See also: last week's Back page page.

Linux links of the week

Python hackers may want to have a look at the Vaults of Parnassus, a comprehensive, hierarchical collection of available Python modules.

Section Editor: Jon Corbet

Letters to the editor

Date: Mon, 24 Apr 2000 13:26:57 -0400
From: "John F. Gibson" <gibson@mae.cornell.edu>
To: letters@lwn.net
Subject: DOJ vs MS, law vs libertarianism

Eric Raymond writes "Economic competition is ... a robust and
ubiquitous phenomenon that flourishes whenever human beings need 
to solve scarcity problems and are not forcibly prevented from 
trading with each other to do it."

This is quite likely true for emerging markets among individuals
with comparable economic clout. However, the situation is more 
complex in developed economies, with economies of scale and large 
disparities between the players. In such an environment, a player
with sufficient resources can effectively prevent trade among others,
through threats, coercion, dumping, buying and squelching potential 
competitors, etc. Monopolies are especially capable of such tactics. 
This is why we have anti-trust law. 

Over the last ten years, Microsoft has effectively prevented trade 
in operating systems by secret pricing structures and threats of
exclusion to manufacturers, in productivity software by closed,
rapidly changing file formats and coerced bundling of Office, in 
browsers by dumping Internet Explorer and corrupting Internet 
protocols. These actions have manifestly harmed consumers, the
software industry, and ultimately the U.S. economy. 

Mr. Raymond writes, "There are any number of counterexamples to the 
... claim that government-made law is essential to economic competition. 
Customary law maintained by the self-interest of economic actors is
quite sufficient." 

Perhaps examples prove that law is not necessary in every case,
but they don't prove it's unnecessary in all. If so, how does one 
explain away the harm Microsoft has already done? We're talking 
about billions of dollars, warped evolution in the software 
industry, the paths of thousands of people's lives, and ninety 
percent of the country subjected, for the past twenty years, to 
an operating system that crashes daily! You can't claim that 
Adam Smith's invisible hand will ultimately prevail and right 
these wrongs retroactively. 

Clearly what is needed is more effective law and better, more timely
enforcement.

Lastly, the First Amendment is in fact a regulation on the behavior
of Congress. It prevents Congress from making laws that restrict certain
freedoms of individuals, the press, etc. The founders recognized the
disparity of power between the majority, acting through Congress, and
minorities. They enacted the First Amendment to ensure fair
competition between unequal bodies. Kind of like antitrust law for
speech.    

John

--
John F. Gibson     gibson@mae.cornell.edu   
Mechanical and Aerospace Engineering, Computational Fluid Dynamics Lab
288 Upson Hall, Cornell University, Ithaca, New York, 14853-7501
Tel:    (607) 255 0360        Fax:    (607) 255 1222

From: ghaverla@freenet.edmonton.ab.ca
Date: Thu, 20 Apr 2000 09:29:44 -0600 (MDT)
To: letters@lwn.net
Subject: Printing and Re: WordPerfect "review"

Hi!

  I haven't run WordPerfect (or any other word processor),
but where are people getting this misconception that printing
is screwed up on Linux/*BSD/...???

  Printing is simple on any UNIX or UNIX-like machine, you
send the print job to the lpr (or lp) program, which sends
it to the print spooler (lpd), which ....  The program 
requesting the printing doesn't have to know a darn thing
about whether the printer is directly connected to "this"
computer, is directly connected to some other computer,
is a network printer, or is something acting like a printer.

  Word processors like to know what kind of device they are 
printing to, but this isn't a problem.  They only have to
look in the printcap file to find what devices are available.
Maybe the information in the printcap needs to be extended
a bit for word processing applications, or augmented, by
another config file.  But this certainly doesn't require the
word processor people to run around writing drivers or any
such nonsense.

  LPRng is actively in development.  Of late, we have someone
looking for a little tuning advice on having a bank of modems
attached to the system (somehow) to be used for outgoing
faxes, with each modem getting its own customized cover
sheet (phone number of the modem).  A little while ago we
had someone write in about using LPRng as the spooler for
a system where some kind of "ink" was placed on cookies
prior to baking.

