Re: [Mingw-users] Yet another mingwrt/w32api provenance discussion [Was: Re: Is WSARecvMsg available in mingw?]

[Charles W. <cwi...@us...>]

On 11/21/2010 10:29 AM, Greg Chicares wrote:
> On 2010-11-21 13:14Z, Charles Wilson wrote:
>> Are the two samples so similar that it makes you doubt mingw64's claims
>> of independent re-implementation, or are they so different that you
>> believe it is obvious their re-implementation IS independent?
> 
> Comparing code found on the web that bears a microsoft copyright, to
> the initial revision in the mingw64 repository--except for comments and
> whitespace, the WspiapiLegacyGetAddrInfo() implementations

Implementations, or declarations?  I'm still confused why Tor's original
post showed a .c from mingw64 and a .h from the SDK.

> are plainly
> identical. I visually compared all eighty-seven nonblank lines in the
> function body, and didn't see a single token that differed.

Ah, sounds like implementation.

> That doesn't conclusively prove any violation. Perhaps both derive from
> some copyright-free original.

Or Person X documented not just the functionality, but also the public
and private token names -- and the data flow:

"To initialize variable X, use the function FooBar(), according to the
obvious meaning described in its msdn documentation.  Check the return
status, and SetError to WINFOOBAD on failure and return.  Then..."

It's very likely, given enough specificity in the documentation, that
Person Y, implementing the mingw64 function, could achieve a great deal
of similarity to the original implementation, without actually viewing
the code or violating copyright.

> I didn't compare code obtained directly
> from microsoft, because I don't want to accept their EULA. But there's
> a concern here that would need to be addressed before we could adopt
> and redistribute this library.

Certainly.  I'm only passing on what I have been told over the past few
months concerning mingw64.  Perhaps NS or one of the other mingw64
contributors will chime in.

--
Chuck