[cfgeeks] Good points (binary modules banning talk)

[Kevin P. Inscoe]

----- Forwarded message from jb1 at btstream.com -----

Date: Fri, 15 Dec 2006 14:25:11 -0800 (PST)
From: jb1 at btstream.com
To: Jeff Trull <linmodemstudent at gmail.com>
cc: LinModems <discuss at linmodems.org>,
	Pietro Filiberti <btech at olympus.btstream.com>
Subject: Re: binary kernel modules banned after January 2008?

After looking through some of this thread, two points seem to stand out:


1. In their arrogance, the proponents of eliminating non-GPL drivers
delude themselves into thinking that they can actually pressure
manufacturers into releasing their source code, whereas those of us who
have followed the linmodems discussions for any time know that the
manufacturers almost universally *don't care* about Linux!. They don't
care about Linux because it's too small a market, which will become even
smaller because they will take the only sensible course of action ...  
they will simply end what little support they now provide. Part of my 
evidence for this is that it's becoming increasingly difficult to find a 
modem, printer (or, I suspect, sound card, advanced video card, etc.) that 
doesn't require a software driver.


2. The proponents of eliminating non-GPL drivers are apparently unaware of
facts Marv pointed out to me (jb1) in an e-mail dated 07 September 2003:
  A few issues seem worth clarifying concerning distinct roles of
  Proprietary Information (PI) and Open Source (OS herein) code.

  All important computer components embody PI protected by either Patents
  or Copyrights.  The PI may be in hardware and/or complementary software,
  which only together provide functionality.  
  ...
  ... But these modems still have much IP embodied in their
  chipsets. This includes the various Vnn compression algorithms, which
  the chipset makers pay royalties on.
  ...
  ... For modems there is an additional Danger. If connected to
  the exterior phone system, faulty modem hardware + software could
  disrupt phone service.  For modem chipsets selling world wide, the
  hardware+software combination must adapt to the needs of diverse
  national phone line standards, ie, the homologation standards.  This is
  an additional reason, for chipset manufacturers to encrypt more critical
  code either as hardware of PCB software. In these forms, code amateurs
  without sophiticated test equipment cannot easily do harm.  

  Additionally one could expect that the same Idiots who creat viruses for
  Internet disrupt would love to disrupt phone systems too.
Perhaps Marv would care to elaborate on these facts.

I'm not competent to "argue our position well on LKM", but whoever is
should point out that the amount of hardware useable with kernels that
prohibit non-GPL drivers is more likely to evaporate than to increase ... 
in addition to all the other arguments against making such a 
non-Linux-like requirement.

On Thu, 14 Dec 2006, Jeff Trull wrote:

> I just saw this on a kernel blog:
> 
> http://thread.gmane.org/gmane.linux.kernel/475654/focus=475721
> 
> The motivation given is to put further pressure on the makers of binary-only 
> drivers to release source code.  I hope that we can persuade someone there 
> needs to be an exemption for our situation, where the binary-only (or 
> binary-mostly) drivers are for hardware that is in many cases no longer 
> manufactured.  Given that the manufacturers no longer have any revenues from 
> the products (and in some cases no longer exist), they have no incentive to 
> release the source, regardless of whether their products work under Linux.
> 
> Is there someone who can argue our position well on LKML?  Knowing the 
> orphaned drivers I'm helping to maintain (PCTel, ESS) may be banned from the 
> kernel lowers my motivation to do any further work on them.
> 
> Thanks,
> Jeff Trull
> 

----- End forwarded message -----

-- 
Kevin P. Inscoe                       Amateur Radio Call Sign: KE3VIN
Deltona, FL 32738                                28.9497N by 81.1952W
kevin [at] inscoe [dot] org                    http://kevininscoe.com
GPG 0x61288D53