2005-09-27 17:49 knight * modules/m_info.c: - OUTPUT_STRING_PTR -> OUTPUT_STRING 2005-09-27 16:15 adx * src/channel.c: - fixed a simple bug spotted by Michael 2005-09-27 16:10 adx * src/channel.c: - hopefully fixed an off-by-one bug that caused server numerics during the connect burst 2005-09-27 11:44 adx * README.PLATFORMS: - added db as FreeBSD port maintainer 2005-09-27 11:29 adx * src/s_conf.c: - slight corrections to /kline nick 2005-09-27 08:43 adx * TODO, contrib/m_ctrace.c, contrib/m_ltrace.c, modules/m_admin.c, modules/m_etrace.c, modules/m_info.c, modules/m_links.c, modules/m_stats.c, modules/m_trace.c, modules/m_whois.c: - fixed --disable-shared-modules 2005-09-25 23:10 adx * src/s_log.c: - use CRLF when writing logs under w32 2005-09-25 23:06 adx * contrib/setup-win32.h: - make ircd_prefix look nicer on win32 ie c:/ircd/blah instead of C:\ircd/blah 2005-09-25 22:57 adx * src/fileio.c: - () were missing here 2005-09-25 22:52 adx * include/fileio.h, src/fileio.c: - continue libc vs API fixes 2005-09-25 22:21 adx * src/: fdlist.c, fileio.c: - this should fix an old win32 bug I was always aware of but too lazy to fix ;) basically we can't treat libc handles from open() as the same "handlespace" as sockets. Under win32, libc handles are artificial (indexes to libc array of open files), whereas sockets are real handles from kernel. They might collide (although quite unlikely regarding NT handle allocation scheme). - solved by using win32 CreateFile / CloseHandle API instead of open / close. Now both sockets and files exist in the same handlespace (kernel objects) 2005-09-25 21:42 adx * src/fdlist.c: - close_fds shouldn't use fd_close. Unnecessary overhead + fd_close calls comm_setselect which can use (for example) the epoll fd; results in a core if the epoll fd is closed already. Just using close() on all descriptors is quite enough. 2005-09-25 21:13 adx * src/fdlist.c: - don't call comm_setselect on non-sockets! 2005-09-25 16:10 knight * src/: ircd_parser.y, y.tab.c, y.tab.h: - Fix same issue with reject_hold_time as join_flood_time had. It was using NUMBER instead of timespec. 2005-09-25 16:07 knight * src/: ircd_parser.y, y.tab.c, y.tab.h: - Fix parse error in config for join_flood_time. 2005-09-25 14:48 db * src/hostmask.c: - trivial comment fix 2005-09-25 14:43 michael * src/hostmask.c: - Fixed find_conf_by_address() so it actually works with encrypted auth{} passwords. 2005-09-25 14:12 michael * src/s_conf.c: - Fixed core with match_conf_password(). Too bad glibc's strcmp doesn't test for pointers being NULL 2005-09-25 06:15 michael * RELNOTES: - Updated RELNOTES 2005-09-25 05:44 michael * src/resv.c: - Another episode of "I propably shouldn't be coding on a sunday morning without having a coffee first ..." Accessing invalid memory isn't a nice thing ;-) 2005-09-25 04:17 michael * src/resv.c: - Another fix to valid_wild_card_simple(). Must increase 'nonwild' for escaped characters aswell. 2005-09-25 02:33 michael * src/resv.c: - Fixed bug in valid_wild_card_simple() where we would treat escaped wildcard characters (i.e. \*, \#, \?) as real wildcards. Cleaned it up a tad while on it. 2005-09-25 02:05 michael * src/s_conf.c: - Fixed the "KLINE nick" bug reported by pi. We now always place a kline on the sockhost field if the client is a local client or an unspoofed remote client. If it is a remote client with a spoof, a kline on the host field will be placed. Remember: sockhost is "0" for remote clients that have a spoofed host. Piotr: please look if everything works as advertised. 