See also: last week's Back page page.

Linux links of the week

The Giant Java tree is a project to create a comprehensive set of Java classes, all implemented with open source code.

Letters to the editor

From: cbbrowne@godel.brownes.org
To: editor@lwn.net
Subject: Free Software (Gift) Exchange Registry - FSEX
Date: Thu, 17 Dec 1998 10:59:14 -0600

I have been known to write essays on occasion :-).

The latest is more directly a "call to action" than most of the others.
See the URL <http://www.ntlug.org/~cbbrowne/fssp.html>

The premise is that I would like to establish a "registry" where
reasonably authoritative lists of addresses of developers of free
software may be collected.  

- Individuals would be encouraged to send gifts to people on the list.  

- By collecting (voluntarily contributed) information on what gifts had
been sent, this allows people to "optimize" their preferences towards
where contributions may "need" to go.  Thus, they have opportunity to
pick developers themselves based on some perception of "need."

[This building of a distributed self-optimizing economic system
somewhat parallels the way Linux development works...  That can't be
a bad thing!]

- This approach *avoids* the bureaucracy entailed by the formation of
a formal charitable organization.  I would expect this to discourage 
corporate "gifts," but avoids a whole host of complexity by its
informality.

- From a tax perspective, gifts may not be tax deductible to the giver,
but by the same token, would not be taxable in the hands of the one who
receives it, so that the overall situation is a "wash."

As the Christmas season arrives, a thought to pass on:

Why not locate a nice Christmas card with a penguin on it (has
everyone noticed that penguins are "in" this year?), drop $20 into
it, and send it to some developer that has built software you found
useful?  (And remember that if Linus' wife has to fight through bags
of mail, she knows martial arts, and may have complaint about this!
Rather better to distribute any wealth...)
--
cbbrowne@hex.net  - What have you contributed towards Linux today?
North Texas Linux Users Group <http://www.ntlug.org/~cbbrowne/fssp.html>

From: Matthew Benjamin <MBenjamin@comshare.com>
To: "'nicholas_petreley@infoworld.com'" <nicholas_petreley@infoworld.com>
Subject: Pretreley Linux Fandom on the Wane?
Date: Mon, 21 Dec 1998 13:21:06 -0500


I found your "rat out of an aqueduct" remarks disturbing.  Seeing as
you have been active in promoting Linux of late, are you now trying to
be among the first to sling mud at Linux?

Quite frankly, I don't think those ISVs supporting Linux are
significantly motivated by the MS anti-trust action.  (The case may be
different among PC vendors, I don't know.)

At my company, there is significant support among developers to do
Linux--because "Linux is cool."  Strategic marketing is potentially
interested in _anything_ that customers say they want, and many are
saying they are interested in Linux.  This has the power to change
opinions at companies large and small.

I, for one, had enough "Windows NT is the future" sloganeering years
ago, and I question its relevance now.  Microsoft, of course, can say
whatever they want.  However, they will find that consumers take a
quite different attitude towards such posturing than they did
pre-Linux emergence, and, indeed, pre-anti-trust.  Microsoft is big,
but they are neither invincible nor the permanent direction of
technical evolution.  As consumers become more sophisticated, in fact,
Microsoft looks less like a technology company, and more like KMart.
That is inevitable, and will not assist Microsoft in the least.

Meanwhile, the Linux and OSS communities have evolved their own (very
effective) mechanisms for generating publicity, and they are so
informative and effective that I believe no trade press FUD campaign
can have the choice-damaging effects that the all-commercial trade
press allowed previous MS campaigns to exert.  This, of course,
resounds to the benefit of consumers.

Linux, for its part, got where it is by being technically viable
_before_ a single one of the usual suspects had anything to do with
it.  Larry Ellison and Nicholas Pretreley can abandon Linux whenever
they like--but since they neither own nor develop Linux, I submit that
this will only leave a desirable market to the new generation of ISVs
and IHVs who have made Linux their business over the past 4 years.  In
accounts where I've deployed Linux commercially, there is no intention
to replace it with anything.  It just works too well, is too
hassle-free.  There is talk about migrating other functions to Linux.

Hard as it may be to believe, Linux, the free UNIX, is winning on
quality, not cost.

Just the same, are you suggesting that you're a rat looking for a way
out of this aqueduct?  I think a lot of Linux supporters would be
interested to know.


Matt Benjamin
mbenjamin@comshare.com 

Date: Tue, 22 Dec 1998 11:05:35 -0500 (EST)
From: Jonathan C Day  <j.c.day@larc.nasa.gov>
To: editor@lwn.net
Subject: "Newsmaker of the Year" article

Dear editor,

   ZDNN's "newsmaker of the year" award was a blatant confidence trick,
IMHO. Their "results", as described in the article, have NO connection
with the results actually given in the poll. In the poll itself, Linus
beat Jenni by over 10%! It seems very clear that the "poll" results had
a fixed outcome, regardless of the votes cast. It is one thing for an
editor to "prefer" one newsmaker over another. Indeed, it would be very
unusual for an editor to have so suppressed their own feelings as to
have no preference at all. It is another to produce an article which is
completely false, in an effort to "promote" that preference.
   Linus won that poll, fair and square. (Well, as fair as on-line polls
ever get. :) He deserves the credit for that achievement.

