Rework FoE Busy handling and enables in global test of win32

[lehned]

There are two commits:
1st: FoE has incorrect BUSY handling without resending a packet and therefore aborting due timeout on wait of a never sended packet.

2nd: cleanup win32 examples and adding CMakeFiles, additional enabling win32 them by default for tests

[nakarlsson]

Does this PR supersede the PRs provided by @blocksebastian?
#623
#624

[nakarlsson]

@lehned , ping

[lehned]

Sorry for the late answer. It supersede #624 not #623, which has additional changes in other files .
As i remember there were an issue with sendpacket (FoE Packet number) counter on BUSY state, which i addressed, also.

[blocksebastian]

Does this PR supersede the PRs provided by @blocksebastian? #623 #624

Yes, that was my old account (which is taken by project paperwork).

[blocksebastian]

Sorry for the late answer. It supersede #624 not #623, which has additional changes in other files . As i remember there were an issue with sendpacket (FoE Packet number) counter on BUSY state, which i addressed, also.
Which issue? Can you explain?

[nakarlsson]

@lehned and @blocksebastian , to make the review easier on this topics, can we close #624 and #56?
Continue to review #623 and #627?

[blocksebastian]

we can close the old #56. I splitted the old one into multiple as this project didn't react and hoped it will be integrated if I make the pull requests smaller and one topic a pull-request.

[lehned]

Sorry for the late answer. It supersede #624 not #623, which has additional changes in other files . As i remember there were an issue with sendpacket (FoE Packet number) counter on BUSY state, which i addressed, also.
Which issue? Can you explain?

The sendpacket counter will become corrupt, in case the client application verifies the packet counter.