2005-09-24 08:38 michael * include/ircd_defs.h, modules/m_away.c, modules/core/m_kick.c, modules/core/m_part.c, modules/core/m_quit.c, modules/core/m_squit.c, src/resv.c, src/s_user.c: - Added separate KICKLEN and AWAYLEN definitions 2005-09-24 05:30 michael * include/ircdauth.h: - Remove old ircdauth.h 2005-09-24 05:27 michael * include/channel_mode.h, include/dbuf.h, include/event.h, include/fdlist.h, include/fileio.h, include/handlers.h, include/hook.h, include/hostmask.h, include/irc_string.h, include/motd.h, include/rsa.h, include/s_auth.h, include/s_misc.h, include/s_serv.h, include/s_stats.h, include/send.h, include/userhost.h, src/hash.c, src/ircd.c, src/ircd_parser.y, src/s_auth.c, src/s_bsd.c, src/y.tab.c: - Misc. headerfile cleanups, fixed prototypes etc. 2005-09-22 12:01 adx * src/hostmask.c: - fixed password issue (Unauthorized connection if someone was connecting with a password but no password was defined in conf) 2005-09-20 11:47 db * src/: ircd_parser.y, y.tab.c, y.tab.h: - bitlen should be pass 1 not pass 2 - minor edit, change tabs to spaces - regenerate parser tables 2005-09-20 11:34 adx * TODO: - updated my TODO 2005-09-20 11:33 db * src/s_conf.c: - removed bogus check - removed now unused variable 2005-09-20 11:15 db * src/: ircd_parser.y, s_conf.c: - rewrote the convoluted mess that was there for classes 2005-09-20 10:59 adx * configure, configure.ac, include/patchlevel.h: - restored optimization to -O2 in preparation for rc1 release - revved patchlevel 2005-09-20 08:41 adx * etc/example.conf, etc/example.conf.quick, etc/example.efnet.conf, help/opers/umode, src/ircd_parser.y, src/y.tab.c, src/y.tab.h: - added reject_hold / join_flood_* settings to example confs - moved join flood to channels{} section 2005-09-19 10:01 adx * RELNOTES: - updated RELNOTES 2005-09-19 06:36 adx * contrib/win32build.bat, tools/win32/kill.c, tools/win32/rehash.c, tools/win32/remotd.c: - added win32 tools 2005-09-19 02:15 metalrock * configure, configure.ac, src/restart.c, src/whowas.c: - Renamed shadowed declaration. - Removed whitespace from includes on whowas.c - Lower case v for voting to be consistent with other options. 2005-09-18 19:07 adx * contrib/Makefile.in, contrib/README, contrib/m_etrace.c, help/opers/index, include/modules.h, modules/Makefile.in, modules/m_etrace.c, src/modules.c: - moved etrace to base, as per rb 2005-09-18 18:36 adx * contrib/Makefile.in, contrib/README, contrib/m_omotd.c, contrib/help/opers/omotd, help/opers/index, help/opers/omotd, include/modules.h, modules/Makefile.in, modules/m_omotd.c, src/modules.c: - moved /omotd to autoload - fixed issue where ctrace,ltrace,etrace were not compiled with contrib on static builds (but were referenced in modules.c -> linking error) - shall we move *trace to autoload? 2005-09-18 18:24 adx * include/fdlist.h, modules/m_help.c, modules/m_rehash.c, modules/m_set.c, src/event.c, src/fdlist.c, src/ircd.c, src/ircd_parser.y, src/y.tab.c, src/y.tab.h: - call recalc_fdlimit periodically - use setrlimit to raise max number of fds to the hard limit - cosmetic changes 2005-09-18 17:48 adx * src/modules.c: - mod_add_cmd(mkpasswd) on win32 too 2005-09-18 16:49 adx * src/s_bsd_win32.c: - coded report_dns_servers for win32 2005-09-18 16:23 adx * src/s_serv.c: - fixed a simple bug in fork_server 2005-09-18 16:09 adx * configure, configure.ac, contrib/setup-win32.h, include/client.h, include/fdlist.h, include/ircd_defs.h, include/s_bsd.h, include/s_serv.h, include/setup.h.in, src/fdlist.c, src/hook.c, src/ircd.c, src/packet.c, src/restart.c, src/s_bsd.c, src/s_bsd_sigio.c, src/s_bsd_win32.c, src/s_serv.c: - merged functions like set_non_blocking, set_no_delay, set_sock_buffers into setup_socket callback - removed USE_SIGIO define + win32 related