Thursday, July 03, 2003

Computer software – proprietary legal strategies Fisher   Various legal regimes available to software developer    Trade Secret Copyright Patent Contract   I do not believe software is improperly protected by patent.  I don’t think the differentiation is so clean (patent vs copyright).  But, patent law must be properly constructed.  Contract law is much worse, as it has been construed.   Trade Secret   Trade Secret is state law and thus varies but not a lot.   Provides protection to secrets – things that fit include- info to contents (secret formula), process (effervescent power), has been extended to ephemeral events, particularly negative ones.  Includes computer programs.   Requirements for protection Information must have been “secret” initially – some courts add novelty requirement plaintiff must have made reasonable efforts to keep it secret the information must be commercially valuable   Requirements for liability Breach of confidence Confidential relationship Reliance on commercial custom and tacit understandings secret was discovered through improper means e.g. overflights, fraudulent misrepresentations; phone taps reverse engineering is permissible   Copyright law Entitlements Reproduction Derivative works First distribution Public performance Public display exceptions Fair use Merger (ideas may not be protected – only expressions of it) Essential or archival copying Recounts Apple v. Franklin (CA3 1983)                         aid had to copy Apple OS in order to run apple compatible sw             Trips agreement extensds             Article 10 – Source and object are protected by copyright             Other articles make more explicit terms and exceptions               Has been suggested American fair use agreement violates article 13 Dimensions of diminishing copyright protection   Q– does TRIPS require every country to honor every court decision? A – imposes minimum levels of protection – not a single copyright law (still other stds country by country)   Why is it declining? Copying non-literal features of programs (read a book, write another mimicking a book) How similar? Whelan (1986 CA3) – improvement of existing work too close Altai – really creates expansive 3 part test                                                                i.      Abstraction (map features comparison)                                                              ii.      Filtration (remove from map unprotectable functions)                                                             iii.      Comparison (what is left)  if parallelism is tight – infringement                                                            iv.      Filtration process makes it harder for copyright owners to prevail reverse engineering for the purpose of interoperability Sega (CA9 1992) – accolade reverse engineered code to get games to run – sega brought suit – excused as fair use – done for interoperability menu hierarchies mimic commands – aka Lotus/MS/borland ineffective enforcement Analyzed BSA numbers – he found correlation to low piracy numbers and high Gore votes   Patent   Cases asserting patents in sw Gottschalk (1972) – can’t patent algorithm Parker (1978) – same thing Diamond (1981) – if embedded in a machine or chemical process whole things is protected Federal circuit relaxation of the test                                                                i.      Step by step US Patent activity (2001) 12,000 sw patents trend is up Doctrines by tech sector Utility – easiest Novelty – higher hurdle Obviousness – higher (sw is raised) Disclosure not as high (sw is lowered)   Contract   shrink wrap license prohibits fair use, reverse engineering etc. click on licenses Copyright law should win over contract – but, interpretations are muddled. So, contract law is the really bad animal. The real hazard is here. comment [] 8:24:46 PM

Code Debate Can free sw exist with proprietary sw Jason Maslou? (Shared resource mgr) Larry Lessig   On the surface there is coexistence in the eco-system –tension is whether free sw will survive   You gots             Free  -- open – shared – proprietary   So lessig is taking free and open vs proprietary   Advantages Zero price Good code Teaching others Strategic behavior  But destroy IP?       Open Free Shared Proprietary Free Y Y M Y Good Y Y Y Y Teach All all Some None Strategy Unlikely Never Y,hard   Destroy N ? ? N   Both free sw and proprietary impose a condition on a user (copyleft)   Proprietary sw imposes condition of no right to modify   Microsoft has acknowledged open/free is competitive threat   From microsoft’s perspective opposing open is good   From govt’s perspective – increased interesting in requirements for open software for obvious reasons   Maintaining a rich ecology keeps options open, assures competition   Microsoft response – government should not support the GPL part of the ecosystem   Bks                              Microsoft                                             Answer primary stimulus for innovation is IP protection  Not only stimulus private sector foreclosed                                               Not all (e.g. IBM) commercial application foreclosed                                 (only MS)   Jason   (first he gives larry a dark cape and light sword)   There is a lot of attention drawn to MS when you talk of windows v linux. Or other product level discussion.   Not talking about that.  Shared source is thinking about issues surrounding Industry, IP, economics, policy…   Cycle of innovation within software community including govt research, academic, industry, customers with shared knowledge   TCP/IP came out of this cycle.  I disagree with Larry that internet came about bks of tcp/ip and architecture – in truth it is because of production of cheap hw and communication techniques   There is no right way to create sw Combination of models has been critical factor for past 30 years   The bulk of open source sw is being commercialized.  That is where ballmer’s comments are focused.   There is an industry move to the middle.   IBM does proprietary and non-proprietary (Notes vs. Linux)  and other examples   Did model of software investment – shows that innovation as a competitive differentiation shrinks when shifting to open source.   