[LWN Logo]
[LWN.net]

Sections:
 Main page
 Linux in the news
 Security
 Kernel
 Distributions
 Development
 Commerce
 Announcements
 Back page
All in one big page

See also: last week's Back page page.

Linux links of the week


Linux Dev.Net is a new Linux portal site aimed at developers. It has the usual mix of news and links along with a jobs page, classified ads, and a planned future "Linux knowledge base."

The Database of Orphaned Open Source Softwareseeks to be a clearinghouse for packages in need of maintainers. They are just getting going and looking for help, currently. If you have anything to contribute, consider helping them out.

Section Editor: Jon Corbet


June 3, 1999

   

 

Letters to the editor


Letters to the editor should be sent to letters@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: Fri, 28 May 1999 17:40:08 +0100
From: Hans Schou <chlor@inet.uni2.dk>
To: lwn@lwn.net
Subject: Oracle/Linux for Enterprise

http://lwn.net/1999/0527/press.phtml
Greg Shiply wrote that "Linux is not powering Oracle databases yet."
Well, I got one. It has been running since 10 October 1998
with a pre-production and now with a "production"

It is the one and only Oracle server we have at the office
which is the headoffice for $ 3 billion company. We use it for
a lot things but currently it mostly for consolidation. We more
90 companies in 32 countries with nearly 200.000 employees
world wide. It is not transactions but it is our main financial
data.

As we are in the beginning with using Oracle I guess we will 
only see 100 simultanious users this year but it will grow.

Oracle has our main data - this is not for fun.

We don't use an old pc which we found in the basement. It's
a Dell 6300 server with 512MB RAM, 40G disk and dual 400 MHz.

It has been taking down sometimes during these months for service
but we have never seen a breakdown. I think it has been running
at least for two months without been taking down for service.

How I convinced my boss? Easy. I told him that NT 4 sp3 would
not run stable with Oracle so while we are waiting for Windows 
2000 we are running on Linux.
Ofcource we will switch over to Windows 2000. It will run
faster and more stable than Linux, I've been told...

Except for one other server all pc's in the house runs Windows NT.

---
BTW a lot of people has Linux for DNS. If you don't call
DNS "mission-critical" then try switch the thing off.

-- 
best regards
+-------------------------------------------------------+
! Hans Schou,      Hamletsgade 4-201,     DK-2200 Kbh N !
! Fax : +45 3391 5310             Phone : +45 3586 1266 !
! mailto:chlor@schou.dk             http://www.schou.dk !
+-------------------------------------------------------+
   One Day - One World - One Operating System - Linux
      12 September 1999 - http://www.linuxdemo.org
   
Date: Fri, 28 May 1999 12:23:40 -0400
From: Walt Smith <waltech@bcpl.net>
To: letters@lwn.net
Subject: forbes article

hi all,

It was pleasant reading the Forbes piece.  It was
one of the most objective short marketing pieces I've seen,
with some real numbers and focus.  The author makes  a case
more from the POV of a spreadsheet than knowledge of the
market potentials. This is refreshing since so much is written
only from wishful fantasies or bad interpretation of numbers.
A compromise perspective between the two is best.

This is why I object to the statement about $49 software.

How much is DOS? How much is Win31?  (yes, it's still sold).
Does Win95, at $90 come with a real technical person to talk to
if there's a problem?  And a real installation manual?  and a C
compiler?
and...   What is the real cost to improve and package the RH product?
Does the author imply that if RH raised the Linux price to $90 it
will result in increased profits.  Has the author looked over the
software
selection and prices on shelves at CompUSA lately?

If the real (though still unproven) market profit is in the support of

this unique product, then doesn't selling inexpensive product as
a lead-in make sense?  It does to me - especially if the lead-in
is actually making money!

regards,

Walt Smith, Baltimore

   
To: editor@lwn.net
Subject: KDE Wars
From: Nathan Myers <ncm@nospam.cantrip.org>
Date: Thu, 27 May 1999 01:12:13 -0700

Discussion of KDE usually involves confused analyses of political 
and licensing issues.  I'd like to bring out a more fundamental 
issue, which deserves attention mainly because it can still be fixed.

The KDE system is constructed on top of Troll Tech's Qt library.  
The Qt library incorporates a fundamental flaw which will haunt 
all development which uses it until the flaw is fixed.

The flaw is this: the Qt libraries depend on the use of non-syntactic
macros with the names "signals", "slots", and "emits".  Any other 
library which uses one of these names in any capacity -- function 
argument, local variable, struct member, or function name -- may be 
disrupted if used in a program which includes the Qt headers.  

In effect, Troll Tech and KDE have elected to create their own private 
dialect of C++, claiming three extra keywords.  The Qt and KDE team 
members could (and can) rescind this choice any time by renaming these 
macros to something like Qt_SIGNALS, etc., but they have elected not 
to do so (yet).  The more software that is written on top of the Qt
library as it is, the more disruptive fixing it will be, so it is 
important that it be fixed quickly.

The solution remaining for the rest of us is to assert our right to 
these names by using them freely, in header files of other libraries, 
as formal argument names, struct member names, member-function names, 
and as local variables in inline functions:

  inline int do_stuff(int signals) { int slots; ...

