See also: last week's Letters page.

Letters to the editor

From:	 Andreas Tretow <tretow@snafu.de>
To:	 letters@lwn.net
Subject: Alternative kernel trees will grow in importance
Date:	 Tue, 08 Jan 2002 13:44:27 +0000

In your predictions for 2002 you write that "increasingly, the kernels 
that people actually run will be produced by somebody else."

Wow, what a prediction. Especially since Linus Torvalds now maintains 
the development kernel 2.5. A kernel that is not designed to be run in 
production environments. The stable release 2.4.x, which is being 
maintained by Marcelo Tosatti will obviously be the one that most people 
will sooner or later run on their systems. As the 2.4.x series matures 
it is only natural that people will increasingly switch from the 2.2.x 
kernels and the older 2.4.x (which were maintained by Linus).

Andreas

From:	 "Gregor N. Purdy" <gregor@focusresearch.com>
To:	 letters@lwn.net
Subject: Kernel version nomenclature
Date:	 08 Jan 2002 09:07:33 -0500

Friends --

>From the "my $0.02" department...

With the proliferation of kernel versions we've seen over the last
few years, I'd like to suggest a slight change to the way folks label
their kernels. Just yesterday the Daily Updates section of LWN had a
news item about Linus' 2.5.2-pre10, with 2.5.1-dg13 also mentioned, but
the note says it is caught up as far as 2.5.2-pre9.

I would suggest that "the first 'dj' kernel based on 2.5.2-pre9"
should be called 2.5.2-pre9-dj1, not 2.5.1-dg13. On a similar note,
I'd love to see vendor kernels named similarly. My RedHat 7.2 box
is running 2.4.9-13. Of course, that really means 2.4.9-rh13. The
13th build/revision/iteration/whatever of kernel 2.4.9 made by
RedHat.

As long as we don't end up with derivative chains too many levels
deep, this should work fine and be easier to follow. If everyone's
leftmost chunks match those of Linus' or Marcelo's official kernels,
then its clear from the name alone what code its based on.


Regards,

-- Gregor
 ____________________________________________________________________ 
/            Inspiration >> Innovation >> Excellence (TM)            \

   Gregor N. Purdy                         gregor@focusresearch.com
   Focus Research, Inc.               http://www.focusresearch.com/
   8080 Beckett Center Drive #203                  513-860-3570 vox
   West Chester, OH 45069                          513-860-3579 fax
\____________________________________________________________________/

[finest@newyork.ny.us]$ ping osama.taliban.af
PING osama.taliban.af (68.69.65.68) from 20.1.9.11 : 56 bytes of data.
>From 85.83.77.67: Time to live exceeded

From:	 Kay Hayen <KayHayen@gmx.de>
To:	 peter.w.lawson@noaa.gov
Subject: Installing applications
Date:	 Sun, 6 Jan 2002 22:31:49 +0100
Cc:	 letters@lwn.net


Regarding setup.exe and Linux for installations, 
Peter Lawson writes:

"Usually it works.  Why can't it be that easy in Linux?"

It could be and sometimes does.  

Yet, the "problem" about Linux is that people using their freedom to do 
things they love. Getting things right for 95% is not one of those. They
either make it work for 100% or just themselves and expect others to
make it work for themselves.

Let me ask rhetorically: If Windows works 95% of the time, why don't 
people want to use it, what's so bad about 5% of the installs killing 
your system? You can always reinstall, can't you? 

I personally love about Linux the make-it-possible or do-it-right 
attitude. This is why apt-get exists. In my eyes, it's the do-it-right
for software installation. Now if Debian was only more easy to
install, but probably it never will, since I e.g. don't think I will have
to reinstall Debian all that soon. Simply because they solved how
to update a running system the right way.

Yours, Kay Hayen

From:	 "Marty Leisner" <mleisner@eng.mc.xerox.com>
To:	 letters@lwn.net
Subject: How easy things are with windows...(NOT!!)
Date:	 Fri, 04 Jan 2002 13:38:30 -0500
Cc:	 peter.w.lawson@noaa.gov, leisner@rochester.rr.com


I've installed more than 100 installations
since I started using linux.

Peter Lawson's letter on 1/3/2002 makes statements 
which I take issue with.   Seems I tinker with whatever
machine is in front of me -- and I know if its an open
source machine I can reach a favorable conclusion if I
spend the time.   Otherwise, i often have to give up and
accept a status quo.

I recently spent hours with a friend who bought
a "cutting edge" Dell 2 years ago (nVideo card, 
DVD player, CD-RW).   

We needed to reinstall windows 98 -- after many hours
of playing around, we finally got the system working
(seems the windows distribution CD rom is useless with
this combination of hardware) -- we downloaded drivers from
the net, used "additional" CDs -- I'm going to write dell
an angry letter about how they sell a machine without a 
"cookbook" way to install which works.

I've seen a number of machines where redhat installed
painlessly and windows was a pain in the arse!!
Also the other way around...

Far too many times I click on setup of some application in
windows, and it doesn't WANT to setup for (for some reason).
At least if I can follow a manual path, I can override the
setup which failed -- which is a very rare case in windows.
Many times when installing an application is a problem, I regress to
a "clean machine".

I would recommend to Peter Lawson to take advantage of
mailing lists and search engines like google and to take
old FAQs with a grain of salt.  On second thought, why
doesn't Mr. Lawson contribute something by documenting
his difficulties and providing a "recipe" to get things working.

I've often found installing RPMs for binaries for a redhat system
a breeze.

marty           mleisner@eng.mc.xerox.com   
Don't  confuse education with schooling.
        Milton Friedman to Yogi Berra