On Wednesday 17 March 2004 21:43, James Yonan wrote: > Juan, > > I think it would be great if you wanted to work on this. Until now, most > of the interest has been in IPv6 over IPv4 tunneling, but I would encourage > you to take the initiative and work on this.
Thank you very much for the info, but Im sorry to decline your offer. I dont have enough time to work on this, though I will be glad to contribute with small patches if it is necessary. Besides, I don't know about Windows IPv6 stuff, so I think I couldn't be able to create portable code. See one more comment below... > Here are some thoughts: > > * This work should happen in the BETA20 branch, as the 1.x branches are now > very stable and shouldn't be touched at this point for more than bugfixes > and localized changes. While the BETA20 branch is mostly focused right now > on point-to-multipoint mode (which is not yet finished), it will currently > build off of the CVS and run in standard point-to-point mode. > > * It's important that this work be portable which of course means that we > want to share as much code between the windows and non-windows versions as > possible. I'm not sure to what extent the IPv6 header files are > standardized, but if they are not, it may be necessary to build a > platform-general abstraction. To the extent that portability is > impossible, the code should ifdef itself out on platforms which don't > support IPv6. > > * OpenVPN uses in_addr_t and sockaddr_in types to refer to IPv4 addresses. > We need to decide whether IPv6 should be a whole new mode, i.e. with it's > own address structs/typedefs and socket6.[ch] files, or whether we should > use an abstraction for address types which can be either IPv4 or IPv6. I think it would be easier to do it as a whole new mode. Other opinions are welcomed of course (^-^) Bye bye... > * There is code in route.c and tun.c which does things like network = > (ip_addr & netmask). This code obviously won't work if the protocol > address size is too big to fit in an intrinsic C type, as it is with IPv6. > > James > > Juan Rodriguez Hervella <j...@it.uc3m.es> said: > > Hellooooo, > > > > It seems that openVPN supports only IPv4 for making > > connections between the client and the server. > > > > I'd like to be able to establish connections over IPv6, > > what should be changed ? I feel quite depressed when > > I look at the source code, because I think is gonna be a pain > > to change the code to turn it to be AF-independent....sigh > > > > This should have been done from the scratch...but I don't > > blame anybody, specially because IPv6 is not widely used yet, > > (^--^) > > > > Is anybody willing to start doing this ? > > Should we start changing everything to accomplish this task > > or is it enough to change "socket.c" ? and if we look at socket.c, > > what's the best way of porting that stuff to IPv6 ? making new > > functions like "gettaddr_ipv6", "update_remote_ipv6".....or > > changing the original functions to deal with both IPv4 and IPv6 > > at the same time ?... > > > > Regards > > > > > > -- > > ****** > > JFRH > > ****** > > > > Passionate hatred can give meaning and purpose to an empty life. > > -- Eric Hoffer > > > > > > ------------------------------------------------------- > > This SF.Net email is sponsored by: IBM Linux Tutorials > > Free Linux tutorial presented by Daniel Robbins, President and CEO of > > GenToo technologies. Learn everything from fundamentals to system > > administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click > > _______________________________________________ > > Openvpn-devel mailing list > > Openvpn-devel@lists.sourceforge.net > > https://lists.sourceforge.net/lists/listinfo/openvpn-devel -- ****** JFRH ****** It is not true that life is one damn thing after another -- it's one damn thing over and over. -- Edna St. Vincent Millay
