free as in free by nrk14057

VIEWS: 8 PAGES: 40

									free as in freedom
         silvio lemos meira
           www.meira.com
       silvio@cesar.org.br
free as in freedom: o livro
oss: vem aí mesmo?
   Open source software has the potential to
    fundamentally change the economics of the computer
    software industry. It will affect how traditional software
    firms structure the incentives that they provide to their
    programmers, the product markets in which they
    compete, and the business models and strategies that
    they use. It will also impact on the career choices
    of software developers, and provides an
    additional means for potential programmers to
    enter the software industry.
       The Economics of Open Source Software: A Survey
        of the Early Literature, Aaron Schiff, Review of
        Network Economics Vol. 1, Issue 1 – March 2002
free software: definição
http://www.fsf.org/philosophy/free-sw.html


   a liberdade de rodar o programa, para qualquer
    propósito (freedom 0)
   a liberdade de estudar como o programa
    funciona e de adaptá-lo para suas necessidades
    (freedom 1)
   a liberdade de redistribuir cópias do programa,
    de tal forma a poder ajudar outros (freedom 2)
   a liberdade de melhorar o programa e publicar
    suas melhorias, de forma a beneficiar toda a
    comunidade (freedom 3)
f/l/oss classification
http://www.infonomics.nl/FLOSS/index.htm


Type                   Free   Redist   Unlimited      Src.      Src.       Public      Viral
                                         Use         Avail.     Mod.       Check
                                                                            Ins
Commercial
Trial SW                x
Shareware               x       x
Royalty Free Bin.       x       x          x
Royalty Free Lib.       x       x          x            x
Open Src (BSD)          x       x          x            x         x
Open Src (Apache)       x       x          x            x         x           x
Open Src (Linux/GNU)    x       x          x            x         x           x          x

                                        Source: Vinod Valloppillil / The Halloween Document
de onde veio isso?
steffen evers
http://user.cs.tu-berlin.de/~tron/opensource/opensource.html
   AT&T, 'Sherman Antitrust Act„, 1949.
       'consent decree' em 1956… “prohibited AT&T to start any other business than
        telephone or telegraph services”
   Unix… policy:
       license the software (allowed by the decree) but not to pursue software as a
        business.
           ``no advertising
           no support
           no bug fixes
           payment in advance''
   sem SUPORTE e correção de BUGS,
    a comunidade TINHA QUE SE AJUDAR…
       …they started “to share with one another. They shared ideas, information,
        programs, bug fixes, and hardware fixes.” User groups were created
        where ever Unix was introduced. Among them were universities in
        many other countries than the USA like Australia, United Kingdom,
        Germany, Japan. “At ten years of age, Unix was genuinely being
        used worldwide.”
isso funciona?

SIM
em que contextos?
        em qualquer um onde seja possível
         incentivar uma comunidade aberta de
         desenvolvedores competentes…
    1.    onde haja um “inimigo natural”…
    2.    ou necessidades não satisfeitas…
    3.    onde não interessa para o capital…
    4.    ou onde INTERESSA AO CAPITAL!…
        em especial, 1-3 é TRADIÇÃO em UNIX
         e muito antes de linux torvalds nascer…
mas há restrições:
a lei de LOTKA
   The number of authors making n
    contributions is about 1/na of those making
    one contribution, where a is often nearly 2.

       Alfred J. Lotka,
        The Frequency
        Distribution of
        Scientific Productivity,
        Journal of the Wash.
        Academy of Sciences,
        16(12):317-323, 1926.
software LIVRE interessa…
ao CAPITAL?




até onde eu sei,
OPEN SOURCE SOFTWARE
não é uma nova forma de comunismo
por que funciona?
{do ponto de vista econômico}
   ninguém sabe direito
       ou tem uma teoria completa para explicar…
       mas as primeiras evidências estão aparecendo
           free sw como “public goods”…




