IPv6 strawman plan: Re: [OpenAFS-devel] Moving Forwards

[Troy Benjegerdes]

Am I going to have any problems with trademark nonsense if I call
this 'AFS6', and use the same ports, but only on IPv6 addresses?
I don't mean to produce a throw-away implementation, but I also 
do not want to confuse anyone that this is anything other than 
alpha-use-at-your-own-risk software. (but hey, gmail's been 'beta'
for years) .... What should I call an ipv6 prototype?

I would like to have a working code that I can run regression tests
against before proposing a standard that interoperates with AFS3
and ipv4.

On Mon, Sep 10, 2012 at 01:19:40AM -0400, Jeffrey Altman wrote:
> While you may fork the code base, you may not fork the AFS3 wire
> protocols.  If you decide to implement your own extensions to AFS3 you
> must follow the rules of the AFS3 standardization and RPC registration
> processes.  If you choose not to follow those rules, then you must
> implement your own wire protocol that will not interfere with AFS3
> clients and servers.
> 
> While the approach you describe below might be sufficient for your
> personal needs, it is unlikely that it would be acceptable to OpenAFS.
> 
> Jeffrey Altman
> 
> 
> On 9/10/2012 12:33 AM, Troy Benjegerdes wrote:
> > 1) replace kernel filesystem type of 'afs' with 'tfs''
> > 2) expand all 32 bit IP address data structures to 128 bits
> > 3) change sockaddr() etc calls to INET6
> > 4) develop ipv4 afs to ipv6 tfs proxy server
> > 5) kernel module housekeeping to allow openafs.ko and tfs.ko to
> > be loaded at the same time to allow clients to mount /afs using
> > legacy OpenAFS, and /tfs moving forward.
> > 
> > Is there anything else I'm missing (besides the flamewar that
> > will probably follow regarding the name change?)
> > _______________________________________________
> > OpenAFS-devel mailing list
> > OpenAFS-devel@openafs.org
> > https://lists.openafs.org/mailman/listinfo/openafs-devel
> > 
>