4 ms·

Bang-on. But it's not just startups, it's companies - period. Stallman, love him or hate him is correct - in these cases, you're the product. If the source isn

by cik 12y ago

Bang-on. But it's not just startups, it's companies - period. Stallman, love him or hate him is correct - in these cases, you're the product. If the source isn't open, or at least if enough of it isn't open, you can't trust the company. My favourite, most gregarious example is Privacy Polices. Sure, in certain jurisdictions (i.e. mine) they're legal documents. But equally, a company could bury in their terms of service something along the lines of 'We, the organization, to the best of our knowledge abide by the Privacy Policy. In the case where this is untrue, our liability is limited to XXX'.

bunderbunder 12y ago
Open source doesn't do anything to help with this kind of stuff. Not one thing. Suppose, for the sake of argument, that Facebook released the full code of their social networking application under the GPL. That promises us very little in terms of our privacy, because they could easily (and perhaps should, from an engineering standpoint) stick all the privacy violation in a separate application for doing the analytics or providing data to 3rd parties. Open source doesn't even do much of anything to limit what kinds of data can be collected through the front end. The original owner of the code retains copyright, so open source licenses place no legal restrictions on what they can do with the code. There's nothing stopping an unscrupulous company from sharing a squeaky-clean version of the source with the world at large, and running different code on their servers. If anything, this is where I find Stallman to be at his least useful. He's only got this one hammer in a world where the vast majority of the most worrisome ethical concerns are not nails. His dogged insistence on trying to turn every moral and political issue in computing into one that can be answered by open source[1] is neither realistic nor practical, and only serves to illustrate that his ideas all became fixed long before Berners-Lee came along and changed everything. [1]: For example: http://www.gnu.org/philosophy/who-does-that-server-really-serve.html http://www.gnu.org/philosophy/who-does-that-server-really-se...
- krapp 12y ago
  Exactly. That model - that open source is more trustworthy - was meant to apply to code you could read then compile from source on your machine. But the web is entirely a network of black-box applications you (legally) interact with in a passive sense - send request, get response. You really have no idea what's going on in the back-end, and no control over what they choose to do with what you put on their servers. And you definitely have no way to verify the code, apart from headers and responses.
- Jedd 12y ago
  Which version of GPL? cik was perhaps misleading when he referenced rms but then talked about 'open source' (rather than 'free software'). I note you continue with the phrase 'open source', to argue against rms's position (which is very much anti 'open source' both as an appellation and concept). It suggests, if you'll excuse my bluntness, that you don't quite get it. You also seem to assume a very narrow use case of that FB source code (were it released to public scrutiny).
  - bunderbunder 12y ago
    I used the broader category because in these kinds of situations Free Software brings nothing to the party. Free Software licensing only places restrictions on what licensees of the software can do with it. The original copyright holder, by virtue of possessing the original copyright, is not subject to those restrictions. That's why dual Free/commercial licensing is such a common choice for companies who are looking to monetize - it's an easy way to maintain commercial control over a product while still getting to call yourself "open".
- iammyIP 12y ago
  RMS is not about OpenSource, he is about FreeSoftware, as in Software that respects its users freedom. Its an ethical mindset for programmers, not a method of distribution. You could easily program and distribute binary blobs for money as FreeSoftware, but only you would know it then. The problem is trust. Can you be trusted to respect your users freedom? That is the point where OpenSource might help to build trust, but it is not mandatory.

by cik 12y ago

bunderbunder 12y ago

Open source doesn't do anything to help with this kind of stuff. Not one thing. Suppose, for the sake of argument, that Facebook released the full code of their social networking application under the GPL. That promises us very little in terms of our privacy, because they could easily (and perhaps should, from an engineering standpoint) stick all the privacy violation in a separate application for doing the analytics or providing data to 3rd parties. Open source doesn't even do much of anything to limit what kinds of data can be collected through the front end. The original owner of the code retains copyright, so open source licenses place no legal restrictions on what they can do with the code. There's nothing stopping an unscrupulous company from sharing a squeaky-clean version of the source with the world at large, and running different code on their servers. If anything, this is where I find Stallman to be at his least useful. He's only got this one hammer in a world where the vast majority of the most worrisome ethical concerns are not nails. His dogged insistence on trying to turn every moral and political issue in computing into one that can be answered by open source[1] is neither realistic nor practical, and only serves to illustrate that his ideas all became fixed long before Berners-Lee came along and changed everything. [1]: For example: http://www.gnu.org/philosophy/who-does-that-server-really-serve.html http://www.gnu.org/philosophy/who-does-that-server-really-se...

krapp 12y ago
Exactly. That model - that open source is more trustworthy - was meant to apply to code you could read then compile from source on your machine. But the web is entirely a network of black-box applications you (legally) interact with in a passive sense - send request, get response. You really have no idea what's going on in the back-end, and no control over what they choose to do with what you put on their servers. And you definitely have no way to verify the code, apart from headers and responses.
Jedd 12y ago
Which version of GPL? cik was perhaps misleading when he referenced rms but then talked about 'open source' (rather than 'free software'). I note you continue with the phrase 'open source', to argue against rms's position (which is very much anti 'open source' both as an appellation and concept). It suggests, if you'll excuse my bluntness, that you don't quite get it. You also seem to assume a very narrow use case of that FB source code (were it released to public scrutiny).
- bunderbunder 12y ago
  I used the broader category because in these kinds of situations Free Software brings nothing to the party. Free Software licensing only places restrictions on what licensees of the software can do with it. The original copyright holder, by virtue of possessing the original copyright, is not subject to those restrictions. That's why dual Free/commercial licensing is such a common choice for companies who are looking to monetize - it's an easy way to maintain commercial control over a product while still getting to call yourself "open".
iammyIP 12y ago
RMS is not about OpenSource, he is about FreeSoftware, as in Software that respects its users freedom. Its an ethical mindset for programmers, not a method of distribution. You could easily program and distribute binary blobs for money as FreeSoftware, but only you would know it then. The problem is trust. Can you be trusted to respect your users freedom? That is the point where OpenSource might help to build trust, but it is not mandatory.