In some email I received from Guy Harris, sie wrote:
> On Thu, Dec 12, 2002 at 01:04:38AM -0600, David Young wrote:
> > For the sake of a consistent interface across platforms, I prefer
> > that *every* tcpdump support both -L and -y. On platforms without a
> > choice of DLT, however, tcpdump should not require the DLT API from
> > libpcap. Instead, it should provide a trivial implementation which lists
> > one DLT for -L and which insists that -y provide that DLT, only.
> > 
> > On platforms it is implemented, pcap_list_datalinks should always return
> > at least one DLT. Also, pcap_set_datalink should permit you to set at
> > least any DLT on the list, but no DLT that is not on the list.
> 
>       ...
> 
> > pcap_list_datalinks belongs in whichever platform-dependent file
> > pcap_set_datalink is in, since it does not make sense to have one without
> > the other.
> 
> So does that mean you think that "pcap_list_datalinks()" and
> "pcap_set_datalink()" should be present only on platforms that support a
> choice of DLT (i.e., present only on current CVS versions of NetBSD, for
> now), rather than having "pcap_list_datalinks()" return the one and only
> DLT and "pcap_set_datalink()" allowing the DLT to be set only to the
> value it already has?

Shouldn't this also be fairly trivial to support for Solaris ?

Well selecting a DLT anyway, I don't know how you'd go about enumerating
what the currently available ones are.

Darren
-
This is the TCPDUMP workers list. It is archived at
http://www.tcpdump.org/lists/workers/index.html
To unsubscribe use mailto:[EMAIL PROTECTED]?body=unsubscribe

Reply via email to