[ Date Index ] [ Thread Index ] [ <= Previous by date / thread ] [ Next by date / thread => ]
On Thu, 12 Apr 2007 00:56:41 +0100 Julian Hall <lists@xxxxxxxxxxxx> wrote: > http://www.ethipat.org/ IMHO, there is an impossible conflict within the ethipat proposal - software patents are inherently unpatentable. How can you have an ethical patent system if patents are allowed on unpatentable software? Conversely, if software is excluded from the patent system (globally), is there a role for an ethical patent system and should software lobbies be involved? > I commend the above petition to you, patents are now stifling > innovation rather than promoting it, costs of dicines, technology and > software are artificially inflated by them. Genes are being patented, > these are not inventions or innovations but living tissue that has > been manipulated. When patents were thought of they were to allow the > little man to benefit from his labours, now only big business can > afford the process. Patents are undoubtedly an immense threat to free software - unrestrained, patents could close down free software for 50 years. If anyone here wants the next generation to be able to use free software they must be prepared to ensure that software patents remain unenforceable. Patents are no idle threat - the danger is real, the problems well recognised and the timescale imminent. Patents are arguably the single biggest threat to free software, eclipsing proprietary development and MS itself - principally because MS is not shy of using patents to advance the MS proprietary agenda. Microsoft know the power of the US patent system - that is what the deal with Novell was all about, despite Novell's protestations. It is no surprise that free software proponents avoid MP3, GIF (although that has since lapsed) and other patented codecs - even whilst the patents are still unenforceable we need to drive forward the development of unpatented open replacements like Ogg Vorbis. I'd much prefer that the UKPO and the EU PO strictly follow the existing rules on the non-patentability of software and convince our US cousins to reverse their current patent policy. Whether there is a technical contribution or not (and whatever that weasel phrase actually might mean), a method implemented in software should not be patentable. It sounds simple, but it actually gets complicated when you start considering that perennial problem: firmware. At what point does hardware design finish and software programming begin? If the manufactured hardware doesn't support flashing the firmware but a method is later devised to do exactly that, does it mean that the patents on the hardware design cannot be enforced because the "infringements" are a software method? (Think XBox or Cisco). If software that is or modifies firmware is patentable (or can infringe an existing non-software patent which amounts to the same thing), now where do you draw the line? How can some software be patentable and some software non-patentable? You cannot make a rule based on the use of assembly language because all languages are converted to assembly before execution. All software is converted to binary at some point. You cannot make a rule that says that firmware is excluded or included because 'firmware' cannot be strictly defined. Software itself isn't fully defined - we know what it is but creating a definition that covers all the corner-cases isn't easy. Now imagine managing this complexity in the ethipat framework. Software only needs copyright and copyright holders must remain free to reassign copyright and renounce it completely for public domain code. To me, at least, any patent claim should become automatically invalid if the claim can be represented in source code. That blows a huge hole in the current patent system wide enough to scare any patent lawyer witless but I believe there would be enormous benefits. Imagine a patent system where the claims had to be verifiable by a strict process of input and output results - if a software implementation became available that matched the test suite results, (which would be almost inevitable), the patent would be legally deemed unenforceable. Could the patent system survive? (Solutions would have to show evidence of independent source code development to rule out simple reverse engineering and all software solutions would have to be available publicly.) Basically, such an idea all but removes patents from the entire computing industry, leaving only pharmaceutics and other non-electronic inventions. Almost anything that can be done in a specific piece of electronic hardware can be done with a suitably reprogrammed generic piece of electronic hardware. After all, whether the job is done in hardware or software, it is all just binary - 1011100 in, 0100111 out. Software is usually slower but it can still do the same job. Patent systems across the world accept that a mere increase in speed is insufficient grounds for a new