code from socket I/O, now they use hooks instead - sanified fork_server 2005-09-18 14:08 adx * include/hook.h, src/fdlist.c, src/s_bsd_poll.c, src/s_bsd_select.c, src/s_bsd_sigio.c: - (hopefully) adapted select, poll & sigio:poll to dynamic HARD_FDLIMIT 2005-09-18 13:33 knight * doc/technical/fd-management.txt: - More documentation cleanups. File management is discussed in file-management.txt so there's no need to have duplicate documentation of it in fd-management.txt. 2005-09-18 12:52 adx * include/hook.h, src/hook.c: - improved /stats h, more useful and prettier - fixed race in fdlist. Fdlist is initialized first, but I just added a callback registration to it, which uses make_dlink_node() and dlink_heap (not initialized yet). Resolved by changing hook.c make_dlink_node's into MyMalloc's and free_dlink_nodes to MyFree. 2005-09-18 11:35 adx * include/fdlist.h, src/fdlist.c: - added a global callback to handle hard_fdlimit change - if the change occurs, notice opers 2005-09-18 10:46 adx * src/: s_bsd_devpoll.c, s_bsd_epoll.c, s_bsd_kqueue.c, s_bsd_select.c: - fixed i/o engines to do at most 128 events at a time (they won't do such much anyway, but that's another story), thus removing HARD_FDLIMIT / getdtablesize() references - still to fix: select, poll, sigio(poll) 2005-09-18 10:27 michael * .indent.pro: - Removed .indent.pro file. Don't get the idea to run the sources thru indent! 2005-09-18 10:26 adx * src/: ircd_parser.y, y.tab.c, y.tab.h: - forgot to recalc_fdlimit when parsing ircd.conf max_clients= 2005-09-18 10:25 adx * configure, configure.ac, contrib/setup-win32.h, include/defaults.h, include/fdlist.h, include/ircd.h, include/m_info.h, include/restart.h, include/setup.h.in, modules/m_rehash.c, modules/m_set.c, modules/core/m_die.c, src/fdlist.c, src/fileio.c, src/ircd.c, src/ircd_parser.y, src/ircd_signal.c, src/listener.c, src/restart.c, src/s_auth.c, src/s_bsd.c, src/s_bsd_win32.c, src/s_conf.c, src/s_serv.c, src/s_user.c, src/y.tab.c, src/y.tab.h: - sanified the whole HARD_FDLIMIT, MAXCONN, MASTER_MAX, MAX_BUFFER, GlobalSetOptions.maxclients, ServerInfo.max_clients mess - now we have 0 hardcoded limits (uhm, maybe hard_fdlimit=32767 on win32) plus 2 runtime settings -- hard_fdlimit obtained from the current ulimit, and ServerInfo.max_clients. /REHASH FDLIMIT can be used to recheck hard_fdlimit (although it is done automatically when you change max_clients) - unified restart/die code - to do: some i/o engines statically assume how many fds they will support on startup, fix them - question: is it possible to change OS limits while ircd is running? account hard limits are already set on the process.. 2005-09-18 08:43 michael * src/channel_mode.c: - Finally fixed ban chaching 2005-09-18 08:07 michael * modules/core/m_mode.c, src/channel_mode.c: - It's possible that a newly added exception matches an already silenced channelmember. Must clear the CHFL_BAN_SILENCED bit here too. 2005-09-18 07:41 michael * include/channel_mode.h, modules/core/m_sjoin.c, src/channel_mode.c: - Must always call clear_ban_cache() when removing a ban/exception, even if quiet_on_ban is disabled, otherwise you would end up with channelmembers having a stale CHFL_BAN_SILENCED flag. Also have to call clear_ban_cache() in case we kill our banlist in a SJOIN. Spotted by adx. 2005-09-18 06:56 michael * include/channel.h, src/balloc.c, src/channel.c, src/channel_mode.c: - Fixed indenting in some places, untabified code etc. 2005-09-17 20:59 knight * contrib/m_services.c: - Code speed clean-ups. I know there's a jumping debate as to whether or not switch/case is truly faster than else/if but in this case I felt it would be