Carol A. Kunze ckunze at ix.netcom.com
Wed Jan 31 20:58:46 UTC 2001

>on Tue, Jan 30, 2001 at 05:23:14AM -0800, Carol A. Kunze
(ckunze at ix.netcom.com) wrote:
>> I do not believe implied warranties should apply to open source
>> software.  IAAL, however, IANYL and this is not intended, nor should
>> it be construed as legal advice. 
>> I will confine my remarks to one point, although there are others.
>> Implied warranties on open source software do not make sense because
>> there is no  license income to support granting a warranty.    
>I think you'll find general agreement with this viewpoint here, once all
>parties understand you're talking about _implied_, not _explicit_
>warrantees.  I believe there's confusion on this point in David
>Johnson's response. 

He didn't sound confused to me.  David?

This is not the first time I have heard the viewpoint that imposing
warranties on open source/free software is a reasonable thing to do.

>Free software is more or less predicated on the principle that a free
>software license can disclaim implied warrantees and laibility.  With
>the slate wiped clean, _express_ warrantees or service contracts can be
>provided, as value-added service, and a revenue stream.
>I also suspect I know where this is headed.  I know Carol from the
>CNI-COPYRIGHT mailing list and her work with UCITA (she graciously
>offerred assistance as I prepared my licensing presentation at last
>summer's O'Reilly Open Source Conference).  I suspect a UCITA plug in
>the offing, though I doubt it will be warmly received here.

>> Let me first note that this view does not include express warranties,
>> described in statutes as "an affirmation of fact or promise, . . .
>> including by advertising, . . . any description . . . .  "  
>> There can be no license income because every copy carries with it the
>> right for the user to make more copies and to distribute those copies
>> in competition with the original supplier.  One cannot extract a
>> profit on software under this scenario - at least not for long.  So in
>> practice, it is free beer.
>Business-issue quibble:  I don't know that this is the case.
>Risk-averse organizations themselves (e.g.:  businesses) typically
>benefit from the service, support, and guarantees available through
>commercial distribution of free software.  Last I heard, Red Hat was
>still generating significant revenues on "direct software sales" --
>shrinkwrap and service-bundled sales. 
>> Red Hat's standard Linux product sells on disks for $29.95 with
>> printed documentation and installation service.  It competes with Red
>> Hat's own free download version.  It also competes with the Red Hat
>> Linux version distributed on disks by Cheapbytes for $4.99. 
>...and with various other bundles ranging in price to several hundreds
>of dollars, possibly thousands, per sale.  The bits are free.  The
>chrome, services, and support, are not.  Note also that it isn't
>necessary for RH to support the development effort for all software
>under its lable, but merely the aggregation, anciallary documentation,
>and administrative support tools RH itself provides.
>> Market economics will not tolerate a profit on the software under this
>> competitive scenario.  In fact, Red Hat is selling the medium, printed
>> documentation and services for $29.95, but the software is free. 
>> If Red Hat could extract a profit, someone would immediately set up a
>> competing business to undercut its price, and then that price would be
>> undercut by another distributor and so on until the price was reduced
>> back to zero.  
>Effectively the cost floor appears to be ~$10-20 (including shipping and
>handling) for a delivered disk set for a distro.  Or free download.
>> Keep in mind, Windows operating system sells for what, about $200?
>> $319 for the professional version last time I checked.  
>OT:  The Register did a rough valuation of what the "fair market cost"
>of a GNU/Linux distro would be, some years ago, arriving at something on
>the order of $5000, IIRC.
>> There are also unanswered questions with respect to open source and
>> warranties.  Microsoft gets paid for every user of Windows (discounted
>> price of $1,199 for 10 users). Compare this to a small business which
>> "buys" a copy of Linux for $29.95 and makes 9 more copies.  Has the
>> supplier warranted 10 copies or 10 users?
>Given the present disclaimers in most free software licenses, the
>correct answer is, barring explicit warrantee, zero.

Warranty disclaimers can be challenged on a number of grounds - the clause
was not conspicuous, the user did not agree, the user did not have notice
of the disclaimer, etc.  There is still risk here, particularly with open
source which is more likely to be distributed informally.

But in fact, I'm discussing what the appropriate default rule is in the
first place, not whether it can be changed by disclaimer. 

>> The issue is worse when it comes to the actual authors who may sell
>> some free software  basically for the price of the disks: "If
>> free-software authors . . . find themselves getting sued over the
>> performance of the programs they've written, they'll stop
>> contributing free software to the world.  It's to our advantage as
>> users to help the author protect this right."
>> -- Bruce Perens, Open Sources:  Voices from the Open Source
>>    Revolution, 1999
>> Given that open source software does not generate license income,
>> implying a warranty is not reasonable.  Put another way, open source
>> software cannot even afford to win a warranty lawsuit. 
>Again, no argument, this is commonly accepted fact.  

David seems to have a different opinion.

It is *not*, AFAIU,
>any rationale for adoption of UCITA.  There are existing conventions, if
>not legal doctrines, restricting the imposition of implied warrantees
>for freely performed services -- e.g.:  good samaterian laws -- and, I'm
>given to understand, for advice, instructions, recipies, technical data,
>etc., published in books and magazines.  This might be an area to
>explore.  If specific carveouts for free software are required, there
>are far better vehicles than UCITA to accomplish this task.
>As you're stating what I believe is a commonly accepted truth in the
>free software world, I have to ask:  what's your point?

I'm not sure how I can be any more clear.  As I said in the beginning, "I
do not believe implied warranties should apply to open source software."  

How about this:  I think a more appropriate rule would be that there is no
warranty on open source software unless one is expressly offered. 

Carol A. Kunze 
Napa, CA
ckunze at ix.netcom.com
707.371.1807 (fax) 

More information about the License-discuss mailing list