  Just my $0.02.

Gord

Matter Realisations     http://www.materialisations.com/
Gordon Haverland, B.Sc. M.Eng. President
101  9504 182 St. NW    Edmonton, AB, CA  T5T 3A7
780/481-8019            ghaverla @ freenet.edmonton.ab.ca

Date: Sun, 23 Apr 2000 06:38:02 -0700
From: Steve Powell <stevenrpowell@sprintmail.com>
To: letters@lwn.net
Subject: WP Office Works Great for Me

I just installed WordPerfect's office suite, and while it still isn't
doing everything I want, it isn't nearly as fouled up on my system as it
seems to be on yours. My menus work fine -- everything works fine. The
only problem I have is I can't save to a .pdf yet. I haven't done
exhaustive tests but I've opened Paradox and looked around a bit in the
help system and what not, and it looks good. QuattroPro opens and can
perform routine tasks. I've used WordPerfect the most, and it was able
to create a document, run spell and grammar checks, etc.

Again, I haven't performed an exhaustive test of anything, but it
certainly isn't as goofed up as your experience suggests. Sounds like
something else is wrong. I installed to Corel Linux, and maybe that
makes a difference. Not that that's any excuse, but in any case I don't
think the package is nearly as far from functional as you found.

--
********************************************************************
Steve Powell
StevenRPowell@SprintMail.com
921 Coast Blvd South Apt 1
La Jolla, CA 92037-4150
Home: (858) 551-2021
Work: (858) 505-3460
********************************************************************

From: "Allan Stokes" <allan@stokes.ca>
To: <lwn@lwn.net>
Subject: soft tissue
Date: Fri, 21 Apr 2000 04:19:29 -0700

Hi Liz,

Story in /. tonight:
<<
"According to this AP story, the remains of a 66 million-year-old dinosaur
suggest that the extinct creatures were warmblooded - not coldblooded as
once believed - and capable of the swift and sustained motion typical of
modern birds and mammals. A whole site dedicated to the discovery of this
specimen is here."
>>

Your comments about encryption:
<<
Evidently the QNX folks decided to roll their own, closed-source, unreviewed
encryption, with the usual results.
>>

Do you see the parallel between your logic and the logic which decided that
dinosaurs were cold blooded?  It's the looking under the street lamp because
the light is better fallacy.  Starting out with only the bones they
concluded that dinosaurs were cold blooded because nothing was telling them
otherwise.  It's only what you don't have that is capable of arguing against
you.

It's exactly the same when look at the bone yard of proprietary encryption
algorithms broken and ignore the "soft tissue" of proprietary encryption
which hasn't been broken.  Or as most people assume "hasn't been broken
yet".  Which is exactly what the dinosaur people assumed about the kinds of
fossil fragments they had not yet found.

Perhaps someday the mathematics of "provable security" will be invented and
they will look back at some of the proprietary work done today and discover
that some of it was actually warm blooded after all.

I see this over and over again.  Things which can be measured compared
against things which can't be measured.  The light is always better under
the street lamp.

Even the experts fall into this trap.  Data directed encryption primitives
(which Ron Rivest plays around with) are often rejected because the analytic
tools available are not capable of finding defects.  Yes, it's true.  We
reject this approach because we have not yet invented the appropriate
mathematical tools for demonstrating that it doesn't work.  If we can't even
shoot at it we trust it even less.

Encryption has become a bravado culture.  You are told to wander out into
the mine field of techniques which mostly fail, and you are only respected
if you emerge safe after being shot at by a hundred different people.  This
really says a lot more about the inadequacy of our mathematical framework
than it does about whether a warm blooded individual who wanders away from
the mine field is capable (or not) of getting lucky with a proprietary
method.

Of course, anyone dumb enough to trust someone who spends too much time
alone in a dark room deserves what they get.  But that doesn't mean they
were wrong.  People spend too much time forming opinions about what is
technically possible (we don't know) and then end up misplacing the emphasis
which belongs entirely on the social issue of what kind of development
processes we choose to trust.

Allan