|
Authored by: Ed L. on Saturday, April 21 2012 @ 10:07 AM EDT |
And, also, an API in programming terms *is* a header
file.
Really? For C, perhaps. Part of a header file, at
least. But C++?? Or Java???.
Seriously. We'll have to do better
than that. To repeat an earlier comment,
Here now we have Judge
Alsup tasked with determining whether or not an "API" may be protected by
copyright. So the judge needs a legal definition of what is an API, both in the
specific sense of a Java API, and if possible, in a general sense that might be
misapplied to all programming languages, past, present, and as yet
unnamed.
Sounds so far as if Google would like a very narrow definition e.g.
"A Java class API *is* the class interface", whereas Oracle would like something
more vague and general and creative. In particular something involving
documentation.
Judge Alsup's decision on this one will reach far beyond
Java, so its important he formulates a "What is an API" definition that meets
with general industry and academic consensus.
This is indeed
murky. Hence the Judge's request for additional briefs. What should they
entail?
--- Real Programmers mangle their own memory. [ Reply to This | Parent | # ]
|
|
Authored by: vonbrand on Saturday, April 21 2012 @ 10:16 PM EDT |
I happen to think that the "plane cockpit" example given to PJ is right on
the money on APIs. It is the "naming" (in the cockpit case, the form and color
of the buttons, the shape and markings on the dials, the color of the
blinkenlights) and their "meanings" (semantics; what happens if you press a
button or move a lever, what each dial shows in what units, what light is
lighted when). Sure, the overall layout is important too for familiarity, but
you could move stuff around and an experienced pilot shouldn't be too lost. In
the case of Java APIs the overall layout would be more or less the grouping of
classes in packages. [ Reply to This | Parent | # ]
|
|
|
|
|