Re: NO! (Re: Does GNUstep infringe on Apple's Intellectual Property?)

[Gregory John Casamento]

Scott,

--- Scott Christley <schristley@mac.com> wrote:
> 
> Well I still disagree, but it isn't my intent to get involved in an 
> email war, so I will make a last reply and leave it at that.  

I too don't have the time/energy for this type of thing.  I would much rather
code.

> As always 
> John, it is best if you talk to a real lawyer regarding your specific 
> situation, but I seriously doubt you will run into any trouble just 
> because you want to make some of your programs available for GNUstep.  

Agree.  You should consult a lawyer, if you have doubts.

> My knowledge of the material is certainly not preventing me from using 
> GNUstep; neither back then as a commercial entity nor now with my 
> research.
>
> The main reason why I am confident in my assessment is because when I 
> wrote the windows32api library and donated it to FSF, I talked at 
> length with rms and FSF's lawyer because I didn't want Microsoft coming 
> after FSF due to copyright infringement.  I encourage you to read the 
> document in that distribution which explains why I took the approach I 
> did in order to minimize the risk, but as the lawyer was keen to point 
> out if MS wants to sue then they will sue regardless.  The discussion 
> with the lawyer was enlightening because it cleared up a number of 
> misconceptions that are common:

I looked at the windows32api library and I don't see the file you refer to. I
downloaded the latest version from the FSF.
 
> * Names, like functions names, class names, structures, definitions, 
> etc. can all be copyrighted.  Obviously it would be difficult to 
> copyright a single word.  The copyright becomes stronger when they are 
> included together in a composite work such as an API.

I've been told by a lawyer that a particular implementation or expression of
them can be, not all possible expressions.   This is the difference between
copyrights and patents.

> * What the law is very clear about is that original written expressions 
> are most definitely protected by copyright law.  A document describing 
> an API like WIN32 or OpenStep is clearly an original written 
> expression, and it most definitely is protected by copyright law.

This agrees with the above.

> * The law is vague and lacks clear precedence about copyrighted "look 
> and feel".  Most people, and I did with the lawyer, point out the old 
> Lotus 123 case.  Understand that that case was along time ago, and it 
> was won by a slim majority.  However, it won based upon the concept of 
> "look and feel" not upon a programming interface.  It was a difficult 
> case because the judges didn't know how to apply law that was geared 
> towards written words to a graphical interface.  Since then there have 
> been many lawsuits regarding look and feel copyrights (do a google 
> search on 'legal look feel' to get many hits); these cases have gone 
> both ways and is still a raging debate.  Extending these decisions to 
> APIs, even if the API is for a graphics, is a bit of a stretch.

The look and feel thing, I thought, was settled long ago by the Apple v.
Microsoft case.   I'm not extending it too APIs without precedent.

Another case Apple v. Franklin helped establish the fact that APIs cannot be
copyrighted.

Dont take my word for it: from the case Dymow v. Bolton, 11 F.2d 690, 691 (2d
Cir. 1926).

Just as a patent affords protection only to the means of reducing an inventive
idea to practice, so the copyright law protects the means of expressing an
idea; and it is as near the whole truth as generalization can usually reach
that, if the same idea can be expressed in a plurality of totally different
manners, a plurality of copyrights may result, and no infringement will exist.

This case was heard by the US Court Of Appeals.   The above supports my
interpretation.

> * There are numerous instances where APIs are clearly considered 
> confidential trade-secrets and the copyright owners will vigorously 
> defend its rights.  A recent example are DVDs; you are not allowed to 
> write a program which will read (decrypt) a DVD without a license, as 
> that one russian? guy discovered.  A DVD isn't strictly an API, but it 
> could be considered an advanced data structure which are often part of 
> APIs.  IBM is another company which keeps some of its APIs close to the 
> breast; they've gone after numerous companies who have duplicated their 
> mainframe and minicomputer APIs, both software and hardware.

Where?  These would be APIs which are not released to the public correct?  If
Microsoft are any other entity release an API, especially when it's released as
NeXT and Sun did, it's fair game.

> * Translation, rearrangement, compilation, etc. maintains the original 
> copyright.  So if you wrote a program which read in the OpenStep spec, 
> did some fancy translation work, then spit out a rearranged version; 
> that output is still covered by the original copyright.  Most of these 
> actions produce a composite work of multiple copyrights; the original 
> plus yours for the "original translation".  This is why you find that 
> special clause in the GPL about how the license does not apply to OS 
> libraries; because most programs need to #include in a system header 
> file (copyrighted by the OS owner) and therefore the resultant compiled 
> code is a composite work.

Yes, because the rearranged version of the spec is a "derived work" since it
was created utilizing the original.  If however, I write my own book about
openstep for each method and each class without referring to or using the spec,
then the work I create is NOT a derived work.

None of the GNUstep code uses any Apple headers.

> Anyways, what I took from my discussions with the lawyer is that APIs 
> are well protected by copyright law.

Name/email for the lawyer, please?

> The notion of a "derivative work" just confuses the issue.  If you are 
> talking about infringing on a copyright then the minimal action to 
> infringe is to include portions of that copyrighted work in your own 
> work.  Take a look at any of the major header files in GNUstep, you 
> will find portions of the OpenStep (method names, class names, 
> structures, etc) specification.  This is clearly inclusion of 
> significant if not all of the OpenStep copyrighted material.

It's our own implementation of it!!  Obviously we need to have headers which
refer to the methods and classes, how else are we supposed to create it??  
Unless someone has cut/pasted Apple headers into the code, then we're fine.

Apple's copyright covers their implementation of it not ALL POSSIBLE
implementations of it.   Taken literally what you're saying is if I create a
class with the same name as one from Apple's framework that I am infringing on
Apple's copyright.  That's ludicrious.  

If this were the case then every implementation of GNU/Linux is running in
violation of AT&T's/SCO's copyright as is BSD (there was a case which disproved
this, in 1991 when AT&T tried to claim copyright over BSD) and so would the
eventual GNU/Hurd system!   Every person running WINE or FreeDOS, or any OS
which uses them, would be in violation of Microsoft's copyrights.  OpenBEOS
would be in violation of Palm/Be's copyrights. etc etc etc etc etc...

No offense, but I firmly believe that you've misunderstood what the lawyer told
you.

> Even though the OpenStep specification was published as an "open" 
> (whatever that means) standard; it was never released under a license 
> which essentially allowed GNUstep to incorporate that API into its 
> header files.  Implementation has nothing to do with it because the 
> issue isn't about the implementation source code, the infringement is 
> in the header files.

Any published API can be implemeted as GNUstep has implemented OpenStep.

> Personally, seeing GNUstep track Cocoa is worrisome to me; it keeps 
> GNUstep on Apple's radar screen.  Obviously I can understand the 
> reasons why people want to do that.  I would much prefer to see GNUstep 
> slowly go its own way, but then I'm busy developing other free software 
> and just a GNUstep user these days. :-)
> 
> cheers
> Scott
> 

I agree that GNUstep need to find it's own way.   We should, however, try to
maintain some degree of compatibility.

Later, GJC

=====
Gregory John Casamento
-- bheron on #gnustep, #linuxstep, & #gormtalk ----------------
Please sign the petition against software patents at:
http://www.petitiononline.com/pasp01/petition.html
Petition to make Lighthouse Application Suite Free Software at:
http://www.petitiononline.com/laafs/petition.html
--- Main Developer of Gorm (featured in April Linux Journal) ---

__________________________________
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
http://sitebuilder.yahoo.com