[Openswan dev] nat-t openswan interop problem Win2003

[Ken Bantoft]

On Wed, 4 Jan 2006, Ken Bantoft wrote:

>
> On Wed, 4 Jan 2006, Jacco de Leeuw wrote:
>
>> Michael Richardson wrote:
>>
>>>     Jacco> Perhaps they fixed it in Windows 2003 R2 or Vista but I
>>>     Jacco> haven't tried.  This is how Microsoft works, you pay for new
>>>     Jacco> features...
>> 
>> Perhaps I wasn't clear but this was meant to be a cynical remark about
>> Microsoft's upgrade policy. I expect RFC 3947 support to be only in
>> Vista or possibly Windows 2003 R2. No, I don't like this either.
>> 
>> This situation has happened before: in 2003 Microsoft employees said that
>> a NAT-T server-side update would be released for Windows 2000 Server, but
>> later it turned out that only Windows 2003 would support server-side NAT-T.
>> This was strictly a business decision because they did release a 
>> client-side
>> NAT-T update for Windows 2000 Professional.
>> 
>>> So, who pays us to add work arounds for features that you didn't pay
>>> us?  It is a serious question.  Why should open source maintainers take
>>> time away from adding new features to support interoperating with people
>>> who were too cheap to pay their yearly tithe to microsoft?
>> 
>> I'm not sure what you are getting at.
>> 
>> I reported two issues. The first one was caused by an honest mistake
>> of the RFC maintainer. Microsoft chose to interpret the MD5 calculation
>> one way, you chose the other way. There is much to say about Microsoft's
>> unfair business practices but this is not one of them. The second issue
>> is an Openswan bug, according to yourself.
>
> I know I fixed this in my GIT tree, I'll backport it into 2.4.5 tree.
>

Let me clarify - I fixed the sending of draft 02_N (which I just commited 
to CVS).  I only just read about the other errors happening with win2k3 
you found.

Ken