Re: [tcpdump-workers] [tcpdump] timestamps: make possible to request high precision timestamps (#377)
Guy Harris wrote: > Oops, OS X tcpdump has -k: >-k Print the comment field from the libpcap packet header. Apple > modification. > so it's getopt_long() time. (I think "comment field from the libpcap > packet header" means "comment option from the pcap-ng packet block".) Yes. My notion was to rid tcpdump.c of all code not related to option parsing, and then introduce a pktdump.c interface that that could start fresh in the short-option parsing space. Could toggle on $0/argv[0]. > I've got a version of tcpdump that uses getopt_long(); I'll work on > making it work right when the OS doesn't have getopt_long(), and check > it in. okay... ! -- ] Never tell me the odds! | ipv6 mesh networks [ ] Michael Richardson, Sandelman Software Works| network architect [ ] m...@sandelman.ca http://www.sandelman.ca/| ruby on rails[ ___ tcpdump-workers mailing list tcpdump-workers@lists.tcpdump.org https://lists.sandelman.ca/mailman/listinfo/tcpdump-workers
Re: [tcpdump-workers] [tcpdump] timestamps: make possible to request high precision timestamps (#377)
On May 3, 2014, at 11:55 AM, Michael Richardson wrote: > Yes. My notion was to rid tcpdump.c of all code not related to option > parsing, and then introduce a pktdump.c interface that that could start > fresh in the short-option parsing space. Could toggle on $0/argv[0]. Unless pktdump goes down the Wireshark path of having dissectors register preferences by name, with a generic option such as "-o" allowing options to be set by name (which would allow -b, -E, -H, -K, -m, -M, -R, and -S to be absorbed into that option), it might still run the risk of running out of single-letter options. ___ tcpdump-workers mailing list tcpdump-workers@lists.tcpdump.org https://lists.sandelman.ca/mailman/listinfo/tcpdump-workers
Re: [tcpdump-workers] [tcpdump] timestamps: make possible to request high precision timestamps (#377)
Guy Harris wrote: >> Yes. My notion was to rid tcpdump.c of all code not related to option >> parsing, and then introduce a pktdump.c interface that that could >> start fresh in the short-option parsing space. Could toggle on >> $0/argv[0]. > Unless pktdump goes down the Wireshark path of having dissectors > register preferences by name, with a generic option such as "-o" > allowing options to be set by name (which would allow -b, -E, -H, -K, > -m, -M, -R, and -S to be absorbed into that option), it might still run > the risk of running out of single-letter options. I don't think that I'd want to let specific dissectors capture single letter options, period. So, yes, -o sounds right, or just use long options for that. -- ] Never tell me the odds! | ipv6 mesh networks [ ] Michael Richardson, Sandelman Software Works| network architect [ ] m...@sandelman.ca http://www.sandelman.ca/| ruby on rails[ ___ tcpdump-workers mailing list tcpdump-workers@lists.tcpdump.org https://lists.sandelman.ca/mailman/listinfo/tcpdump-workers
