Welcome Microsoft to our GPL club. Let's try to help you with some tips:
- While taking help from a prominent kernel contributor as g-k-h was very wise and he, no doubt, prevented you from falling in many pitfalls; you would still need a lot of patience. It takes time until people review your patches, ask questions in LKML, propose different approach, ask for rational etc, etc. Take your time, it's nothing personal against you -- it's simply the (somewhat) painful process for including code in the Linux kernel.
- Also, many times the results of this process generate a very different outcome than what the initial coder was thinking. You can just ask Ingo Molnar (as an example) how much time he's been working on the RT patches, how many iterations has passed and how much the result differ from the initial code contributed by MontaVista.
- In some cases the code may have to be maintained outside of the official kernel for a long time. Don't take it personally, it happened to the best/biggest Linux companies. Take as an example GFS2 and OCFS2 -- two clustering filesystems that are in active use by customers of RedHat and Oracle respectively for several years. Both are still not part of the Linux kernel (you can dig LKML for the gory details).
- There were even some extreme cases where code was not (shock, panic, awe...) merged at all! Even after IBM (which is one of the 3 biggest kernel contributors) invested a lot of time into EVMS (yes, it was long time ago..), its solution never made it to the official kernel. LVM2 took its place. Sometimes life can be tough. If worst happens and your code is left out in the cold -- don't despair, maybe next time...
Ah, and just another detail. There would be no doubt people in our community who would raise the patents issues and use them as an argument to reject your patches. While I cannot offer you a complete solution to this problem (until the patent system is revised to the ground), a small and simple idea may mitigate your problem -- make every patch sent to LKML contain the magic line:
Signed-off-by: Steve Ballmer
Cheers,
Yes: this post was written with a tongue in the cheek.
However: all claims are valid -- nobody has promised a free ticket into the kernel to any company who is willing to contribute code...