| Age | Commit message (Collapse) | Author | |
|---|---|---|---|
| 2004-10-28 | manpage to go with sys/netinet6/nd6_rtr.c 1.31 -> 1.32 | Jun-ichiro itojun Hagino | |
| markus ok | |||
| 2004-10-27 | Calculate Hz and round up; ok henning@ | Darren Tucker | |
| 2004-10-27 | use clock_getres(3) and calculate precision from that, and fill the | Henning Brauer | |
| precision field when we reply in server mode accordingly. from phessler | |||
| 2004-10-27 | typo, adresses -> addresses | Jared Yanovich | |
| 2004-10-26 | kill old duplicate code. | Marc Espie | |
| 2004-10-26 | more accurate name: DirRmOnly -> SharedItemsOnly | Marc Espie | |
| use a visitor instead of hardcoding items/users/groups. | |||
| 2004-10-26 | move the code that handles shared items into a separate Module. | Marc Espie | |
| - It will be used after deletion/adding during updates. - We might avoid loading it if a package contains no new directories. Actually a lot of packages don't use it. | |||
| 2004-10-26 | ease the parser a bit. | Henning Brauer | |
| parse() now wants the first argv member to be the first argument it parses, i. e. it does not skip over argv[0] any more, caller has to account for that. the caller does the usual getopt followed by argv += optind; argc -= optind; dance so this is accounted for. in parse() don't use a seperate curarg counter, just in/decrease argv/argc claudio ok | |||
| 2004-10-26 | document -n | Henning Brauer | |
| 2004-10-26 | in the "show summary" view print the peer descr instead of the IP (if a descr | Henning Brauer | |
| is given in the config, fall back to IP otherwise) new -n switch to force display of IPs claudio ok | |||
| 2004-10-26 | do /%u instead of /%6u for max_prefix, looks better, claudio ok | Henning Brauer | |
| 2004-10-26 | make the ntp version check even better, use NTP_VERSION_MIN and _MAX | Henning Brauer | |
| 2004-10-24 | accept replies with a NTP version number of 4 or lower, instead of only 4 | Henning Brauer | |
| as before, we can handle all versions | |||
| 2004-10-23 | beginning of update module: basic check that stuff we are replacing | Marc Espie | |
| is still okay. | |||
| 2004-10-23 | system() method that stores command normal output so that we can show | Marc Espie | |
| it at end of pkg install/deinstall... use for @exec/INSTALL stuff. | |||
| 2004-10-23 | store pkgname in shared items, so that delayed printing can be used | Marc Espie | |
| (and use it). Optimisation: don't clone() directories | |||
| 2004-10-22 | oups | Henning Brauer | |
| 2004-10-22 | in server mode reply with stratum from the peer that we currently prefer | Henning Brauer | |
| plus one | |||
| 2004-10-21 | Backout poll changes; the first mount gets an error for some reason. | Todd C. Miller | |
| 2004-10-20 | merge changes from mod_ssl 2.8.20, in particular a fix for CAN-2004-0885 | Henning Brauer | |
| 2004-10-20 | basic framework for delayed messages. | Marc Espie | |
| 2004-10-19 | fix printing of groups - instead of surrounding each neighbor statement | Henning Brauer | |
| which is part of a group by "group blah { .. }", so that each group decl occurs multiple times (that is valid, but confusing and ugly), stuff pointers to each peer into an a array and sort it based on group id. then print all neighbors whoch are not part of a group followed by each group with all its members. annoyed me for quite some time, now finally enough to sit down and fix it claudio ok | |||
| 2004-10-19 | allow neighbor definitions to depend on interface state. | Henning Brauer | |
| with this, if a neighbor is configured as dependent on carp0 for example, the neighbor will remain in state IDLE as long as carp0 is not master. once carp0 becomes master the session(s) depending on it immediately go to CONNECT (or ACTIVE, if they're configured passive), reducing failover time. claudio ok, with some input from ryan as well | |||
| 2004-10-19 | need to prepend destdir to pkgbase for libspecs checking. | Marc Espie | |
| 2004-10-19 | Extra progress meter for the handling of shared items: first show | Marc Espie | |