in the case of large networks using services. Since else/if's have to run through all conditions this can cause a response time problem with large numbers of queries so switch/case would be better in this instance. Besides I was always taught in college that the use of "hairy ifs" was always bad in this case. 2005-09-17 18:33 adx * src/channel.c: - fixed another lame bug of mine *sigh* 2005-09-17 18:16 adx * src/s_bsd_kqueue.c: - fixed connect not working with kqueue 2005-09-17 16:43 adx * src/channel.c: - fixed another problem with joinflood (integer overflow) 2005-09-17 16:32 adx * src/channel.c: - this probably fixes the problem with ClearJoinFlood not being done on a channel (therefore stopping further flood notices after the first one) 2005-09-17 15:38 adx * configure, configure.ac: - added better epoll detection 2005-09-17 15:11 adx * contrib/help/opers/index, contrib/help/users/index, help/Makefile.in, help/opers/index, help/users/index: - corrected help - still not sure everything is covered 2005-09-17 14:32 knight * contrib/help/Makefile.in, contrib/help/opers/index, contrib/help/users/.lock, contrib/help/users/index, help/Makefile.in, help/opers/index, help/users/index: - Gone back to static index files for help pages. Since help files change very little and new ones are added rarely this isn't a big problem. The commands 'expand' and 'column' caused infinite loop issues on FreeBSD. 2005-09-17 13:56 knight * contrib/help/users/.lock: - Lock users directory so it always downloads. Fixing Makefile issue. 2005-09-17 13:51 knight * contrib/help/Makefile.in: - Restored original Makefile behaviour. The rm -f statements had nothing to do with the Makefile errors. It turns out the errors were related to: echo $$help >> users/index.tmp; Which ONLY occured on Slackware. These leads me to believe that coreutils is horribly broken on Slackware (no surprise). It works everywhere else (Net/Open/FreeBSD, Gentoo, Debian and Red Hat). 2005-09-17 13:28 knight * contrib/help/Makefile.in: - Added 'touch' conditions for distributions with broken coreutils. 2005-09-17 11:56 knight * contrib/help/Makefile.in: - More fixes... 2005-09-17 11:46 knight * contrib/help/Makefile.in: - Removed rm -rf of user.tmp even though UNIX systems are SUPPOSED to force removal even if it doesn't exist, not break. 2005-09-17 11:00 michael * configure, configure.ac, src/ircd_parser.y, src/y.tab.c, src/y.tab.h: - Fixed core on REHASH - Always use -O0 for betas/rcs 2005-09-17 09:51 knight * contrib/Makefile.in: - Minor change. 2005-09-17 09:10 knight * contrib/help/Makefile.in, help/Makefile.in: - Restored column code AGAIN. Metalrock I saw no infinite looping of ANY kind and I tested this thoroughly with different versions GNU make and BSD make as well as calling the Makefiles directly and indirectly. Everything works fine. 2005-09-17 06:30 michael * contrib/ip_cloaking.c, include/s_user.h, src/ircd.c, src/messages.tab, src/s_user.c: - Dynamically rebuild usermode string for RPL_MYINFO when adding/removing a mode 2005-09-17 00:46 metalrock * src/channel_mode.c: - Clean up some of the spacing. 2005-09-17 00:40 metalrock * modules/m_rxline.c: - rxline and unrxline. 2005-09-17 00:39 metalrock * modules/m_rkline.c: - rkline, not kline. 2005-09-16 21:22 metalrock * contrib/help/Makefile.in, help/Makefile.in: - Revert changes to help Makefiles as they run into an infinite at make install. 2005-09-16 18:01 knight * messages/ircd-swedish.lang: - Change e-mail to chatjunkies.org address. 