    The Economics of Open Source Software for a Competitive Firm, Richard Hawkins
o caso da ibm & apache
apache e “o resto”
free sw e o
“mundo capitalista”
   empresas contribuem para projeto(s) open
    source de sistemas operacionais
       mercado de plataformas
           windows
           o “resto”
   pontos de vista…
       para entender a indústria
       e as motivações
       sabendo que, para o usuário, são as aplicações
        que interessam: plataformas são infra-estrutura!
teoria e conseqüências
   plataformas e aplicações SÃO regimes de
    compatibilidade (como carros e estradas)
   os efeitos de REDE valem e há retornos crescentes
    (realimentação positiva) em vários níveis
           desenvolvedores de aplicações
           varejo de software
           usuários
   as tecnologias competem e há efeitos de lock-in
       …“By taking control of a standard and making it proprietary
        [i.e. Windows-specific], Microsoft can design the standard to
        reduce rather than increase interoperability….”
            http://econwpa.wustl.edu/eps/le/papers/0103/0103003.pdf
teoria e conseqüências
   o “resto” se transforma em nicho quando o
    número de aplicações diminui (apple…) E
    corre sério risco
    de desaparecer!...
   para COMPETIR,
    têm que CRIAR
    uma alternativa
    SÉRIA!
Open Source Software (OSS)
é uma tal alternativa
   é muito difícil “seqüestrar” um projeto
   combinação de esforços
    = desenvolvimento rápido
   empresas têm incentivos
       criar plataforma alternativa e atrair aplicações
       desenvolver aplicações e
        atrair clientes e usuários
       influenciar no desenvolvimento da plataforma
       liberdade para desenvolver hardware
        inovador/NOVO para a plataforma
Most popular OSS software
{berlecon research, europe, 2002}
 Server operating system

    1. Linux (78% of those using OSS SOS), 2. Free/Open BSD (9%)
 Databases

    1. MySQL (54%), 2. SAP-DB (12%?), 3. PostgreSQL (11%)
 Web site creation and operation

    1. Apache (72%), 2. Perl (48%), 3. PHP (32%), 4. Squid (26%),
     5. OS CMS (10%), 6. Python (8%)
 Desktops

    1. Linux (49%), 2. StarOffice/OpenOffice (33%), 3. KDE (31%),
     4. Gnome (23%), 5. Mozilla (23%).
  Benefits from OSS usage -server OS
  {berlecon research, europe, 2002}
Importance of criterion for decision in
favour of OSS and against proprietory
software
Security in Open versus Closed Systems
Ross Anderson, Cambridge University {2002}

   The debate about open versus closed systems started out in
    the nineteenth century when Auguste Kerckhoffs pointed out
    the wisdom of assuming that the enemy knew one's cipher
    system, so that security could only reside in the key.
   It has developed into a debate about whether access to the
    source code of a software product is of more help to the
    defence, because they can find and fix bugs more easily, or
    to attackers, because they can develop exploits with less
    effort.
   This paper answers that question. In a perfect world, and for
    systems large and complex enough for statistical methods to
    apply, the attack and the defence are helped equally.
    Whether systems are open or closed makes no difference in
    the long run.
         sim…
  mas como funciona
      a engenharia de
f/l/oss development?
f/l/ossD vs sw engineering
walt scacchi
    First, SE is a team endeavor that is focused on the development of large
     software systems through a software development life cycle.
    Second, the software life cycle (model) constitutes a framework that directs or
     orders the processes of SE that every software development project should
     traverse.
    Third, the focal processes of SE include software requirements engineering,
     specification and prototyping, design (functional, architectural, modular, or
     object-oriented), testing (verification and validation), configuration management,
     maintenance (or evolution), and project management.
    Fourth, these processes may or should employ formal notations and reasoning
     schemes for consistency and completeness, though which computer-based tool
     to use to support such notations and schemes is unclear.
    Fifth, software quality results from the systematic performance of software life
     cycle processes that create, reuse, manipulate, or update software artifacts
     (including formal notations, graphic diagrams, and source code), according to
     project planning, cost estimation, and management control efforts.
    Sixth, the level, goal, or threshold of software quality (e.g., end-user
     satisfaction, number of defects discovered post delivery) that is sought or
     attained determines the level of software productivity that is achieved, as well
     as the overall cost of the software development effort.
ciclo “profissional”/“empresa”
de vida de software




