As noted by another poster, there are a limited number of ways to code a solution in a given language. Experienced coders will often come up with the same or similar results. Variations in white space, layout, variable names and comments would be expected. The "side by side" example looks like a clear "cut and paste" kind of coding where source code to the necessary solution was available and the snippet was copied to save typing. It's a common practice, but most folks are not laboring under concerns that the end product will end up under scrutiny in a court room.
It's been a long time since that's been true.
The original implementation of C++ called CFront, written by the creator of C++ itself, was just a preprocessor that translated C++ code into C. But as C++ grew in complexity (specifically when they added exception support!) C++ compilers went native.
Software is a real can of worms..
There are only so many ways to skin the cat, so to speak.
No sense in rediscovering that which is already tried and true..
but then, what would patent lawyers have to do?
Let a jury decide? Yeah, that’s the ticket.
but recall that C++ ultimately gets compiled down to C then to assembly.
not necessarily so. maybe other compilers do, but not MS.
.Net languages compile to IL to be parsed by the JIT assembler emitting native code. Non .Net is compiled
directly to native.
Developers at MS are fully aware to never introduce Open Source into MS product code. Its a serious offence. Not sure what other corps do along these lines.
The author of the article needs to have every claim treated with a huge bag of salt. He is a hack for Oracle.
That being said - example 1) You can’t copyright an API! Oracle can with that all it wants - but it is a loosing battle. It’s been through the courts before, and doesn’t wash.
An API is a specification on how to talk to a functional unit. It is NOT the functional unit itself. Implementation of that API is the functional piece of software that can be copyrighted.
This is right up there with SCO trying to claim copyright ownership of error.h from the C world. It didn’t work for SCO and won’t work for Oracle.
Further - there is real power in Google’s argument about being a trivial piece of code out of millions of lines of code. They are admitting that there are something like 9 places where they may have copied a few lines of code (this example given is of that magnitude.) 9 places constituting around maybe 100 lines of code out of multiple millions makes it trivial. No way Oracle can call that a derivative work.
Lastly -as others have pointed out. There are many places where there is only one reasonable way to code something. Implementing an API puts you in a further straight-jacket.
Ignore anything that Florian Mueller publishes -because it is going to contain a very slanted point of view. I would suggest following Groklaw.net if you want to see a more accurate discussion of the blow-by-blow of the Google/Oracle lawsuit. They’ll also educate you on what the law is and what is important.