Sam,
Great stuff you ask about! Your post reminds me that when I first read the manual, the further I got into it the more often I re-read the first couple of pages that discuss proprietary, open architectures. For the carpetologists of the world, the concept takes a little time to thoroughly sink in.
The theory you are asking that we test is if a company's source code isn't revealed, it's not possible for their software to be open. Remembering that I'm not a techie, my understanding is that we're not talking about the source code. Instead, we're talking about the stuff that allows one software program to work with another. In other words, we're talking about the interfaces, not the source code.
From page 48 and 52 of the original and revised manual, respectively, "It is open if its interfaces are published and other vendors are encouraged to integrate their products with the gorilla prodcut to create a whole product for a target customer." On the next page in both versions, we see that an architecture is closed if "only licensed vendors can build products using the architecture. ... If an architecture is open, on the other hand, its protocols are published, and any vendor who chooses can uild products to it specifications. This is how Intel, Mcrosoft, and Oracle operate."
I'd like someone like Erick to tell me if I'm right or wrong about the following analogy. I decide to build an engine. It's got my own technology which I own, making it proprietary. I publish some conncectivity specifications. If vendors build their related products (such as a set of fan blades, a lawnmower blade, or an automobile drive assembly) to my specifications, they will be sucessfully powered by my engine. In so doing, I've handed out to everyone for free the specs that allow vendors' products to work with my product, but I don't give them the information that allows them to improve upon my own product.
--Mike Buckley |