                    http://www.sei.cmu.edu/staff/wjh/16CSPOval.GIF
How is F/OSSD faster than SE?
   Internet time and F/OSSD projects also tend to
    produce incremental software releases at a much
    faster rate, even to the point of releasing unstable but
    operational daily system builds. This denotes not only
    a reduction in product release cycle times compared
    to SE practice, but also a significantly restructured life
    cycle process and process cycle time reduction. Thus,
    many large F/OSSD projects have demonstrated a
    sustained ability to produce and deliver complex
    software systems that grow in a manner faster than
    attributed to traditional SE practices that focus
    attention to project development schedules and
    management control.
How is F/OSSD faster than SE?...

    Many of the largest and most popular F/OSS systems like the Linux
     Kernel, GNU/Linux distributions, GNOME user interface and others are
     growing at a super-linear (exponential) rate, as is their internal
     architectural complexity…
    As OSS developers are themselves often end-users of their systems,
     then software requirements and design take less time to articulate and
     negotiate, compared to SE projects. In the tradition of SE, developers
     are not expected to be users of the systems they develop. As such,
     they must elicit requirements and validate system design with end-
     users who are generally not SE professionals, and thus must negotiate
     what they will be able to do, on what schedule and budget, and with
     what staff resources. Thus for F/OSSD projects, the time it takes to
     determine required system functionality is shorter, and often less
     demanding that expected in SE projects.
         sim…
mas… quais são os
    problemas???
On the Nonmaintainability of Open-Source Software
Stephen R. Schach, Vanderbilt University
A. Jefferson Offutt, George Mason University {2002/3}


     We downloaded 365 versions of Linux. For each version in turn, we
      examined the 17 kernel modules and counted the number of lines of
      code in each module. Then we counted the number of instances of
      common (global) coupling between each of the kernel modules and all
      the other modules in that version of Linux.
     We obtained two primary results. First, we found a linear dependency
      between lines of code and version number that is significant at the
      99.99% level; Second, we found that the number of instances of
      common coupling grows exponentially with version number. This result,
      too, is significant at the 99.99% level. That is, the exponential growth in
      common coupling is again an inherent feature of successive versions of
      Linux.
      We previously related common coupling to fault-proneness.
      Consequently, combining our two results reveals a worrying trend.
do ponto de vista capitalista
   faz sentido tratar o problema como um de
    competição tecnológica e…
   a plataforma OSS é uma oportunidade ÚNICA
    para que se crie, de fato, no longo prazo, uma
    alternativa SÉRIA a windows… e
   ISSO É BOM pra microsoft!
       mantém a empresa de redmond paranóica…
       beneficia seus parceiros…
       clientes e usuários!
adoção por governos
bens econômicos:
categorias
   bens privados:
       aqueles para os quais há rivalidade e exclusão
        (o meu NÃO é seu e TÊM preço)
           maçã, sorvete, automóvel
   bens públicos:
       para estes não há exclusão nem rivalidade
        (o meu É seu e NÃO TÊM preço)
           defesa nacional, conhecimento,…
           pergunta: ensino superior DEVERIA estar aqui?...
bens econômicos:
categorias
   bens comuns:
       há rivalidade mas não exclusão
        (o meu NÃO é seu, mas NÃO TÊM preço)
           peixe nos oceanos, o meio-ambiente,..
   monopólios naturais:
       há exclusão, mas não rivalidade
        (TÊM preço, mas o MEU é SEU)
           proteção contra incêndio, tv a cabo
bens PÚBLICOS…
   aqueles que estão disponíveis grátis
    para todos (=para os quais não há exclusão)
       tais bens têm um elemento de
        externalidade, porque algo DE VALOR não
        tem nenhum preço atribuído…
           i.e., pessoas recebem benefícios sem ter que
            compensar NINGUÉM por tornar disponível
            um recurso escasso e valioso…
           aí é onde os governos normalmente entram!
               defesa nacional é um exemplo…
 Government Preferences for Promoting Open-Source
 Software: A Solution in Search of a Problem
                                   BERNARD REDDY
                                    DAVID S. EVANS
                       NERA Economic Consulting - Cambridge Office
           Michigan Telecommunications and Technology Law Review, Forthcoming


