[OH Updates] Fwd: the solderpad hardware license

[Andrew Katz]

Hi Andrew


On 28 Mar 2012, at 22:09, G. Andrew Stone wrote:

> 
> 
>> 1. Can a software copyright enforce copyleft onto the hardware its running on?  For example, "if you own copyright to hardware design files AND you are using this software you must license the hardware design files using license XXX".
>> 
>> From a legal perspective, there are a number of problems with this. In some jurisdictions, you are limited from adding additional restrictions to a licence grant over and above what you can control by granting conditional permissions for example. I realise this is just an example but this sort of construction does generally allow for simple circumvention. For example, what happens if you assign the copyright of the hardware designs to a friendly but separate entity, and then license them back to yourself with an explicit no-sublicensing clause. This is the sort of thing that Liberty and Death in the GPL may try to deal with, but you get yourself tied up in knots very quickly. 
>> 
> 
> Isn't what I'm suggesting exactly like the licenses that say "not to be used in weapons" (for example)?  Are you saying that these existing clauses are not enforcible in major jurisdictions?

Nope. That's a licence condition: if you use the software in a weapon, you fall outside the scope of the copyright licence, and I can sue you for copyright infringement. This is different - it's saying that you to take advantage of this hardware design, you must give up some additional rights (also license other documentation you have). What is a matter of debate (and there is no clear answer to this), is how close the connection can be, for the licence to be legitimate. So it's probably ok to say "you have to license any patent rights you have to downstream recipients to the extent that they impinge on this software, as a condition of using the software", but it's probably not ok to say "you must sell your house as a precondition of using this software". This is an example, so don't try to unpick it - the legal theories on which this is based are very diverse depending on jurisdiction - in the UK it would be questionable under principles of non-derogation from grant, and competition law, for example. All I'm saying is that this is a nasty messy box, and I don't want to open it (and coming up with some formulation that actually worked, in any one jurisdiction is likely to be very difficult. In all jurisdictions, likely to be impossible. 

> 
>> 
>> 2. Sure the physical device cannot be copyrighted... but the files can.  Is there a viable legal strategy where the physical hardware is shown to be sufficiently similar to the OSHW (maybe using DNA or fingerprints as a precedent) that corporate data (emails, source repository, etc) can be subpoenaed to investigate whether the engineers began from the OSHW files?
>> 
>> I hate to think where this might lead. You're starting to construct a legal framework which would shut down legitimate on-sale of physical articles like cars and TVs etc. Microsoft and Oracle have managed to keep the sale of secondhand software at an absolute minimum, despite the intention of the Computer Programs Directive to expressly allow it through exhaustion of rights. Do we really want a world where acquiring secondhand hardware (cars, appliances, art....) is considered to infringe IPRs by default.
>> 
> 
> ??? That is not what I am suggesting.  I'm saying nothing about resale.  I'm suggesting research into the following scenario: that manufacturer X claims no copyright violation because he never even used the copyrighted files.  Yet how likely is it that the manufacturer would have produced a board with the exact same routes (to the mil) -- or for sufficiently complex boards selections of chips/packages -- independently?  So from that you prove he did in fact use the files, therefore must adhere to the copyright agreement (whatever that may be).

OK, this is fine, so long as what is being transferred is something that is subject to an IP right in the first place, so watermarking it to track back is fine. Unfortunately, items like circuit boards are a bad example, as they are essentially 2D works and are reproduced as 2D works, so reproduction would be a breach of copyright onto which a licence can impinge. However, if you take a non-artistic design (like a stormtrooper helmet in the UK), and reproduce that in 3d from 2d plans, that is not a breach of copyright, so it makes no odds where the design came from - watermarking may prove derivation, but it doesn't prove copying, and I'd argue that most hardware is like that. What I am concerned about is that this watermarking - ostensibly to enforce copyleft - may lead to the unintended consequences of legislation being passed which plays right into the hands of those who want to apply IP-like controls to physical objects. 









>  
>> Incidentally, this is the reason why the RMS has to tie himself in knots by arguing that the Pirate Party's proposals for copyright are too short, and that there should be an extra-long term of copyright for computer programs so that copyleft can continue to work!
> 
> Yes ironically, copyleft needs strong copyright to keep it away from the public domain :-).
> 
> Cheers!
> Andrew
>  
> 
> 
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhardwaresummit.org/pipermail/updates-openhardwaresummit.org/attachments/20120328/0928cbf7/attachment.htm>