2005-09-16 16:42 adx * src/s_bsd_epoll.c: - fixed epoll 2005-09-16 16:13 michael * src/s_bsd_epoll.c: - Defining _GNU_SOURCE is pointless here 2005-09-16 16:09 adx * src/s_bsd_epoll.c: - this should fix epoll compile 2005-09-16 16:07 knight * src/s_bsd_epoll.c: - Remove duplicate include 2005-09-16 13:44 knight * contrib/help/Makefile.in, help/Makefile.in: - Have the index generate in a column'd style like in 7.0.x for easier reading on IRC clients in 80 char terminals or people who don't like scrolling. 2005-09-16 13:34 knight * contrib/help/Makefile.in: - Merged r1.11 with r1.8. The Makefile.in from r1.8 worked perfectly fine, I do not know what errors Stacy and pi had but they were not Makefile related and the changes made in r1.9 broke the help index. This situation has now been fixed. 2005-09-16 12:24 knight * src/s_bsd_epoll.c: - Fixed compiling errors pertaining to missing epoll_functions(). The issue was was left out of the headers list so s_bsd_epoll.c got confused. 2005-09-16 12:10 knight * src/channel.c: - Another compile error from restrict_channels. 2005-09-16 12:00 knight * modules/core/m_join.c: - Fixed compiling error relating to restrict_channels. *slaps adx* 2005-09-16 11:28 knight * modules/m_info.c: - Added restrict_channels to info struct. 2005-09-16 10:56 adx * etc/example.conf, etc/example.conf.quick, etc/example.efnet.conf, include/s_conf.h, modules/core/m_join.c, src/channel.c, src/ircd_lexer.l, src/ircd_parser.y, src/lex.yy.c, src/s_conf.c, src/y.tab.c, src/y.tab.h: - added channels { restrict_channels = yes/no; }; to reverse RESV logic for closed/corporate networks with a static channel set - fixed oper_pass_resv logic - needs review 2005-09-16 10:27 knight * messages/ircd-dutch.lang: - *oops* :x 2005-09-16 10:21 knight * messages/ircd-dutch.lang: - Updated dutch translation file, probably needs proof-reading. 2005-09-16 10:11 knight * contrib/help/Makefile.in: - Makefile clean-up. The users index temp file was not being removed when it *needs* to be in order to add the contrib user help pages. 2005-09-16 10:07 knight * contrib/help/Makefile.in: - Makefile clean-ups. However for some strange reason the help page FLAGS is not entered into the index for users for some strange reason, but the symlink is created. Someone please review this lol. 2005-09-16 10:07 db * src/ircd_parser.y: - fix my own stupid core 2005-09-16 08:35 knight * modules/m_info.c: - The module m_info has now been completely updated to reflect the current configuration file. Everything is in exact order so if you modify this make sure it's in the right place. Note: We really should re-write how this thing outputs. the current output method is quite ugly and malformed most of the time (e.g. [] never hug the descriptions). So if someone would like to do that be my guest. 2005-09-16 05:39 adx * src/s_bsd_win32.c: - wndhandle MUST NOT be static here, s_bsd.c uses it for set_nonblocking 2005-09-15 21:40 metalrock * TODO: - Update my TODO 2005-09-15 14:03 knight * modules/m_gline.c: - *sigh* 2005-09-15 13:56 knight * modules/m_gline.c: - People who use G-Line voting need to be able to set_local_gline() as well. 2005-09-15 12:50 db * modules/m_gline.c: - set_local_gline was never being called when GLINE_VOTING was disabled this triggered a warning about set_local_gline being defined but not used. 2005-09-15 07:46 adx * RELNOTES, etc/example.conf, etc/example.conf.quick, etc/example.efnet.conf, include/s_conf.h, modules/m_info.c, src/ircd_lexer.l, src/ircd_parser.y, src/lex.yy.c, src/s_conf.c, src/y.tab.c, src/y.tab.h: - got rid of general { maximum_links = ...; }, it didn't work anyway. - default to MAXIMUM_LINKS_DEFAULT (normally 0) if a class{} doesn't specify max_links 2005-09-15 06:43 adx * src/: ircd_parser.y, y.tab.c, y.tab.h: - allow accumulating more