Source Code, Cultural Sources And Privacy

The GPL has a patent-style requirement of disclosure for the source code of binary executables that are publicly distributed.

If software was only written in interpreted languages, if tools existed to manipulate binary executables with the same semantic detail that source code can be edited with, or if computing machinery was programmed using switch combinations that were published in books, such a requirement would not be necessary. The distributed form of the software would be complete, it would be the software.

But compiling source into a binary executable, whatever its practical benefits, is a form of obfuscation and a means of locking software to a particular piece of hardware. For those who would modify or move software that they have received it is an inconvenience that is major enough to hamper or prevent their use of the software. This is why the GPL requires disclosure, to protect the freedom to use the software.

So requiring source for published software is not an intrusion into privacy, it is not coercion and it is not unfair economically. It is a neccessity. The source code for software is the software, it is the only complete representation of the software. Disclosure of it is a necessary measure to ensure that the software is actually distributed. Otherwise binary executables become the public tip of a private iceberg of code that can impact on the private or public life of those that use it.

Private development and use of software should be possible. But it is important not to confuse my right to develop software privately with a right to prevent you using that software privately when I publish it. Your private use of my software is public relative to me but this is only a matter of perspective. DRM and binary executables make private use of software impossible; they make your private use of my software a public use of it. Their effects should be nullified.

For cultural works, private development and use of work should be possible and supported by licenses. The Creative Commons licences do not allow for this. Possibly the practicalities of this and the effects of the minutiae of copyright law mean that this does not have much effect in practice, but it would be good to have the licences acknowledge both a private domain of use of culture and when this domain crosses over into the public sphere.

Technorati Tags: ,

Posted in Free Culture
7 comments on “Source Code, Cultural Sources And Privacy
  1. mako says:

    Nice comments Rob. 🙂

  2. Matt Lee says:

    “If it isn’t source, it isn’t software.” – Cygnus.

  3. A gift cannot be a burden or obligation for either receiver or giver.
    If I give you a binary file, that’s the end of the matter.
    There can be no further obligation on either person’s part. You didn’t commission the gift, and I didn’t contract to support the gift.
    You cannot compel me to give you more, such as the source. Nor can you constrain my use of what I still have.
    I cannot constrain you in your use of the gift. It is yours to use, publish, reproduce, derive, etc.
    You must consider the source a separate work that you may well covet, but have no right to.
    It is only copyright that gives people notions otherwise, as it provides the legal power to implement them. Unethical, and highly corrosive power.

  4. jer says:

    If I give you a binary file, that's the end of the matter.

    Except, as you even admit, when ‘higher’ principles are involved. Indeed I believe you claim that an individual cannot be compelled to release source unless the right to life is at stake.
    But how narrowly do you define ‘right to life’: Can only death compel disclosure? Immense suffering? Mild suffering? Deprivation of happiness? The definition makes all the difference doesn’t it?
    Without introducing degrees of rights violation your argument is unpersuasive. I think that your delineation and ordering of general classes of rights is too ‘neat’. It seems to me that, at least when lacking specificity, these classes rights are not comparable at all; only instances of these classes should be compared. For instance a ‘major’ violation of liberty may outweigh a ‘minor’ privacy violation.
    Accordingly, I am not convinced that distributing binaries without source as a form of subjugation is ethical.

  5. Things can be just as bad whether citizens are prosecuted for enjoying cultural freedom, or prosecuted for failing to sufficiently document their software.
    Any society that needs stormtroopers to execute dawn raids against kids (whether for sharing published software or publishing derivatives without source) in order promote science, technology and the arts for the benefit of its citizens is in a handcart on the way to hell.
    Check out my comment here:
    Live and let live eh?
    Remove privilege and you can remove the force needed to enforce it.
    If you want someone’s source code, offer something of equal value in exchange.
    Ask yourself why the heck you need stormtroopers to force people to provide source code to any binaries they publish.
    Try a free market instead of communism. It’s natural. No compulsion necessary.

  6. Rob Myers says:

    My argument, as made in a post somewhere below, is that software *is* source code. If binaries could be edited as easily as source, or software was all written in interpreted languages, this would be a non-issue. But as it is not, binaries are a means of making my private use of software public; this is surely wrong by your own ethics.

  7. It is semantic sophistry to justify a privacy violation based on some functional equivalence or common classification between human readable source code and machine readable binary.
    If they’re equivalent, the binary is as good as the source – so you don’t need the latter.
    If they’re not equivalent, then the binary is not as good as the source – so you need to exchange greater value for the latter.
    Machine code is ‘software’.
    Source code is ‘software’.
    It does not follow that ‘software *is* source’. Software is anything that can be read by a machine, whether source or binary.
    It does not follow that simply because works share the same classification that if an author publishes one that they forfeit privacy to every other.
    You may wish that if an MP3 is published that the digital master is forfeit.
    You may wish that if an orchestral work is publicly performed that the composer’s score is forfeit.
    Wishes do not constitute rights.
    If I put “LDA &0932; ADC #12; STA &0933” on a piece of paper with my name on it and post it on a public noticeboard, this does not give you a right to break my door down and search my house until you find a piece of paper with “b=a+12; // Add year” written on it. Nor can you expect the state to fine me until I produce something in a high level language that results in the same object code.
    Frankly, I’m astounded that anyone can consider themselves entitled to violate an author’s privacy and seize their unpublished work without even the offer of payment, simply because they feel entitled to know how a published work was produced.
    Do let me know how publishing a binary without source ‘is surely wrong by my own ethics’.
    Remember, we’re talking from a post-copyright abolition perspective and whether a law would be necessary to compel disclosure of source to published binaries.
    We’re not talking about whether it’s particularly heinous for contemporary copyleft licences to oblige publication of source.