Jonathan Day

Subject: Re: (Online News, 12/10/98 05:17 PM)
Date: Tue, 15 Dec 1998 01:15:44 -0700
From: Alan Robertson <alanr@henge.com>
To: Tom_Diederich@cw.com

Mr. Diederich:

In the article referenced above, you state:

	However, because Linux predates Windows NT and has
	problems taking advantage of capabilities packed in
	current-generation hardware, Enderle said he doubts the
	operating system will ever become a mainstream desktop
	alternative. 

I believe this information is incorrect and misleading.  Version 1 of
Linux came out in 1994.  Version 1 of Windows NT predated that by
several years.  Linux takes advantage of every chip feature (including
3D graphics) of each of the various PC manufacturers, in addition to
those of Sun's chips, COMPAQ's Alpha, SGI's MIPS chips, IBM/APPLE's
Power PC chips, and numerous other chips in addition to the three major
Intel vendors.

According to Microsoft, older releases of Linux run Netscape roughly 30%
faster than NT runs Internet Explorer.  This is not likely to be due to
being unable to take advantage of the hardware.  Additionally, (in the
second of the so-called "Halloween" documents) Microsoft has also stated
that doing OS development for Linux is cheaper and faster than
corresponding development for NT.  This is likely to be a result of
Microsoft NT designers carrying over large parts of the VMS system they
had designed before.  [Microsoft hired away Digital's key VMS designers
to create their New Technology operating system in the late 80's - this
resulted in a lawsuit against MS for stealing Digital technology, which
was settled by MS agreeing to port NT to the Alpha]

Linux is slated to run the new Merced chip in 64-bit native mode within
a month or so of its introduction.  Intel is rumored to be already
running Linux on it in their labs.  This is a reasonable thing for Intel
to do, since if they port Linux to it, then they can have a solid test
base for their new chip architecture, without relying on outside vendors
(since they already have the source).  In some ways, it is a CPU
designer's dream come true -- a portable, retargetable compiler, a
highly portable modern OS inside, and no need to motivate an outside
software company to do the work, or disclose details of your design to
them.

As you are no doubt aware, Microsoft does not claim that Windows NT will
run native on the Merced line until sometime in late 2000.  64-bit mode
will take longer.  If history runs true to form, it is nearly certain
that there will be some delay from a prediction made this far in
advance.  If recent history were used as a guide, one might conclude
that a significant delay is likely.

Although it started later than NT, it already runs in 64-bit mode on the
DEC Alpha, and the Sun UltraSparc CPUs.  On Sun's UltraSparc line, it
currently runs on 14-processor complexes.  Sun has committed to provide
test time for Linux developers on their 64-way multi-processor systems.

It seems to me that it if one were to make a comparative claim about the
age and modernity of Linux and NT, that a very different comparison
would be in order.  If you would like references for any of the
information I presented above, I would be happy to provide them on
request.


	Thank you for listening,

	-- Alan Robertson
	   alanr@henge.com

Date: Tue, 22 Dec 1998 07:51:26 +0800 (CST)
From: Hung(2) Chao(2)-Kuei(4) <ckhung@cyut.edu.tw>
To: editor@lwn.net
Subject: Re: Bittersweet victories


Dear Editor,


Most of us know that it will be good for the proprietary UNIX vendors
to become Linux VARs. Let's make this option clear to the proprietary
UNIX vendors. And there is something they need help each other with
along the way.

For one thing, the general public need to realize that the key to
productivity does not lie so much in obtaining and installing a lot
of software programs, be it proprietary or open source. Very often
it is how well they are configured that makes a real big difference
in productivity. As OSS goes main stream, 2 more dimensions of
activities, which were previously too costly with proprietary software,
emerge to affect productivity, namely choosing which programs to
download and hiring people to customize programs. Choosing, configuring,
and customizing free software (how about calling it "3C's for OSS")
requires IT proficiency. It is necessarily too customer-specific 
for the free copying of OSS to do much help (or harm, if one takes
the vendor perspective). In short, it is where business can make
money while customers can really see their IT expenditures turn into
productivity.

We know this all along, but the proprietary UNIX vendors (and
proprietary apps vendors, too, for that matter) need be shown a
clear .. uh .. road ahead. Instead of spreading FUD against Linux
in vain (read: resistance is futile), it will benefit themselves
and customers much more by lining up their propaganda towards
showing the difference that effective 3C's can make. Their business
can sail better along with rather than against the OSS currents,
by emphasizing that their experience in UNIX helps them supporting
Linux better, and that "that other operating system" lacks
flexibility in 3C's regard. 

And we as a community should credit vendors for advertising 3C's,
perhaps even more so than vendors porting apps to Linux. Personally
I am glad to see Corel releasing WordPerfect binary for Linux, but
can't help worrying about the sweet relationship turns sour and
bitter when an OSS alternative takes over. By helping the
proprietary software business seeing further, we help making our
community perceived as more helpful to the business. Besides, most
of us (and the future Linux professionals) depend on a prosperous
VAR business model. Yet a very large potential portion of the market
seems still very unmature and only vaguely defined as "support".
The OSS community will also benefit in the long term if the losing
proprietary UNIX vendors systematically advertise 3C's.

Any comments on these points are appreciated. If these points are
valid, I will expand this note and keep it at:
	http://www.cyut.edu.tw/~ckhung/published/oss3c.shtml
Thank you and Merry Christmas!

Chao-Kuei Hung
ckhung@cyut.edu.tw
http://www.cyut.edu.tw/~ckhung/