of Debian submenus, which didn't seem to include everything. What made this especially heinous was that if a Debian menu had been included in KDE, I could have made a launcher out of it. At this stage, though, I don't believe that's enough. Debian should follow the lead of every other major distro and offer the exact same menu layout throughout. All you need is for graphical packages to install an information file in, e.g. /etc/debmenus, and in the post-install stage run a script which creates from it th e necessary menu entries in all the window managers and environments.

I've got most of the sound and KDE stuff off my chest, though frankly its deeply disappointing. It's the first time I've experienced functional bugs in any KDE version, and I started with 0.99. The only other time I've seen a major bug was a cosmetic issue with KDE 2.1 (?) in SuSE 7.3 which caused vertical stripes to appear on widget background s.

Again I've dealt with the appalling foul up of Euro-support. The support packages should have been installed by default when I selected en_IE@euro. The AltGr-4 keymap should have been set up. As far as I'm concerned these are functional bugs.

The PPP tools could definitely have been set up better. The default setting is only an invitation to newbies to use root for web-browsing. They could be set up using sudo, or else set up them with rwsr-sr-- permissions and root.pppusers ownership. That way, at the user creation screen you could ask if people should have permission to connect to the net, and make them members of the pppaccess group if permission was granted.

GTK, and consequently Mozilla, looked atrocious due to the oversized fonts (look at Windows, MacOS, BeOS, other Linux distros - they all have fonts a round 11px), and changing the default font in GTK is a bit of a struggle for newbies (how obvious is Theme Selector after all). I changed it to Helvetica at 12, and now things look okay.

The fact is, I'm going to have to invest a considerable amount of time just to get things to the same level that Mandrake and RedHat give straight out of the default install. This is not something that will attract new people. Oth erwise the system seems reasonable. I'll have to wait a while before I can make any pronouncements with regard to stability. Anecdotal evidence is extremely positive, but my initial experience hasn't matched. I was a little disappointed with the way files were arranged. I had hoped Debian would lead the world away from RedHat's madness and stick KDE and Gnome in their own subdirectories, e.g. /usr/kde2 -> /usr/kde-2.2.2 and /usr/gnome1 -> /usr/gnome-1.4.1. The fact is, that given what I've had, and will probably get when RedHat 8.0 inevitably starts going around the magazines, it's hard to be upbeat about the Debian desktop.

Conclusions
I'm sure you're aware that this isn't going to be glowing. Debian's installer is several years out of date, and needs a serious overhaul. It's not fit for commercial consumption, and is only good enough for established Debian users and poor wannabe PFYs like myself. This is not a sustainable situation. Apt-get is good, but RPM has caught up with it for the most part thanks to apt-rpm and urpmi. I'll take everyone's word for it and say that Debian is, for the most part, stable. I like the fact that the packagers are willing to hold back and patch existing stable software to get a decent system, and not one that seems to be in permanent beta. This is why I went for it in the first place.

But people who chose Debian aren't rewarded. Installation and post-install configuration is a bit of a nightmare. Debian should organise people to collect code from the Webmin, Linuxconf and Mandrake configuration programs and create Debian's own configuration framework. At this stage of Linux development it's compulsory, even RedHat has finally copped on to this. Indeed, I would recommend following RedHat in several arenas. I believe Bluecurve is free, Debian should package it - it gives everything a nice polished look. People can then change things if they want to. Having worked in MIS a bit, I know that people will always find a way to muck about with display settings, even if word-processors give them palpitations.

I think peopl e should get together and form a DebianDesktop group, committed to creating a package which will install several different themes, configurations and menus. People can be asked near the end of the install if they would like their desktop customised - if they answer yes, this package could be installed. Similarly work should be done on intelligent installers and hardware auto-detection (though the latter is obviously going to be especially difficult for a multi-platform system). The priority should be the simple installer though, hardware detection can wait.

The inspiration for this article was an article I saw on this site a while back bemoaning Debian's loss of mindshare, attributing it in part due to the lack of attention in the media. Most of the pertinent points were made in the article and accompanying comments. An open-source distribution needs mindshare to survive, but the media won't cover distros which don't have the latest whiz-bang desktop software. If Debian formally released a distribution based on the Test tree compiled with GCC 3.2 for 686mmx, its marketshare would explode. Just look at Gentoo, a hideous installation process, but a system equivalent to a Honda Civic with added spoiler, exhausts, alloy wheels and, of course, go-fast stripes. In other words, something for the lads to show off.