Why is source code important in debate             Transparencey increases trust -         ~60% say source access is critical for view/modify -         <5% will view system source -         <1% will modify system source   MS is sharing source code with customers, partners and governments worldwide   Nesson   Let’s focus the debate by each of you taking one minute to state question you would like the other to address:   Larry (w/o light saber but liking the cape)   Not a lot of difference of what we said.  Transparency is the word I want to focus on – is Microsoft being transparent?    Jason   We have been extremely transparent on feelings for gpl – we have said we don’t like the license and have told on the web site what we don’t like about it.    Relatively to govt – MS does not believe govt should place procurement preferences on development model of sw – rather the value they receive in the software  they should choose on the merits of technology.   We never advocated DOD not use open software – just urged they carefully review the license of gpl   Larry   Is it just money govt’s can consider   Jason No.  Cost is just one piece of the proposal.  Your decision of what is valuable to you is not just monetary.   Larry   So if developing country decided that open source is better for their country –   Jason   I agree with you up to the point that should be legislated ….   [4926]   …   Jason   Is direct commercialization bad?   Lessig   No.  Are software patents bad?  Yes – my complaint is the awfulness of the patent process as it applies to software   Jason Lessig   Brad Smith and I have talked about imagining a better software patent system – and that would be good.  Software patents are bad now because existing system is so poor.    …. comment [] 6:46:17 PM

Free software and commons-based peer-production   Yochai     Free Software Characteristics                                                                i.      Apache has 60+ percent                                                              ii.      Linux  - 30% of IS market                                                             iii.      Proprietary SW 1.      use permitted in exchange for $$$ 2.      Learning often prevented to prevent copying/competition 3.      customization usually only within controlled parameters 4.      no redistribution permitted so as to enable collection by owner                                                            iv.      Free SW 1.      limit’s owners control 2.      use for any purpose 3.      study source code 4.      adapt for own use 5.      redistribute copies 6.      make and distribute modifications a.       notification of changes b.      copyleft 7.      identifying characteristic is cluster of user permitted, not absence of price (not talking free beer here)                                                              v.      anatomy of Free SW 1.      Raymond, Moody 2.      one ore more programmers write sw and release It on the net 3.      others use, modify extend or test 4.      mechanism for communicating, identifying and incorporating additions/patches into a common version 5.      volunteers with different levels of commitment and influence on testing, fixing, and extending   Institutional framework                                                                i.      Property 1.      property is institutional core of market and firm-based production a.       parameters of exclusion permit charging a price and controlling output of employees. 2.      public domain/open access a.       dedication to domain makes sw free b.      allows anyone to use, modify and redistribute c.       weakness: ease of defection/reappropriation by downstream actors may cause demoralization and ex ante non-participation by peers 3.      copyleft a.       cluster of licensing provisions that rely on the control property rights provide to make software “free” while protecting against some defections that an open-access commons approach permits b.      distributions is under same terms as original work was licensed c.       clear notifications of changes and attribution d.      covenants run with the program                                                                                                                                        i.      if you break, you lose the license e.       gpl and open source definition do not discriminate between commercial and non-commercial sw f.        Major current questions: what counts as “modification” 4.      Ads vs disads of copyleft v public domain a.       Reduces incentives to adopt a proprietary strategy b.      Reduces opportunities for “defection” c.       Retains the integrity of contributions as part of the peer-review process. Commons-based peer-production Peer production                                                                i.      Various sized collections of individuals                                                              ii.      Effectively produce information goods                                                             iii.      Without price signals or managerial commands Human parallel to distributed computing?                                                                i.      Various @home projects                                                              ii.      Gnutella, freenet Examples                                                                i.      Academic research 1.      we select our projects 2.      build on what others have done 3.      distribute as widely as possible                                                              ii.      wikipedia                                                             iii.      mars crater analysis                                                            iv.      http://slashdot.org                                                              v.      google 1.      they build web sites by figuring out what people link to…                                                            vi.      distribution 1.      gnutella 2.      distributed proofreading                                                           vii.        Economic analysis Motivation                                                                i.      OSS economics literature maps the diverse appropriation mechanisms 1.      intrinsic (make me better off) a.       hedonic b.      community ethics 2.      