[conspire] We're home

[Rick Moen]

There's been less of a flood of (private) mail in my absence than I
expected, but inevitably I get stuff like this complaint from a
(probable) DJBware user.  I at least try to be nice, in responding:


----- Forwarded message from Aleksandar Kostadinov <avalon at friendofpooh.com> -----

Date: Tue, 12 Jun 2007 15:00:43 +0300
From: Aleksandar Kostadinov <avalon at friendofpooh.com>
To: rick at linuxmafia.com
Subject: suggested correction

Hi

In regards of http://linuxmafia.com/~rick/faq/index.php?page=warez

tcpserver utility could have a stupid license but it is too much to call it
proprietary. Since everybody is able to copy modify and distribute the
source it's far from being proprietary.

Regards,
Alexander

----- End forwarded message -----
----- Forwarded message from Rick Moen <rick at linuxmafia.com> -----

>From rick at linuxmafia.com Sun Jun 17 05:21:32 2007
Date: Sun, 17 Jun 2007 05:21:33 -0700
From: Rick Moen <rick at linuxmafia.com>
To: Aleksandar Kostadinov <avalon at friendofpooh.com>
Subject: Re: suggested correction
In-Reply-To: <a886f3c10706120500k9263f45o2bb1193652ee67e8 at mail.gmail.com>
X-Mas: Bah humbug.
User-Agent: Mutt/1.5.11+cvs20060403

Quoting Aleksandar Kostadinov (avalon at friendofpooh.com):

> In regards of http://linuxmafia.com/~rick/faq/index.php?page=warez
> 
> tcpserver utility could have a stupid license but it is too much to
> call it proprietary. Since everybody is able to copy modify and
> distribute the source it's far from being proprietary.

Hi, Alexander.

I certainly would be the first to agree that Dan's licensing is very
generous, making it a fine example of the most benevolent _sort_ of
proprietary licence.  

The term need not be seen as pejorative, merely descriptive:  As Dan
himself points out, his licence gives him the ability to prevent other
people from distributing binary-packaged variants of his code that don't 
meet his standards.  From his perspective, that degree of central
control is a good thing -- and I would not dispute that, but it does 
illustrate why it can fairly be classed as proprietary.

You might ask:  What exactly makes a codebase proprietary, and why?  In
this context, the term denotes a codebase that for whatever reason may
not be forked and independently developed and freely redistributed in
any form by all recipients, or may not used for arbitrary purposes
without fee by all recipients.

tcpserver (and most of Dan's other projects, by his explicit intent)
fails the first of those two criteria, because Dan wants to prevent
distribution of binary-packaged variants of his code that don't meet his
standards.  That is his right as author and copyright owner, of course.

At the same time, it is exactly the trait that is gradually causing his
codebases to become progressively more outdated, because he alone has
the legal right to fully maintain them.  With his larger and more
dynamic projects, e.g., qmail and djbdns, this has already become a
large problem, and the problem can only become worse over time unless he
changes the licensing (which seems unlikely).

Consider, for example, the fate of qmail:  Dan's last official release
(1.03) is about a decade old, and Russ Nelson et al.'s "netqmail 1.05" 
source-code patchkit isn't a lot younger and has had only one, now
rather aging, release with no obvious prospects for more.

And that is the most important reason (among others) why proprietary
licensing is (in my view) contrary to users' interests over time:
Proprietary-licensed code inevitably runs into maintenance problems
eventually, because the one party legally permitted to perform that work
dies, loses interest, etc.

Which was intended to be part of the point of my essay.  I hope you did
not take offence, as none was intended.

-- 
Cheers,          "A Discordian is a Taoist with a very strange sense of humour 
Rick Moen         and the inability to sit still."
rick at linuxmafia.com   -- Rabbi Kwan Chi Sun Lieberwitz, _Jews for Buddha Cabal_
----- End forwarded message -----