ip= entries in an exempt{} block - added missing semicolons in ircd_parser.y 2005-09-15 06:18 adx * src/s_conf.c: - allow MaxTotal(iline->class_ptr) == 0, we already treat this as "no limits" 2005-09-15 05:21 adx * include/defaults.h, src/s_conf.c: - by default a class should have no total clients limit, MAXIMUM_LINKS_DEFAULT = 1 was very confusing 2005-09-15 03:17 adx * src/s_bsd_win32.c: - fixed compile error 2005-09-15 02:27 adx * contrib/win32build.bat: - fixed win32 build after adding epoll i/o 2005-09-14 23:14 metalrock * src/: y.tab.c, y.tab.h: - Update y.tab.h and y.tab.c after parser changes. 2005-09-14 23:08 db * src/ircd_parser.y: - make password encrypted *default* condition. Too many people make mistkaes. 2005-09-14 22:23 knight * TODO: - Semi-simplified my TODO a bit more. 2005-09-14 22:09 metalrock * INSTALL: - Grammar correction. 2005-09-14 21:37 knight * RELNOTES: - Updated RELNOTES 2005-09-14 21:27 metalrock * TODO: - Update TODO since previous commit collided with knight's 2005-09-14 21:25 metalrock * RELNOTES, doc/serverhide.txt, etc/example.conf, etc/example.efnet.conf, modules/m_info.c, modules/m_whois.c, src/channel.c: - Don't show local channels on remote /WHOIS. - Updated documentation. 2005-09-14 21:20 knight * TODO: - metalrock's TODO has been solved, time to get him more work =P 2005-09-14 21:19 knight * doc/: channels.txt, index.txt: - Documention Update: Removed channels.txt; no longer necessary after metalrock squashed his TODO. 2005-09-14 20:47 metalrock * RELNOTES: - Updated RELNOTES 2005-09-14 19:36 metalrock * etc/example.conf, etc/example.conf.quick, etc/example.efnet.conf, include/s_conf.h, modules/m_info.c, modules/m_whois.c, src/ircd_lexer.l, src/ircd_parser.y, src/lex.yy.c, src/s_conf.c, src/y.tab.c, src/y.tab.h: - Updated copyrights. - Added hidden_name to allow customization of hidden servernames. 2005-09-14 13:25 knight * messages/ircd-swedish.lang: - Updated swedish translation file thanks to a fellow operator at ChatJunkies. 2005-09-14 12:29 db * etc/: example.conf, example.efnet.conf: - removed ~ from ~operwall in examples. Too many people getting caught with that and wondering why operwall no longer works. 2005-09-14 10:13 knight * contrib/ip_cloaking.c: - Moved key definitions back to the top of the file since adx feels people might not notice to change them even though they are directed to read the README.cloaking file before doing anything. 2005-09-14 10:01 knight * contrib/: ip_cloaking.c, m_services.c: - Source code and comment clean-ups. 2005-09-14 06:27 adx * configure, configure.ac, src/s_bsd_devpoll.c, src/s_bsd_epoll.c, src/s_bsd_kqueue.c, src/s_bsd_poll.c, src/s_bsd_select.c, src/s_bsd_sigio.c: - make all I/O engines compile (whether they work is another question). - removed unnecessary #includes - please test it on freebsd, linux, linux with --enable-sigio, anything with --enable-poll / --enable-select, solaris, windows 2005-09-14 06:00 adx * configure, configure.ac, include/setup.h.in: - added check for epoll() 2005-09-14 05:47 adx * configure, configure.ac, src/s_bsd_poll.c, src/s_bsd_select.c: - miscellaneous I/O fixes 2005-09-14 05:40 adx * configure, configure.ac: - updated configure to offer --enable-epoll 2005-09-14 05:35 adx * src/s_bsd_poll.c: - fixed compile error in poll 2005-09-14 05:34 adx * src/s_bsd_sigio.c: - fixed sigio 2005-09-13 20:22 knight * src/s_bsd_kqueue.c: - Fixed some compiling errors. 