We can also insert "#undef signals", etc., directives.  Eventually, 
as they find it increasingly difficult to build programs that rely on
useful non-KDE libraries, the KDE developers will be forced to give up 
their claimed monopoly on those names, and begin to act as responsible 
members of the cooperative software development community.

The alternative is to yield these new keywords, and carefully avoid 
using them in any code we write which someone might someday want to 
link into a KDE program.

Because competently-written C++ code has an order of magnitude fewer 
bugs than C code written with the same competence, the KDE project has 
a chance at a proud accomplishment.  I hope they do not squander that 
chance by arrogantly insisting on trying to add new keywords to the 
language, as this could only poison their product's future.

Nathan Myers
ncm@nospam.cantrip.org  http://www.cantrip.org/

   
Date: Thu, 27 May 1999 15:53:45 -0400
From: Joseph J Klemmer <klemmerj@webtrek.com>
To: letters@lwn.net
Subject: On Window Managers and Desktop Envirenments

	I don't know about anyone else out there but I'm getting a
little tired of the GNONE/KDE "debate".  They are both what they are
and they both have their strengths and weaknesses.  In my personal
opinion, and I stress that this is MY personal opinion, they both are
neat things to play with but they are poor work environments.  Oh,
they probably do well with the people coming over from the dark side
but they have to many bells and whistles and junk that just weighs
them down.

	For anyone else out there who might want to use their system
rather than just spend hours configuring it they might want to look at
the XFce/XFwm window manager and desktop.  It's a CDE-like desktop
with some outstanding features (D'n'D, total GUI configuration, very
small and fast, etc.) that doesn't get in the way of getting the work
done.  You can find it at http://www.xfce.org.

	FWIW, I think if all the effort and work that's gone into
GNOME and KDE were spent on something more productive Linux would
already have kicked the dark side off of the corporate desktop.  But
what the hell do I know anyway.

Thank you for the time,
Joe "No, I *don't* have anything better to do!" Klemmer

---
The most exciting phrase to hear in science, the one that heralds new
discoveries, is not "Eureka!" (I found it!) but "That's funny ..."
                -- Isaac Asimov
   
Date: Sat, 29 May 1999 14:07:42 +0200 (MET DST)
From: David Kastrup <dak@neuroinformatik.ruhr-uni-bochum.de>
To: letters@lwn.net
Subject: Qt freeness


There have been numerous complaints that Miguel de Icaza has
misrepresented the status of Qt, and there has been repeatedly stated
that KDE/Qt has been declared as "free" from the OSI and others.

To this the following comments apply:

It is still technically correct that no version of Qt has yet been
released under the new QPL license under discussion.  No proper
release of Qt up to now satisfies the Open Source Definition.  Qt 2.0
probably will do so, but it has been announced for more than half a
year already and still is not released, if I am not mistaken.

But this does not even border on the problems people see with KDE.
The problem with KDE is not that its license is non-free (being GPL
and LGPL is very much ok for free software), but that its license is
conflicting with the license of Qt, making redistribution of the
KDE/Qt combination a doubtful enterprise.  While the new license of Qt
will in all probability be Open Source compatible (what it was not
before), the main problem is that KDE's license will probably still
not be Qt-compatible after the change.

The current proliferation of "Open Source" licenses like QPL and MPL
and others has the disadvantage of effectively crippling free software
development, as the resulting products cannot be easily combined to
mutual advantage.  The continuing KDE/Qt controversy even after
announcement of the plan to use the new QPL for future Qt products is
just one sad example.  I would strongly suggest to people planning to
license their software as Open Source and feel the necessity of
cooking up their own license for this, that they consider at least
double-licensing the software with a choice of GPL.

David Kastrup                                     Phone: +49-234-700-5570
Email: dak@neuroinformatik.ruhr-uni-bochum.de       Fax: +49-234-709-4209
Institut für Neuroinformatik, Universitätsstr. 150, 44780 Bochum, Germany
   
From: "Matt.Wilkie" <Matt.Wilkie@gov.yk.ca>
To: Press@ActiveState.com
Subject: MS and Perl
Date: Wed, 2 Jun 1999 09:13:59 -0700 

Hi,

I sincerely hope ActiveState will work very diligently and with great
integrity to insure that all work on the Win32 port of Perl will remain
true to it's *nix heritage. This means a lot of open source code, no
"embrace and extend" (we do *not* need another Java/html/browser
 war) -- true and complete cross-platform compatibility and openness.

I freely admit I possess doubts about any organization's ability to
maintain integrity when it's pockets are being lined by MS (or any other
single organization of similar stature). Please do not merely categorize
this as "paranoia" and "ms-bashing". There are many precedents which
give foundation to this nervousness.

While I intend to continue to learn and utilise Perl on both Windows 
and linux, I will be staying away from ActivePerl until the test of time
can prove my fears groundless or justified. I won't be alone on the
watch. I entreat ActiveState to remember that while your coffers are
being filled with a wave of new Windows users, there are many of
us who are reserving judgement.

-matt 
 

 

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