UNIX Systems Diagram

The UNIX Systems Diagram on page 10 of LJ Issue #69 is missing a UNIX flavor that can be expected to influence the future of Linux. SGI's IRIX is not included in the diagram, although on page 8 it is reported to run more web sites than AIX or HP-UX, which are included.

What is significant about IRIX is SGI's plan to phase out IRIX over the next few years and move completely to Linux on their hardware. You can bet that they will be contributing more components of IRIX to the Open Source community. They have already started contributing some IRIX code, starting with their XFS high-performance journaling file system. I'm waiting for their 4Dwm to be ported to Linux; their Desks Overview has a better multi-window management tool than others I've tried, such as fvwm2, KDE, or AfterStep.

—Mark Plimley, markp@blueneptune.com

Re: PNG Book review

Thus spoke Ian Clarke, I.Clarke@strs.co.uk

I was somewhat concerned to notice that a link is provided above your online review of this book to that books page in Amazon's online store. This link seems to inform Amazon that the referral was provided from Linux Journal. I know that Amazon do reward people who encourage others to buy books from them, and I must say this makes me somewhat suspicious of your glowing review of this book, and also calls into question LJ's journalistic integrity.

I don't have anything to do with how LJ does their layout, either in the printed magazine or online. I also don't get anything from Amazon with respect from the link you mention. I wrote the review for a stock price with no input from LJ on what that review should contain. I wrote the review without influence as to content from either Amazon or LJ. It's just a damned good book.

Does Linux Journal profit in any way from people buying this book through Amazon having followed the link from your review? If so, how do you justify disguising what is essentially an advertisement as an impartial review?

I don't work for LJ. I'm a freelance writer. They ask if I can do a book review. If I say yes, they send me a copy (unless I already have it, which I'd already purchased a copy of this particular book *before* they asked me to do a review of it). They don't tell me what the review should say—it's completely up to me. And it's always been that way. Even before on-line partnerships with resellers like Amazon came along.

Additionally, you assume there is a profit motive in making that link. No one is getting rich off of “associates” programs with Amazon or other book sellers. It's a convenience thing for readers more than anything else. I doubt LJ would notice the drop in income if they removed the link.

It's not a disguise. It's a good book. That's why the review is so glowing. If you'd read any other technical books on Linux you'd know this book makes them all look like children had written them. Greg is a very good writer.

Further, given Amazon's recent exploitation of the American Patent system, something that is anathema to anyone who supports the principals of Open Source software, LJ's relationship with them appears even less well thought-out.

I don't tell LJ who they should associate with, and they don't tell me who I should. I don't like patents much, but I'm not boycotting Amazon because of patents. I simply don't buy books from them for the same reason I don't buy computer equipment from Wal-Mart.

If you disagree with the review (i.e. you think it's not a good book), that's one thing. But my involvement with how it's displayed or associated with advertising within LJ is non-existent. I suggest you take this up with Linux Journal directly.

—Michael J. Hammel, mjhammel@graphics-muse.org

LJ does receive some sort of pittance if someone buys the book clicking through us. But we never tell our reviewers what to say. The link is there on both good and bad reviews.

Linux GUIs

I just read the article on GNOME and Its Future. Some comments:

1) The OS interface must be usable without a mouse. I have not found any GUI interface for Linux that meets this criteria. Even Microsoft has keyboard equivalents in its various Window OSs. People talk about right/left clicks like a keyboard never existed. This may not seem important, but work can be more efficient if the mouse is used sparingly.

2) The abhorrent focus on the Internet has obscured the vision of product developers. Many people do not have Internet access and never will unless it is totally free. Internet-related modules should not be loaded by default nor should they be an integral piece of the OS. For example, Red hat's installation loads “chat room”-related modules/packages by default. Why? Is everyone supposed to use “chat rooms”? I do not think so.

3) Personally, I do not like the Internet-style interface. I do not have Internet access at home and will not until I can get DSL for Linux at an affordable price. I do think the Internet can be a useful part of people's lives, but it needs to alter its direction somewhat.

4) I use a computer for computing not entertainment. I use Linux for computing not entertainment. If I did not care about computing I would use a Windows product. Linux development seems to be emulating Microsoft rather than continuing to build/improve a real OS.

—Burk, Galen G CHPPM-Wash DC, Galen.Burk@NA.AMEDD.ARMY.MIL

Meaningless Survey

As a statistician I am irritated by those surveys you are publishing since several issues and specially by your Readers' choice awards. The method of the survey is completely flawed.

