[LWN Logo]

 Main page
 Linux in the news
 Back page
All in one big page

See also: last week's Back page page.

Linux links of the week

Script-fu.org is a new site dedicated to the scripting language used by the GIMP. Here's all the stuff you need to get started doing magic things with our favorite image editor.

Linuxapps.com has been through a major upgrade. It has a new look and a new interface, worth a look.

January 28, 1999



Letters to the editor

Letters to the editor should be sent to editor@lwn.net. Preference will be given to letters which are short, to the point, and well written. If you want your email address "anti-spammed" in some way please be sure to let us know. We do not have a policy against anonymous letters, but we will be reluctant to include them.
Date: Thu, 21 Jan 1999 21:47:49 -0600
To: editor@lwn.net
From: James Thompson <jet@avalon.net>
Subject: Mac OS X Web pages...

Hi Folks,

Have you sauntered over to the Apple Website and looked at Mac OS X
screenshots lately?

The Mac OS makes an impressive front end for Unix.

If it runs as good as it looks...

If Linux developers could come up with a GUI that makes Unix easy to use,
my complaints would melt.  Apple has gone a long way towards that.

Mac OS X may replace my current Linux DNS server (sniff) for those chores
because of it's ease of use.
James Thompson
NRN Consulting
Iowa City IA
From: Dick Hein <rwh@lmi.net>
To: editor@lwn.net
Subject: Re: Discussion on FUD
Date: Mon, 25 Jan 1999 00:43:39 -0800

As to what it is, I think this piece -


...pretty well sums it up.

Concerning use of the term to describe criticism of Linux - I don't see
anything wrong with using it when appropriate.  FUD =is= practiced,
after all, and is no stranger to the likes of the Redmond crew.

Dick Hein / rwh@lmi.net / Mountain View, California.

Date: Sun, 24 Jan 1999 13:54:42 -0000 (GMT)
From: Dirk Koopman <djk@tobit.co.uk>
To: editor@lwn.net
Subject: It's not all good...

For my sins I have been running Linux as a programming platform since 0.99.x
days and whilst I have seen many things improved, there are others which have

Recently I was asked to prepare a machine for someone who wanted very cheap
internet access and home word processing, he was a newby and happy to try

Naturally I thought of Linux, so I installed RedHat 5.2 and Star Office onto
a spare 32Mb Cyrix DX2/66 clone together with X and isdn.

To cut a long and painful story short, the system was unusable. Star Office
took forever to load and then, in a VERY short time of use, proceeded to use
all 48Mb of swap space (as well as the available RAM) and ran like a 
woodlouse (i.e. somewhat slower than a dog).

Loading Netscape Communicator wasn't much better, at least it left a
(small) bit of RAM, but then started to eat swap as you surfed.

What makes this worse is that I remember running a similar configuration
(but with an Intel DX2/66) in early 1.0 days with X and Netscape and having
program loading competitions with sceptical M$ fans and beating them hands

In disgust, I loaded Win95 and IE 3.2 and although it can't be said to be a
fast machine (in any way) it is at least usable, programs (even Office)
load in a tolerable amount of time and the system is stable enough for casual
home use.

Is this warning for us all? try a loading competition now with Netscape on
Linux and IE (or even Netscape) on identical modern specification machines,
you may be surprised by the results!


Dirk Koopman
Dirk-Jan Koopman, Tobit Computer Co Ltd 
At the source of every error which is blamed on the computer you will find
at least two human errors, including the error of blaming it on the computer.

Date: Thu, 21 Jan 1999 12:55:17 +0000
From: David Jao <djao@MATH.HARVARD.EDU>
To: editor@lwn.net
Subject: Misleading report on Sarah Flannery's encryption algorithm

ZDNet, BBC News, Slashdot, and now LWN have all published reports on
Sarah Flannery's new encryption algorithm which are in some way
inaccurate. The discovery of a new direction for exploration, and the
coming of age of a new researcher are very deserving of praise, but the
possible immediate practical impact of this particular algorithm has
been way overblown.

Details on the discovery are sketchy at this point. I am basing this
letter on the few details of the algorithm that I have been able to pick
up from Slashdot posts and the various news sites, so take it with a
grain of salt. That said, it seems that no news site has an accurate
conception of the proper significance of Sarah's discovery.

Sarah has discovered a matrix algorithm which is provably equivalent to
RSA in security, uses about 20 times less CPU, and 8 times more memory.
This discovery is of great interest to researchers and theorists, but
has little practical impact on user level software in the forseeable
future. No production program being distributed today uses RSA for bulk
encryption of data. Programs like PGP use RSA to encrypt a small session
key, which is in turn used with some other cipher (not RSA) for bulk
encryption. The practical value of fast RSA encryption is questionable,
because RSA is simply not used to encrypt large quantities of data in
today's world.

Furthermore, there are already enough free (i.e., unpatented)
cryptographic algorithms available today that the addition of one more
is not too exciting. Today we already have Diffie Hellman, El Gamal, DSA
(for signing), Blowfish, and CAST. GNU Privacy Guard has progressed very
well using these existing free algorithms. Even RSA itself will fall out
of US patent protection in less than two years. The value of one more
free algorithm, while positive, is not, as you say, "incalculable."

I know you guys are not cryptographers, so I forgive you (and other news
sites) for their mistakes. But now you know the truth. By itself, this
particular discovery of Sarah's will not make a difference to users in
the next 10 years. We all should be most encouraged by the prospect of
future research from Sarah and others (possibly based on Sarah's current
work) that will dwarf this recent achivement in practical benefit.

Date: Thu, 21 Jan 1999 11:31:18 -0800
From: Jay Jakosky <jakosky@usc.edu>
To: editor@lwn.net
Subject: French vs. US cryptographic restriction.

"The biggest news of this week in security has been the about-face
of the French government policy on encryption. Part 1 and Part
2 of the government documents in French describe a "project"
(essentially, a proposed new law) which promises, eventually,
complete freedom of the use of cryptography within France. In the
meantime, until the law is enacted, the maximum allowable keysize
for cryptography has been increased from 40 bits to 128 bits,
certainly trumping the U.S. Government's recent increase in
allowable keysize to 56 bits. Exportation of cryptography is still
controlled by virtue of existing agreements with other countries."

You are neither the first nor the last to refer to 56 bits as the limit
of the strength of  U.S. domestic cryptography. Strength of cryptography
EXPORTED from the U.S. is now limited to 56 bits. France in no way
trumped the U.S. since they still have a limit on domestic cryptographic
strength. I'm sure you know this but your statement (quaoted above)
portrays a different message about the comparative freedom of France and
the U.S.



Eklektix, Inc. Linux powered! Copyright © 1999 Eklektix, Inc., all rights reserved
Linux ® is a registered trademark of Linus Torvalds