4 ms·
I would contend that the definition of the word "interface" is, in essence, interoperability. > a point where two systems, subjects, organizations, etc. meet a
by mieseratte 8y ago
I would contend that the definition of the word "interface" is, in essence, interoperability.
> a point where two systems, subjects, organizations, etc. meet and interact.
An interface is that which allows two systems to work together, in conjunction. That is, interoperability:
> the ability of equipment or groups to operate in conjunction with each other.
Whether or not the intent of our legal code is to allow for drop-in-replacement interoperability of systems is something I have no domain knowledge with which to comment.
- garmaine 8y agoThat’s a pretty useless definition. In this context interoperability means being able to substitute one product for another because they have the same API. Sometimes this is the intent (e.g. SQL). Sometimes not.
- didibus 8y agoDo you feel licenses should address that distinction? So say, only APIs under certain licenses would allow alternate implementations of them?
- salawat 8y agoConsidering that an API is a mere symbolic/semantic gateway closer to a mathematical expression, I'm sorry, your interpretation would seem to be more useless. Nothing says the underlying functionality exposed by an API has to functionally line up. I could implement a persistence layer through Pair(K. V) or any other myriad sets of logic, but just because I use that set characters to render that API in a virtual machine, I should be open to extortion? Programming languages, and API's in particular are useless alone. This is just an attempt to try to monetize a gigantic segment, which if allowed will lead to an inevitable fracturing of just about every semantic convention behind an ever growing wall of lawfare. It's bloody math. Stop trying to treat it like a bloody Widget. For God's sake, you're saying Oracle should be able to gatekeep Pair(K key, V value). Great. So as long as anyone uses that Oracle gets to sue them? Great way to hobble the industry. Or are you seriously going to try to make the argument that all we need to do is change the package names/method signatures of the entire JDK without altering it structurally and we're gold? This is a space where legalese starts to enter its' own butthole without reconciling what you're trying to do with what you're actually doing.