What you are using is accounting the votes of people who go to your site. This has a name in statistics “straw ballots”. The problem with them is that they are very sensitive to the problem of vocal minorities spreading the word and massively voting to the straw ballots. In the 1932 US presidential elections they predicted a crushing victory of the republican candidate while the first Gallup polls predicted Roosevelt's victory and got very close of the percentage of votes he got. Since then no serious statistician uses straw ballots or gives a dime for their results.

The Gallup method consists in _you_ selecting a fair sized sample of your readers and asking them to answer a questionnaire. While militant types will have a higher answer rate and thus the results will have a certain bias towards the software they support this is far less than the bias in straw ballots.

So, please if you can't use Gallup polls at least include a BIG disclaimer warning the reader about the inaccuracy of the results.

—Jean Francois Martinez, jfm2@club-internet.fr

Actually, we don't use a disclaimer on the Reader's Choice awards, because we feel our readers have the intelligence to figure out the bias of straw polls themselves. Doing the survey on the web is cheaper, easier and more convenient for both our readers and us than doing a mail-out survey.

IBM Ad in your Jan 2000 Issue.

I just received my Jan 2000 issue of LJ. I have just one question. Why is Tux made to look like a Chicken in the IBM 3D Ad? The fact that he looks like a chicken makes the “World Domination” theme confusing. Are they talking about “World Domination” for Linux or KFC?

—Jason Lawrence, jason@19.com

Stock Info!

First let me say LJ is GREAT!

I have been following the different Linux-based stocks and have been reading the articles on the Linux insiders and startups. I thought; it sure would be nice if somehow LJ could cover one or two companies from the perspective stock holders point of view. It would also be good for investors and the companies advertising in LJ if there stock symbol was listed in the advertisers address information. For those who are looking for investment capital. It seems that I remember a brewery in New York where the owner went public on his own and did very well. Maybe a story on something like that would help small Linux programmers to get a foot up with there work.

Just some thoughts,

—Fred Walker Jr., fastfred@wil.net

re Feathers McGraw


How dare you make Feathers McGraw an honorary member of the Linux community (“About the Cover” LJ 1/2000) on the basis of his goal of World Domination!

Not only do I not recall his goal as being one of world domination ( though it may well have been), but he was an out-and-out crook: in Nick Parks' “Wallace & Gromit - The Wrong Trousers”, Feathers McGraw stole Wallace's robotic trousers (with Gromit in them) and then used them to help him steal a big diamond!

Feathers McGraw is a thief, an verifiable criminal (a master criminal according to http://www.aardman.com/wallaceandgromit/homesweethome/feathers/index.shtml), and you want to bestow upon him the great honour of an honorary member of the Linux community?! I am both shocked and horrified!

Feathers McGraw should be seen for what the really is: an attempt by subversive evil forces, such as those found in a certain part of Redmond, Washington, U.S.A., to infiltrate and corrupt the near untouchable ranks of our great Linuxdom. Let us, therefore, close ranks to keep such potential infiltrators and perpetrators of great mischief out!

Enjoy 2000! Graeme Nelson

P.S. It is good to see the even the eminent Phil Hughes makes typos! In his “Watch Out For The Snakes” letter (LJ 1/2000), he calls Python “easy-to-debut”: as things usually only debut once, I guess his finger slipped upward and hit the <t> instead of the <g>. You see, just 2 pages after awarding Feathers McGraw honorary Linux community membership, his is already working his heinous schemes of decay, keeping Phil from declaring Python to “easy-to-debug” and thereby encouraging untold millions away from M$ products. Instead Phil declares it to be “easy-to-debut” and so causes it to be viewed with suspicion, or even worse, causing it to be viewed as mediocre, and therefore equal to or worse than M$ products.

—Graeme Nelson, graeme@cheerful.com

jan 2000 cover

Good mag; I'm well pleased with my subscription but... er.... your cover for Jan. Umm.... you do realize that's the evil penguin from Wallace and Grommit! Seriously though, thanks to all for an excellent mag. Keep up the good work.


—K Clethero, k.clethero@taranaki.ac.nz

Readers choice awards

Maybe for those of you with Pentium III's mpg123 is THE audio application, but for those of us with more modest; computing power amp ( and xamp) work much better. Also I think somewhere mention should be made of the 2 things that make text Linux preferable to GUI namely gpm and command completion

—mraivio, mraivio@oln.com

Linux Journal Letter to Editor: “World Domination in 1900”

So, Mr. Smartipants, unless you are playing a “Ha ha, only serious!” joke on us, I think you have some explaining to do. Your January 2000 issue has “January 1900” footers on some pages (with red-line correction). Embarrassing, after joking “Y2K compliant & certified” on the December issue cover.


