|
Authored by: jesse on Friday, April 20 2012 @ 08:24 AM EDT |
In this context, discussing the requirements between the caller (using the API),
and the library (implementing the API), the term "contract" is used as
an analog to the business contract.
The two parties involved are the caller, and the library. The contract specifies
how the caller (an application function) is to behave when using the API, and
how the library is to behave. It does not specify HOW either party accomplishes
the action. Violations of this "contract" are considered bugs, and can
result in undefined behavior of the application.
When all elements of the "contract" are satisfied (parameters valid,
exceptions handled) then the application is considered working properly.[ Reply to This | Parent | # ]
|
|
Authored by: Ian Al on Friday, April 20 2012 @ 10:12 AM EDT |
I understand it as a promise to deliver functionality if the offered rules are
followed. The programmer and the API implementer are the parties to the contract
and the contract is the usual abstract promise to deliver, but formalised as
written text.
---
Regards
Ian Al
Software Patents: It's the disclosed functions in the patent, stupid![ Reply to This | Parent | # ]
|
|
|
|
|