One that includes why "ways <inventor> could never have imagined" is desirable so it would convince people who don't already believe it to be the case?
If they used the accelerometer to measure how hard they hit people over the head with iPad when mugging them, and triggered a string of copycat attacks, that wouldn't be desirable for Apple or society.
It's almost arguing against firewalls - why restrict computer access when developers can find so many new and interesting uses for it? (Buffer overflows, spam and more is why. Default-deny is the good practise alternative. Assuming that's my stance, your suggestion isn't a convincing argument at all).
Yes, and the "X can be used for bad" argument you put forward is one that could be used against any advance in technology, with or without open standards or APIs.
And your firewall example is not relevant. Hackers aren't using firewalls against the companies that employ them, they are circumventing the firewalls. What you said is more like "people can climb over fences easily, so we should stop building them around our yards"
Yes, and "X can be used for bad" would be as weak as "X can be used for good" is weak in favour of something.
My example is relevant because it directly counters edw519's implicit claim that "if an idea can be used in surprising ways, that is good" by citing open network connections as an example of something used in surprising ways which are bad, and our opposition to that by using a "not open" stance as the agreed best position.
To be convincing that openness is good, it would need to be established that significant good comes from it, or that more good than bad comes from it, or that at least serious bad cannot come from it or similar.
Saying that X can be used in previously unexpected ways is neither here nor there as an argument for open APIs or collaboration, whether or not I like open APIs and collaboration.
I think that this argument is much stronger than your original one, and in fact, one that I would make myself. Although I still don't see how it helps your firewall example, because I can't understand how circumventing a firewall can be considered "using it in a surprising way."
Actually, on rereading your original comment, I think I misunderstood that paragraph. The exact wording is still very confusing to me, but what I think you were trying to say is that edw519's implicit argument (for all X, X should be open because it can be used in surprising ways) can be applied to network connections, where network connections are the X. I thought you meant that firewalls were the X, and that idea is what I was arguing against.
If they used the accelerometer to measure how hard they hit people over the head with iPad when mugging them, and triggered a string of copycat attacks, that wouldn't be desirable for Apple or society.
It's almost arguing against firewalls - why restrict computer access when developers can find so many new and interesting uses for it? (Buffer overflows, spam and more is why. Default-deny is the good practise alternative. Assuming that's my stance, your suggestion isn't a convincing argument at all).