Such a system would have the benefit of bringing a lot more bug-reports into the system, g iving a better stable distro. Mandrake are sucking a lot of the talent Debian needs through cooker. They've openly thought about making the distribution packaging process totally open and building a value-added distro around it like Progeny. If this were to happen it would place Debian into a very tough place.

The new Debian needs to blow people away. It needs to be Granny-proof. It needs an installer that people can bluff their way through, with an attractive, well configured desktop on the other side. Debian maintainers should check out the competition now and again, to see where they can improve. Because if they don't, Debian will lose developers, and become less and less of a force in the Linux world

You're right, "autism level" is not discrete. A diagnosis based on a large variety of symptoms and variables, not just those listed in the DSM for that particular disorder.

BTW, the ":-)" in my post is an emoticon : "An ASCII glyph used to indicate an emotional state". In this case the emotional state was humor.

I guess he should notify everybody first, or the system could set a simple auto-reply in the order of "This still works, but only for a little more while.". A simple change would be to change the ".su" part of the email to ".(whatever country your father currently resides in now)", assuming his e-mail buddies know in which country he actually lives. But no one has money to do that I guess.

So he got the email address two years after the USSR fell? Makes sense, I doubt the Communist Party would have wanted its citizens to be able to talk using an uncensored western communication medium. I wonder what was the rationale behind creating a .su domain at all?

And why .su? .ussr or .cccp would be a lot damn cooler, IMO. :)

One time I mounted a floppy disk that had some problems under Linux. I know it had problems because when mount mounted it, it said it was mounting it read-only because there were errors detected in the filesystem. I copied some data off it, then unmounted it. When I unmounted it, I got a message that said, "Attempting to write to a read-only filesystem. uhh..."
I don't know why something was attempting to write to the disk on umount.

I think one of my favorite messages was the message that, that old Apple ][e printing program (what the hell was it called? Printshop? I can't remember) would display when it was calculating what commands to send to the printer next. It would flash the word `THINKING' on the screen in a huge font and alternate the "colors". (It wasn't actually a color monitor--it was monochrome but it had the capacity for bright text as well.)

And let's not forget the error message Haikus (sp?) that people were playing with before. I really liked some of those. I think that some real programs (for programmers, not regular users) should try error messages delivered as Haiku. It would be neat. The point would be to have more than one Haiku for the same thing; then always follow that with the no-nonsense error message down below the Haiku. The Haiku would be like a little treat to offset all the compiler errors. Sort of like a quote of the day.

Now back to Linux. Quite a long time ago, I foolishly tried to run X Windows on a 486 with only 16MB. It didn't exactly crash, it just kept the hard drive going non-stop. Even when I didn't move the mouse for over a minute. When I would go to shutdown X Windows, over in the console where I had typed startx&, there'd be a status message like this:
Sending server the TERM signal,
waiting for server to shutdown...
Sometimes, that's all there'd be. But because this machine had only 16MB and the hard drive was thrasing so much from running X, everything was drastically slowed down. Sometimes, after about 30 sec-1 min, I'd see:
Server too slow to shutdown
Sending server the KILL signal,
waiting for server to die...

In plain vanilla MS-DOS, if DOS fails to read from a floppy disk in a certain special way, what can happen is this:
C:\>dir a:
[the disk churns and time passes...]
[some more time passes]
Volume in drive A is unlabled
[again time passes]
[still more time passes]
Fail on INT 24

C:\>

If you're stupid or unlucky enough to be in a situation where you have to make a batch file for COMMAND.COM, then all the error messages are cryptic. Because, as the script executes, there's no indication of what line caused the error message (unless you do an echo on and there are certain situations where that won't help. All you see is stuff like:
Bad command or filename
Syntax error
File not found
all the way down the screen.

Oh yeah, and then there's INTERCAL, where the compiler error messages are actually intended to be not understandable.

Actually, the message ?Syntax error was really cryptic the first time I saw it, since I had no idea what the hell `s-Y-n-tax' was. (I was 12.) In general, every (status or error) message a computer gives you is cryptic without context. Ironic since computers can't really understand context, eh?

I liked this error message Amigas used to give: "GURU MEDITATION #XXXXXXXX.YYYYYYYY". For the curious, it's explained better here.