2005-09-13 16:12 knight * messages/ircd-german.lang: - Removed a few numerics that really didn't need to be in the german translation file since there is nothing to be translated Spotted by: Michael 2005-09-13 14:15 adx * include/fdlist.h, src/s_bsd_devpoll.c, src/s_bsd_epoll.c, src/s_bsd_kqueue.c, src/s_bsd_poll.c, src/s_bsd_select.c, src/s_bsd_sigio.c, src/s_bsd_win32.c: - reworked i/o implementations, much cleaner now, fixed some bugs - fixed possible crashes (although very unlikely to happen), we shouldn't dereference fde_t * in the I/O loop, they can be freed by the handlers, so *always* use fd values and lookup_fd() - added epoll - probably many of the I/O providers will not compile now, others might work wrong. We are before an rc, it'd really be nice if someone tested them! 2005-09-13 11:41 knight * messages/ircd-german.lang: - Updated German localisation file thanks to a fellow operator at ChatJunkies. 2005-09-13 10:40 knight * contrib/m_services.c: - Removed ERR_ERRONEUSNICKNAME and RPL_NICKNAMEINUSE numerics. 2005-09-13 04:05 adx * src/s_user.c: * show RPL_YOURID _after_ RPL_ISUPPORT 2005-09-13 01:28 adx * TODO: * one other I missed 2005-09-13 01:24 adx * TODO: - updated my TODO (2) 2005-09-13 01:22 adx * messages/: ircd-polish.lang, ircd-standard.example.lang: * updated translations 2005-09-13 01:18 adx * TODO: - updated my TODO 2005-09-12 23:53 db * include/s_conf.h, src/s_conf.c, src/s_serv.c: - Aren't 3 line functions useless? removed get_con_freq() it was only used in s_serv.c and easier to just do in place. 2005-09-12 18:07 knight * TODO: - Heh missed two other ones I wanted to have. 2005-09-12 17:59 knight * TODO: - Added my list of TODOs for the future. 2005-09-12 16:42 db * src/: ircd_lexer.l, lex.yy.c: - added missing lexer entries for CIDR_BITLEN and NUMBER_PER_CIDR 2005-09-12 16:08 db * src/messages.tab: - This should have been a string ! 2005-09-12 16:04 db * contrib/m_services.c: - Don't do this flood control here, it's not necessary and could core the server if the nick is remote. Spotted by Michael. 2005-09-12 15:47 knight * src/messages.tab: - Missed an extra conversion specifier for RPL_HOSTHIDDEN. Spotted by Michael 2005-09-12 14:17 knight * src/messages.tab: - Forgot to commit messages.tab :x 2005-09-12 14:05 knight * messages/ircd-standard.example.lang: - Match numeric 440 to DALnet's Bahamut string. 2005-09-12 13:50 knight * messages/ircd-standard.example.lang: - Pet peeve commit, line width difference 2005-09-12 13:41 knight * contrib/ip_cloaking.c, include/numeric.h, messages/ircd-standard.example.lang, src/messages.tab: - Numericified hidden host message (assigned as RPL_HIDDENHOST) to numeric 396 which afaik is unused and matches ircu2. 2005-09-12 13:03 knight * contrib/m_services.c: - Header clean ups. - Numericified error messages. - Match clean_nick_name to m_nick.c fixes Reviewed by: Diane Notes: Probably needs more auditing and clean-ups for clean_nick_name since I couldn't figure it out lol. 2005-09-12 09:35 adx * src/hash.c: - oops, accidentally committed my debug code 2005-09-12 09:34 adx * src/: client.c, hash.c: - fixed find_person 2005-09-12 08:04 adx * contrib/m_services.c: - fix it more 2005-09-12 01:59 metalrock * messages/custom.lang: - Add a space after server to make it look nicer. 2005-09-12 01:38 metalrock * src/lex.yy.c: - need to update lex.yy.c after changing ircd_lexer.l 2005-09-12 00:41 adx * modules/core/m_nick.c: - allow remote nicknames begin with a digit, this is VERY important in introducing some protocol changes related to UIDs, collision solving etc. And probably one of the reasons old hybrids were pissing off the ratbox ppl ;) 2005-09-12 00:33 adx * include/hook.h, modules/m_stats.c, src/hook.c: - limited hub/leaf stats to stats H - added /stats h for listing registered callbacks 2005-09-11 23:07 adx * src/client.c: - allow UID addressing for lusers in find_person