So, has anyone seriously estimated the cost and timeframe to get a production version of RxTCP6 up and running? This seems like a project to put on kickstarter.com or some other site like that, and start collecting money.
(And can we roll RxK5/RxSSL encryption in if it's not already fully baked in somewhere? ) Let me also propose something else which is probably disgusting and wrong, but I want to hear someone else comment ;) What if, instead of changing all the RPC's, we continue to use a 32 bit field, and use some flag to identify RxV6, and then use 24-30 bits of what used to be an IP address as an index and/or key for a key->IPv6 address mapping structure. Some of the address space would be for dynamic mappings, while others would be IDs for a new PTserver/Ubik name to IPv6 address mapping... I'll volunteer to write/debug/hack a DNS server that uses this Ubik database ;) On Wed, Dec 08, 2010 at 06:04:12PM -0800, Russ Allbery wrote: > Coy Hile <[email protected]> writes: > > > I'm well aware that currently (as of 1.4.x), running OpenAFS on IPv6 is > > not supported. However, the recent discussions on nanog about the last > > few /8s being allocated to RIRs has got me thinking about what would be > > required to run OpenAFS over IPv6. Just how much work is it? > > It requires essentially rewriting the network layer of AFS, due to the > number of places that assume IPv4 addresses, with corresponding changes > all the way up the protocol stack for field lengths in various RPCs that > hold IP addresses that would no longer be long enough. > > That's why normally IPv6 support is discussed in conjunction with > implementing Rx over TCP, since that effort would require touching many of > the same pieces of the code. > > -- > Russ Allbery ([email protected]) <http://www.eyrie.org/~eagle/> > _______________________________________________ > OpenAFS-devel mailing list > [email protected] > https://lists.openafs.org/mailman/listinfo/openafs-devel -- ---------------------------------------------------------------------------- Troy Benjegerdes 'da hozer' [email protected] CTO, Freedom Fertilizer, Sustainable wind to NH3, [email protected] Benjegerdes Farms TerraCarbo biofuels The challenge in changing the world is not in having great ideas, it's in having stupid simple ideas, as those are the ones that cause change. Intellectual property is one of those great complicated ideas that intellectuals like to intellectualize over, lawyers like to bill too much over, and engineers like to overengineer. Meanwhile, it's the stupid simple ideas underfoot that create wealth. -- Troy, Mar 2010 _______________________________________________ OpenAFS-devel mailing list [email protected] https://lists.openafs.org/mailman/listinfo/openafs-devel
