On Wed, 2001-12-12 at 14:57, Earnie Boyd wrote:
> Henrik Stokseth wrote:
> > TAMURA Kent <kent@...> wrote:
> > > What is a correct way to add missing things into w32api? It is
> > > illegal to look at headers of Platform SDK provided by
> > > Microsoft, right? Is it legal to look at Platform SDK documents
> > > instead of its headers? What is a legal way to determine the
> > > value of unknown macro?
> > I'm glad this issue was brought up. IMHO open source people are way too
> > cautious when it comes to make their equivalents of stuff made by microsoft.
> Well, IMO, you can never be too cautious on this issue and you seem to
> agree with that in your next statement.
> > have you guys ever heard of the 'clean room approach'? it is fully legal.
> > when some manufacturers tried to make clones of the first pc, they had one
> > problem. most parts were available for anyone, but the bios used in the IBM
> > PC was copyrighted by phoenix corp. IIRC. so what did they do? used the
> > clean room approach: you have a group of people go through the bios and
> > analyse and write a fully detailed document on it and give it to another
> > group of people who have never looked at the BIOS firmware, which then will
> > start developing a compatible bios. this method CAN NOT be legally pursued!
> This would be the correct method. Is someone willing to provide the
> fully detailed document? I would be glad to add you as a developer
> whose sole purpose is to document what isn't already documented, and
> that being primarily the constants used. You can't have prior knowledge
> of the PSDK headers, even to the point of having downloaded them for use
> so as to avoid the temptations to look. Also, I would like for this
> documentation to be non-copyrighted, i.e.: Public Domain.
Why is it necessary someone _really_ have to do this? If one makes a
MSVC++ program that only outputs the macro values it's not even
considered reverse engineering! Like:
and give it to someone with MSVC++ to compile it and give the result
Of course that these kind of programs can be generated automatically, or
even more informative ones.
I don't see anything illegal on this, since we would be using the MSVC++
for its purposes - to compile programs. And the person that runs it
doesn't even have to see a windows header.
> > One of the things that bothers me most with Mingw is the lack of Microsoft
> > compatible headers. I think the Win32 headers now is fairly good, but things
> > like DirectX headers are not usable.
> It takes committed developers with time and burning desire to do it. So
> far, no one has.