Hi Simon,
Simon Josefsson <[email protected]> writes:
> Thanks - I'm neutral on these patches, as I don't really know what a
> positive or negative test for them would be. However, speaking on
> inet_ntoa and different code paths reminded me: a good todo work item
> would be to merge the ping4 and ping6 tools. I don't think there are
> any reasonable arguments for having different main() etc code paths for
> these two tools, they ought to be close enough to use the same overall
> logic and differ when needed depending on IPv4 vs IPv6. Does anyone see
> any strong argument against that? I think some small (hopefully
> unintentional) variations between these tools have sneaked in because
> they aren't synced, and I think it would be nice to make them more
> consistent. Not sure if you want to work on this, but thought I should
> mention it.
I don't see an issue with it. I think I have seen some versions of ping
use a single binary for IPv4 and IPv6 and others use separate. I forget
which implementations do which. Perhaps the older ones use separate
binaries? I'll have a look at merging them.
Also, I wonder if we can add some headers to gnulib. Currently Inetutils
has a "replacement" header for <netinet/ip_icmp.h>. This should be
available on BSD and GNU libc systems.
Likewise we check for arpa/{ftp.h,telnet.h,tftp.h} and define things if
they aren't available. Though those files have BSD licenses on them
which would likely mean messing with gnulib-tool...
Collin
