Hi there, Michael Michael Tautschnig wrote: > 1. Quoting Aníbal from the message the URL of which you gave above: > "Yes, I've been aware of arpwatch-ng and was thinking about changing > arpwatch upstream soruce to arpwatch-ng's." > Why isn't this approach being pursued?
Because he has had three NMU's, his last upload was on 2004 and I have got no reply from him since I pinged him and Anibal about a month ago. I just don't want to step on his toes. It would be a takeover, and he just asked to be in the Uploaders field for arpwatch-ng. > 2. The description a) seems to indicate that arpwatch-ng just has a few 64bit > patches that were best integrated into arpwatch anyhow; b) makes me simply > wonder what "improving" means, as I'm running it on 64bit systems just fine!? There's other improvements however, and quite some, just look at http://freequaos.host.sk/arpwatch/ (pasting changelog here for completeness on the BTS): arpwatch NG 1.7: update autoconf system to support x86_64 better [UPDATED] arpwatch NG 1.6: use a central report function table to ease customization [FIXED] minor cleanups and updates [FIXED] arpwatch NG 1.5: try to report error on startup better _ arp.dat _ ethercodes.dat [FIXED] arpwatch NG 1.4: try to report _all anomalities via the report function _not syslog [FIXED] mode 2 _ make action list parseable [FIXED] further static'fy local functions in arpwatch.c [FIXED] ethercodes updated from nmap-4.11 and removed old ones [UPDATED] arpwatch NG 1.2: on make install also install man-pages [FIXED] ethercodes updated from nmap-4.00 [UPDATED] arpwatch NG 1.1: allow for attaching pcap / tcpdump_style filters [ADDED] improve and update man-page [FIXED] arpwatch NG 1.0: allow getopt() to complain about unknown options [FIXED] rework help output and send to stdout, not stderr [FIXED] previous versions _ all changes included in NG 1.0 _ DROP1: code was in bad and old shape [FIXED] ethercodes were old - updated from NMAP [FIXED] DROP2: reporting to stdout added [ADDED] ethercodes updated from nmap-3.81 [UPDATED] DROP3: report using raw mode _ so later filters can change output as they like _ [ADDED] close stdin _ stdout _ stderr in daemon again [FIXED] clean up reporting subsystem [FIXED] DROP4: compile fix for arpwatch.c [FIXED] segfault fix for report.c -> arpwatch.c: uninitialized function pointer [FIXED] include overflow security fix [FIXED] DROP4.1: compile fix for NON-GNU systems: strndup() missing [ADDED] DROP5: rewrite make install target [FIXED] remove #ifdef DEBUG altogether [FIXED] simplify initialization code in arpwatch.c [FIXED] DROP6: apply lots of vendor patches from debian [MERGED] eg continue on unconfigured interface, option -p for non-promiscous mode, ... - see changelog misc fixes [FIXED] DROP7: more debian patches [MERGED] drop privileges to user _ specify sendmail-prog _ mail-to option minor security fix for replacement strndup() [FIXED] DROP8: added fancy mac adress printing [ADDED] DROP9: ethercodes updated from nmap-3.83 [UPDATED] add -F as mail_from option [ADDED] shorten checkpoint time of arp db [FIXED] also try mktemp in mkdep script [FIXED] DROP10: fix bug: arp.dat was not checkpointed or updated while running [FIXED] Not only 64bits users will benefit from these changes. And yes, I do have hundreds of installs on 64bits that do segfault with vanilla arpwatch, so I can confirm it does. In short, I don't care that much about how to fix this, but I want it fixed and without getting in arpwatch's maintainer's way. -- .''`. Ex nihilo nihil fit : :' : `. `' `- Proudly running Debian GNU/Linux -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