patent. For a general overview: http://www.nosoftwarepatents.com/en/m/basics/index.html "If people had understood how patents would be granted when most of today's ideas were invented and had taken out patents, the industry would be at a complete standstill today." Bill Gates (1991) http://www.nosoftwarepatents.com/en/m/dangers/index.html Read the ethipat site but also read www.nosoftwarepatents.com and come to your own decision: Can the patent system ever benefit software - ethically or not? For more background see: http://www.dcglug.org.uk/linux_doc/patent.html http://www.dcglug.org.uk/wiki/?id=Software+patents I believe that only a strict observance of the *current* treaties that explicitly prohibit patents on software - WITHOUT any caveat for arbitrary definitions like "technical contribution" or "business use" - can ensure innovation in software and prevent a complete logjam in the coming decade(s). If a side-effect of this is that firmware of all kinds is classed as software I won't mind a jot - open standards are the way forward and anything that breaks the stranglehold on mobile device development can only be a good thing. The public may think that their mobile device is made by a high street brand but it is not, most handsets are made by a small number of independent manufacturers under licence and rebranded for the high street in a flashy plastic cover. It is possible to deal with the actual manufacturer and bypass the brands, their marketing costs and restrictive business methods. It is possible to create a standard mobile device and potentially revolutionise the mobile market in the same way as IBM did with the PC. No more incompatibilities masquerading as unique selling points, no more locked handsets - more choice, more freedom, more opportunities for free software running on the mobile device and easier installation of that free software, independent of the precise device model. Why shouldn't it be as easy to install Debian on a mobile device as it is for a desktop PC? Why should it be as hard to migrate your phone records to a new device as it was to transfer your data from a BBC micro to an Amstrad PCW? Customised cables, customised protocols, incompatible data storage - it's all bunk. Standard USB cables exist that can fit any mobile device, the sockets are small enough and the software exists to implement USB2.0 on all mobile devices, turning them all into usb mass storage units - a usb key with a phone function. Why are we prevented from doing this? Incompatible hardware using closed designs and aggressive patents that try to cover the software used to create, store and retrieve YOUR data. (Remember: the data on that SIM is YOUR data - manufacturers have no right to restrict you in how you choose to access, modify, update, delete or disseminate YOUR data. You generated / created that data, you choose what happens to it. Why accept anything less? Would you accept that attitude from Debian or Ubuntu?) Free software cannot afford for patents to be enforceable against software because patent infringements can ONLY be judged in a court of law and the costs are impossible. If you beat down the first 7 patents, they'll just come back at you with another 14. The guy with the deepest pockets wins, end of. I'm a pharmacist by trade - I know the benefits of patents in pharmaceutics (even then patents only work with finite time limits that are absolute, strictly enforced and as short as possible) and I can see a role for patents in fields that involve the consumption of physical resources to produce implementations of the patented idea. However, software is not a physical entity (it can be stored on physical media but software itself has no physical presence, it is only a pattern of electrons) and therefore cannot be constrained by legal constructs that are designed for physical entities. The laws of property do not apply to software - you cannot steal software, you cannot commit piracy with regard to software just as I cannot accuse you of theft if you use words that I have used. Plagiarism, yes. Theft? No. Software takes form only as an expression of speech and as such is perfectly suited to protection by copyright alone. I do not agree with ethipat and I will not be signing. I do agree with GPLv3 and I will be licencing all my future code exclusively under GPL3 once the draft is finished. My current GPL2 code will remain available under GPL2 or 3 at your discretion, but future code will migrate to GPL3. (This email was produced, displayed and stored using 100% recycled electrons. It may be freely quoted as long as a link is made back to the original.) -- Neil Williams ============= http://www.data-freedom.org/ http://www.nosoftwarepatents.com/ http://www.linux.codehelp.co.uk/
Attachment:
pgpEdzlMefNzT.pgp
Description: PGP signature
-- The Mailing List for the Devon & Cornwall LUG http://mailman.dclug.org.uk/listinfo/list FAQ: http://www.dcglug.org.uk/linux_adm/list-faq.html