[Diritto] link e decss

[Alessandro Rubini]

> Ne approfitto per ribadire come, secondo me, la scrittura di un
> programma è del tutto analoga alla creazione di un'opera letteraria,
> ovvero all'espressione di un'idea secondo un linguaggio formale, che
> si chiama C o Perl anzichè Francese o Italiano, ma che non cambia la
> natura dell'azione. Ed è su questa impostazione che dovrebbero fare
> perno le battaglie contro i brevetti sw, perchè sennò ci si va ad
> "imbrodare" in discorsi difficilmente difendibili.

Ma secondo gli oppositori anche questi sono difficilmente difendibili.
Da patents@liberte.aful.org:

    I suspect the case will be overturned.  The judges in the case were
    ignorant of many software engineering concepts, which renders their
    decision suspect.
    
    For example, they stated that their are no ideas disclosed in object
    code, which distinguishes object code from source code as speech.
    This is false.  There are entire conferences on reverse engineering
    object code and binary code into source code and system architectures
    - IEEE has an annual conference just on such tools.
    
    If they got this wrong (due to the court's lack of interest in getting
    its law clerks to investigate software engineering when ruling on
    software cases), who knows what else is wrong in their assumptions.

E poi:

    > That decision was right, there are no ideas disclosed in object codes.
    
    There are no ideas disclosed in gzipped files.
    
    There are no ideas disclosed in assembly language listings for a
    CPU whose assembly language I don't know.