extrinsic a.       supply-side – human capital, reputation b.      demand-side – service contracts, widgets                                                              ii.      diverse motivations 1.      care for different things other than money                                                             iii.      initial implications 1.      where component contributions are too fine grained to transact around, peer production dominates (too small to do for money, great to do for greater good) 2.      discussion made along lines of formula                                                           iv.      peer production limited not by total cost or complexity of project, but by 1.      modularity (how man can participate) 2.      granularity (min investment necessary) 3.      cost of integration                                                             v.      value in economic terms 1.      human capital difficult to specify/define (can’t price it) 2.      by comparison to firms and markets peer-production has a.      information gains                                                                                                                                       i.      human capital highly variable 1.      time, task, mood, context, raw information materials. Project                                                                                                                                     ii.      difficult to specify completely for either market or hierarchy control                                                                                                                                   iii.      in peer-production agents self-identity for, and self-define tasks b.      allocation gains                                                                                                                                       i.      best agent on optimum resources                                                           vi.      Commons problems 1.      different kinds of commons have different solutions 2.      information only a provisioning problem, not an allocation problem 3.      primary concerns a.      defection through unilateral appropriate undermines intrinsic and extrinsic motivations b.      poor judgment of participants c.       providing the integration function 4.      solutions a.      formal rules b.      tech like slash c.       moral/social norms d.      redundancy or avging out e.       iterative peer production of integration f.        reintroduction of market and hierarchy with low cost and no residual appropariate Business models Surfers                                                                i.      Cost reduction and improved quality 1.      google 2.      live 365 3.      ibm, hp widgets                                                              ii.      translation into the price system 1.      services/customization/massification toolmakers                                                                i.      sourceforge, OSDN                                                              ii.      massive multiplayer games thematic analysis diverse motivations with complex relationships to money peer-production, not OSS anti-defection mechanisms                                                                i.      formal rules, tech restraints, social norms business opportunities                                                                i.      integration without residual appropriation                                                              ii.      surfers and toolmakers comment [] 3:41:49 PM

Business Method Patents Revised plan: Fisher to talk on Patent law for first hour.   Most other countries in world are hestitating to follow American path on these patents.   5 million patents in world, 82% held in US, Japan and Europe   Will describe pertinent features of American law with some splash of differences found in Europe.   Utility Patents – 95% of US patents Founded on section 101 – Inventions patentable Discovery useful process or product (there is a process patent on a cat exerciser – wiggling a laser point in front of a cat) Also describes product-by-process patent (a fine sub-group of products) Plant Patents Eli Lilly explores effect of Rosy Periwinkle on diabetes Ineffective there – but effective on leukemia Increases remission rate on childhood leukemia Doesn’t get patents – as property Another extension Plant Protection Act of 1930 Plant Variety Protection Act of 1970 1981 case (US Sup Ct) – said animal developed to eat oil – is patentable 2001 Pioneer Hi-Bred – now patentable under general utility patents One backing away Nuclear weapons were not patentable   Design Patents   Example of how extension process workes             Surgical procedures ·        tradition o       Can patent drugs o       diagnostic and surgical procedures not patentable ·        change o       Patent office started patents o       New technique for cataract surgery o       Doctor discovered non-linear cuts in line with shape of eyeball (ascertained optimal radius of curve) and sought patent protection o       He asked for fee to educate others §         But doctors are immunized for liability, device makers liable (?? Need to rework) Business Method Patents ·        Most of 20th century, couldn’t patent business methods ·        1990’s PTO starts granting BMP ·        State Street Bank case challenged patentability of bm o       Pool mutual funds into partnership for tax benefits o       Have to track relative magnitude of shares o       So, sw program relatively tracked value o       And got patent o       Competitor sought declaratory judgement action §         Court of Appeals ·        Machine calculation as application of algorithm is patentable ·        Repudiated business-methods exception to patentability §         Thus, surge in patents §         Drop off in rate since 2001 o       Famous patents §         Priceline §         Single click shopping ·        Amazon v. Barnes & Noble §         Behavioral programming ·        i.e. customizing web experience based on prior use – targeted ads ·        Backlash o       Bounties §         Bountyquest – rewards on prior art o       Litigation §         Patent Requirements ·        Subject-matter coverage ·        Novel ·        Non-obviousness ·        Utility