| the reading of remaining stuff from installed packages, then the cleaning-up as it progresses. | |||
| 2004-10-18 | Documents @endfake and -Q semantics. | Marc Espie | |
| 2004-10-18 | take window size into account. | Marc Espie | |
| window sizes changes are delayed, nothing to do about it, since perl keeps the signal around until it is safe to activate it... | |||
| 2004-10-18 | experimental progress meter. | Marc Espie | |
| not active unless you specify -x for pkg_add/pkg_delete. (option is bound to change) | |||
| 2004-10-18 | Don't show libspec checking unless very verbose. | Marc Espie | |
| Split addition into `parsing' and actual addition. Don't show parsing info unless verbose, so that pkg_add only displays the actual pkg addition... | |||
| 2004-10-18 | use OpenBSD::Getopt so that we can do -v -v. | Marc Espie | |
| Be much terser by default, start showing dependency information and various things only with at least one -v. Build a `flattened tree' of dependencies information in pkg_add, and show why we install a given package (idea from deraadt@). | |||
| 2004-10-18 | fix report of final problems | Marc Espie | |
| 2004-10-17 | behave according to documentation, e.g., delete groups/users only if -c. | Marc Espie | |
| fix typo. | |||
| 2004-10-17 | implement -q for icmp, deregister, print usigned values using an unsigned | Otto Moerbeek | |
| format. ok markus@ | |||
| 2004-10-16 | oups, unrelated changes crept in | Henning Brauer | |
| 2004-10-16 | Add support for @endfake and pkg_add -Q, to be documented later... | Marc Espie | |
| 2004-10-16 | prefixlen 0 is a special case for prefixlen2mask() as well, claudio ok | Henning Brauer | |
| 2004-10-15 | print ip_len for -v; ok otto, hshoexer, henning | Markus Friedl | |
| 2004-10-15 | Only set IPTOS_LOWDELAY on IPv4 interfaces; pointed out by phessler, ok henning | Darren Tucker | |
| 2004-10-14 | Add missing {} on if statement; ISC#1744 | Ryan Thomas McBride | |
| ok jakob@ | |||
| 2004-10-14 | Have ntpd use IPTOS_LOWDELAY; ok henning@ | Darren Tucker | |
| 2004-10-13 | document @sysctl | Marc Espie | |
| 2004-10-13 | @sysctl | Marc Espie | |
| 2004-10-13 | set rootdelay in replies. | Henning Brauer | |
| inherit rootdelay from the delay from the last client update from the peer that we picked last time to adjust the local clock. in some cases we use the average offset between two peers' client updates, then use the average delay between the two as well. | |||
| 2004-10-13 | in struct ntp_msg, rename "distance" to "rootdelay" to closer match RFCs | Henning Brauer | |
| and such | |||
| 2004-10-13 | thinko, htonl() -> ntohl(). as we don't use the value in question effect zero | Henning Brauer | |
| 2004-10-13 | fall cleaning | Henning Brauer | |
| 2004-10-13 | correctly set refid in replies with NTP protocol versions < 4. | Henning Brauer | |
| code path for NTP4 remains unchanged, we already set refid correctly there. NTP3 and older uses an IPv4 address as refid. use the IP of the server we last synced to if it was a IPv4 one. sometimes we use the average offset between two, in that case just pick one for the IP. this scheme naturally fails when we query IPv6 servers and have to reply to IPv4 NTP3 (or even older NTP versions) clients - refid stays at 0 then. this is a protocol limitation, nothing we can do about it. | |||
| 2004-10-13 | when we get back a IMSG_HOST_DNS message from the parent the peer ID within | Henning Brauer | |
| might have become invalid (because the peer showed up, dns request sent to parent, peer vanishes, and then the reply comes back), so do not fatal() in that case but just log_warnx(). provoked by brad | |||
| 2004-10-12 | nicer display | Marc Espie | |
| 2004-10-12 | Use a udp socket to communicate with the dhcp servers. The bpf interface | Claudio Jeker | |
| can't be used to forward the packets. Bpf would relay the packets to wrong interface. Reports and tests by Hans Kremers and Colin Harford OK henning@ | |||
 |
index : src | |
| OpenBSD base system |
| summaryrefslogtreecommitdiff |