—Walter Cooke, wcooke@paragon.bm

The December cover and the follow-up date in January were jokes. We wanted to poke a little fun at all the Y2K uproar and at ourselves.

Why the Y2K explosion was a dud.

On an old PC with Linux, I entered the following New Year command:

% clear ; date ;
and what appeared on my screen was the following:
% 01-01-00
% ? ... ?
% Mon Jan 1 9:49:22 EST 1900
% No electronic computers existed in the year 1900.
% I compute, hence I exist.
% Sat Jan 1 9:49:23 EST 2000
% Descartes ... eat your heart out.
All this in less than a second of computing time.

Allan Widom

—Allan Widom, widom@acausal.physics.neu.edu

LinuxPPC vs PowerPC Linux]

In the Forum, page 112, February 2000, Stephane Morvan opens his presentation of PowerPC Linux “flavors” by stating “LinuxPPC, along with MKLinux and NetBSD, are the three flavors of Linux that can be found for Macintosh computers.”

There is a major flaw in this statement that has been overlooked by the author of this article, Linux Journal, and the media in general over the past year. “LinuxPPC” was once used in reference to the “Linux for PowerPC” development community. However, with the introduction of the for-profit company LinuxPPC, Inc., a general confusion and lack of separation between the development community and the commercial product resulted. This is compounded by the fact that LinuxPPC, Inc. has failed to release the www.linuxppc.org domain to the PowerPC Linux development community and continues to link to their specific commercial products.

In an effort to regain autonomy, the once “LinuxPPC” development community has migrated to the “PowerPC Linux” development community. This title needs to be recognized and supported in order for every commercial venture offering Linux for PowerPC to be truly competitive, offer quality products to Linux consumers, and most importantly, make certain the development efforts of the community are readily recognized as independent of the commercial ventures.

As Caldera, SuSE, and Turbo Linux have fought an uphill battle to gain recognition from the media who continue to focus, and sometimes use the term “Red Hat” in reference to Linux in general (or even the reverse where “Linux” refers to “Red Hat”), we simply ask that proper recognition be given for all distributions and development efforts.

—Kai Staats, kstaats@terraplex.com

We try to catch blanket statements made by authors but sometimes we miss. Sorry about that. In our April Distribution Watch, we will discuss all the various flavors of Linux of the Power PC. Thanks for writing.

Desktops of the Future

I was pleased to see “as we may think” mentioned by Mr. Salus. However Memex, the hypothetical machine described by Dr Bush, was not a general purpose computer. It was a database appliance which would literally search through a file cabinet, locating records by filenames, just as a human file clerk would do. What's interesting about Dr. Bush's paper isn't the machine, which he contrived as an example of what might be possible with 40's ear technology, it's the system of linking documents together with key words and targets thus making it possible to search the database without looking at every file. Now that we've all been exposed to the web for a few years clicking a link to find related documents doesn't seem unusual but in 1945 it was a fairly new and exiting idea.

“As we may think” has been donated to project Gutenberg, and the entire text of the article is displayed on many web sites. It's an important part of our history, I recommend it to anyone who is interested in the how & why of science and technology. http://www.isg.sfu.ca/~duchier/misc/vbush/vbush.shtml

—N.Masters, enormas@gj.net

Grief with LinuxCad

I don't know if this is the right forum for this, but hopefully it will save other people some grief.

Software Forge Inc., which advertises LinuxCad in the LJ, falsely states the capability of LinuxCad. On a direct question through email whether LinuxCad supports STL output format, Software Forge Inc. lied and said “yes”. Further, software Forge's web page falsely states that LinuxCad contains equivalent 3-D capability as AutoCad, which it does not. On top of that, many of its existing functions don't work. After many unanswered email and phone calls (request left on a answer machine) I got the the email response “Wait until the next release.”

I have requested a refund.

I have purchased quite a few commercial software for Linux (Applix, Star, VMware, Motif ...) which have all been top-notch. This is the first product that has come close to being a scam.

On another note; I have being using Linux since 1992 with my first copy a TAMU distribution. The LJ is a good magazine and does well on presenting to a wide, diverse audience.

Fill free to edit this and do as you wish with it.


Dale M. Snider, dsnider@nmia.com