Year 2000 WG, August 11 1997, Munich Minute taker: Philip Nesser Chair: Erik Huizer Agenda 1. Opening 2. Review of Research Results So Far 3. Status of Areas 4. ID Setup 5. AOB 6. Scope Discussion Erik recaps Memphis meeting most eloquently. Areas Email No remarks DNS Timers in security stuff(make it modulo 2^32) Serial Numbers Robert Elz promises report within one month Network Management A few MIB's specify date but no format Version control in some MIB modules SNMP in general should be fine Problem in ASN.1. We should send a liason report to ITU. Virtual Terminal TN3270 and Telnet are fine, modulo authentication protocols SSH should be fine X windows still needs more investigation should be passed to the Open Group. Janecek, Peter from the Open Group formally takes the handoff. Info Services & File Transfer Report on HTTP. 1.1 requires 4 digit years from host requirements (RFC1123) but allows acceptance of RFC 850. HTML should be fine as of RFC 1866 Nothing on File transfers (ftp, tftp) or other Info services. Martin Carpenter volunteers News Nothing from Erik Fair. Similar to mail. Try sticking it to Stan Barber. INN has been given a clean bill of health Real Time Services AVT & MMUSIC should be asked to look into it. Security Talk to Bob Moscowitz Directory Services Rick Wesson reports 6 RFC's that allow 2 or 4 year digits. They should be reported to the LDAP and X.500 people for update. Erik will talk to ASID WG. Disk Sharing NFS was given clean bill of health by Barbara Jennings However SUN reacted with some comments. AFD/DFS should be passed off to the Open Group Autoconfiguration BOOTP was okay. Still need DHCP & DHCPv6. Talk to DHCP WG & IPv6 regarding autoconfiguration. Games & Chat Still no report NTP Standard 32 bit rollover in 2036. Routing BGP4 is okay. No reports on anything else. All IGRP stuff and multicast routing need to done IP/TCP/PPP, etc Robert Elz . Interim report. Still busy checking all STD RFC's No problems so far except 3 small exceptions. Standard 44 (RFC891) has a 1987 bug it contains a 4 bit year offset from 1972! Standard 25 Daytime protocol has example with 2 year digit. Standard 26 Time Server Protocol has the 2036 rolllover problem. Robert will continue to work and include RPC items. Questions: Question about X.509. Problem has been fixed in version 3. ID Set-UP Should we use prescriptive language? We SHOULD use things like words l like MUST. Redo introduction after work is finished. Should do executive summary with itemized list. We should delete all of the appendices when we go to RFC and replace with a short summary. We should point out that various network management platforms compare MIB versions by a two digit date will need to be fixed. The WG decided not to do implementations. If we find standards that require a two digit year, we need to have them revised. Any other Business Any interest is doing a pilot system? A few people are doing it. Will feed back information to WG by DC. Year 2000 bake off is good idea. Should we come up with an algorithm to assume a year based on being given a 2 year date. Need to comment about relative ages. (email age versus age of a human) IBM will send in their version for SNA uses sliding window approach Decision: WG will include suggested algorithms in the RFC. Chair should send a message to IESG regarding their responsibility to check new RFCs for y2k problem. Whats the scope of the WG? We only do RFCs and also other popular de facto standards on the Internet that are not handled by other standards bodies. We should get the legal council to write a disclaimer. New draft by the end of September. ===============