I think (and I think I've mentioned this a few times :) it's not possible to implement a good (proper, nice, pretty) async DNS library to do *client*-side hostname lookups. Because on the client side you want to read things like /etc/hosts and the default lookup domain (search xx in /etc/resolv.conf). You could maybe kludge these for some specific OSes, but somehow I doubt any non-libc coder actually wants to do that.. Until POSIX (or whatever new standard) defines such a thing, the only possibility is to use libc's gethostbyname() (or whatever) that does the hard work. For async behavior either threads or forking is needed. In mosh's case I think either one works fine, since it's not a library..
On 8.7.2013, at 8.22, Keith Winstein <kei...@mit.edu> wrote: > Hello Lawrence, > > Unfortunately not. (See previous reply.) But you can follow this issue at > https://github.com/keithw/mosh/issues/210 > > The main blocker for this is that we need an asynchronous DNS resolver in > mosh-client to be able to discover if the server's IP has changed in the > middle of the connection (and whether we should try another IP). > > There are a few options here, none of them super-great. (Link with c-ares? > Use gethostbyname() from a thread or forked process?) > > If you or anybody else has an opinion about the best way to do this, happy to > discuss or take a patch to add this support. > > Best regards, > Keith > > On Fri, Jun 28, 2013 at 8:34 AM, Lawrence 'The Dreamer' Chen > <thedrea...@lhaven.net> wrote: > Can Mosh work if the mosh-server IP changes as I roam between networks? > > Namely when I'm on campus wifi, I'll have a 10.x.x.x address and the > hostname for mosh-server will resolve to one as well. > > But, when I'm off campus. I'll have whatever IP is assigned by the > network I'm on, and the hostname for mosh-server will resolve to its NAT > IP (which I'm told is subject to change, but so far hasn't. I'm the DNS > administrator, so I should get notice of any changes...) > > Not sure if its because the NAT is only on the outside, or that internal > routing doesn't send the response back out through the NAT... > > -- > Name: Lawrence "The Dreamer" Chen Email: thedrea...@lhaven.net > Blog: http://lawrencechen.net > > _______________________________________________ > mosh-devel mailing list > mosh-devel@mit.edu > http://mailman.mit.edu/mailman/listinfo/mosh-devel > > _______________________________________________ > mosh-devel mailing list > mosh-devel@mit.edu > http://mailman.mit.edu/mailman/listinfo/mosh-devel _______________________________________________ mosh-devel mailing list mosh-devel@mit.edu http://mailman.mit.edu/mailman/listinfo/mosh-devel