                                       Abstract

Governments around the world are making or considering efforts to promote open-
       source software (typically produced by cooperatives of individuals) at the
         expense of proprietary software (generally sold by for-profit software
       developers). This article examines the economic basis for these kinds of
    government interventions in the market. It first provides some background on
      the software industry. The article discusses the industrial organization and
   performance of the proprietarysoftware business and describes how the open-
    source movement produces and distributes software. It then surveys current
      government proposals and initiatives to support open-source software and
    examines whether there is a significant market failure that would justify such
     intervention in the software industry. The article concludes that the software
   industry has performed remarkably well over the past 20 years in the absence
      of government intervention. There is no evidence of any significant market
       failures in the provision of commercial software and no evidence that the
  establishment of policy preferences in favor of open-source software on the part
                    of governments would increase consumer welfare.
                  http://papers.ssrn.com/sol3/papers.cfm?abstract_id=338564
A Solution in Search of…
   Compared with many other industries, the software
    industry is relatively unconcentrated. One
    conventional measure of industry concentration is the
    total share of sales accounted for by the four largest
    firms. In 2000, the four largest firms in the proprietary
    software industry accounted for 26.7 percent of total
    revenues. According to the latest Census data,
    nearly 47 percent of all manufacturing
    industries have a four-firm concentration ratio
    greater than that of the software industry.
A Solution in Search of…
   A second measure of industry concentration is the
    Herfindahl-Hirschman Index (HHI). HHIs can range
    from zero (a large number of firms with infinitesimal
    market shares) to 10,000 (a monopoly with 100
    percent of the market). In 2000, the HHI for the
    software industry was 244,47 a relatively low HHI
    when compared with industries such as automobiles
    (2,506) or breakfast cereals (2,446)
   There is also a great deal of turnover among the
    leading firms indicating that firms generally have
    not had entrenched positions in the software
    industry overall. Of the top ten companies in
    1990, five did not make the list in 2000…
a intervenção do governo no
setor, no longo prazo…
   vai ter que significado?...
       DRM (digital rights management)
       TCPA (trusted computing platform alliance)
       Palladium
           a microsoft tem que liberar o código porque é exigido
            por lei…
               mas e daí?
               ...
   será que algum governo vai conseguir
    seqüestrar algum open source project?...
olhando lá pra frente…
   é preciso REDISCUTIR o problema de
    propriedade intelectual na era da informação
       patentes e copyright podem não estar mais
        atendendo às necessidades do TODO
           o SISTEMA pode ter sido distorcido pelas PARTES
   LINUX é uma boa parte desta discussão e
       dentro de uma boa gama de aplicações…
           é muito mais barato do que a competição… desde que
            se tome os cuidados básicos que se tem que tomar
            com algo que é LIVRE como uma criança…
mas é preciso pensar. sempre.

people  demand freedom
 of speech to make up for
 the freedom of thought
 which they avoid
    as pessoas demandam liberdade de expressão
     para compensar a liberdade de pensamento que
     elas mesmo evitam.
        Soren Aabye Kierkegaard (1813-1855)
dever de casa: grupos de 4!
   descreva, usando de 10 a 12 mil caracteres,
    o processo de funcionamento de um projeto
    open source. para tal, descubra onde tais
    projetos são sediados, como o time é
    montado e gerenciado, como as discussões
    são realizadas, como as decisões são
    tomadas, como o processo é administrado…
    e como VOCÊ poderia fazer parte…
   entrega: próxima segunda, impresso!
free as in freedom
         silvio lemos meira
           www.meira.com
       silvio@cesar.org.br

								
To top