if the target client is not +i 2005-09-11 22:06 adx * include/s_user.h, src/s_user.c: - turned uid_get into a callback, effectively allowing modules to override the default generator. For instance this makes permanent uids idea quite possible and easy to do in some 100-200 lines module 2005-09-11 21:57 adx * include/numeric.h, messages/ircd-polish.lang, messages/ircd-standard.example.lang, src/messages.tab, src/s_user.c: - imported RPL_YOURID from ircnet, UIDs are now public. - in the future commands should be reworked to accept uids as arguments, and always use them in s2s communication. 2005-09-11 20:24 adx * TODO: - updated TODO, removed some obsolete technical entries etc 2005-09-11 14:33 knight * etc/: example.conf, example.efnet.conf: - Merge 1.83 with 1.81. 2005-09-11 14:28 knight * etc/: example.conf, example.efnet.conf: - Badly written C/N-Line block for IPv6 server example, could cause confusion *sigh*. 2005-09-11 11:59 michael * contrib/m_clearchan.c: - The second list traversal of locmembers is unavoidable. Bleh 2005-09-11 11:38 michael * contrib/m_clearchan.c: - Optimized kick_list() by getting rid of two extra loops 2005-09-11 08:13 knight * src/s_conf.c: - Compile error in s_conf.c fixed. A period was used instead of a comma in a print statement :x 2005-09-11 07:31 michael * src/s_misc.c: - Imported date() bugfix from ircu. TIME would give you a weird looking TZ offset in case lt->tm_year differs from gm->tm_year 2005-09-11 06:44 michael * src/ircd.c: - Fixed typo 2005-09-11 06:40 michael * modules/m_stats.c, src/ircd.c, src/ircd_lexer.l, src/s_conf.c, src/s_user.c: - Improved error reporting here and there, also made error messages a bit more consistent 2005-09-11 04:04 adx * contrib/m_mkpasswd.c, contrib/m_services.c, src/client.c: - fixed compile error in m_services, slight corrections here - use rand() instead of random() in m_mkpasswd, basically because random() generator is not initialized by the ircd and is unsupported on win32. - corrected find_person() prototype 2005-09-10 15:38 adx * contrib/ip_cloaking.c: - do not send CHGHOST if called from register_local_user 2005-09-10 15:08 adx * contrib/m_services.c: - fixed svsnick a bit 2005-09-10 14:45 michael * contrib/win32build.bat: - Don't forget to create kline.conf 2005-09-10 13:44 michael * src/s_bsd_win32.c: - Fixed uname() bug with windows 2003 2005-09-10 10:02 knight * contrib/m_services.c: - acptr -> target_p for consistency 2005-09-10 09:51 michael * src/modules.c: - Small improvement to load_all_modules() 2005-09-10 08:05 michael * src/dynlink.c: - Replaced %lx conversion specifier with %p 2005-09-10 05:48 michael * include/defaults.h, include/modules.h, src/ircd.c, src/modules.c: - Got rid of that stupid set_basename function 2005-09-10 03:46 michael * contrib/help/Makefile.in: - Fixed broken 'make install' in contrib. Reported by Stacy & pi 2005-09-09 17:19 knight * doc/challenge.txt: - Another doc update. The scripting section should only refer about the challange.irc script and tell users where it is. The code should note be posted in the doc as that can cause minor confusion imo. 2005-09-09 16:06 adx * contrib/: README, win32build.bat: - slight corrections 2005-09-09 14:00 knight * contrib/: README, confgen.pgsql, confgen.pl, m_services.c: - More clean ups of SVSNICK; needs more testing obviously - Removed confgen perl script and sql dump as it is useless in 7.2 since it was designed for 7.0.x 2005-09-09 13:55 adx * include/modules.h: - finish previous commit 2005-09-09 13:48 adx * src/modules.c: Fixes for STATIC_MODULES: - moved help+uhelp from contrib to autoload - do not install test_msgtab, silly 2005-09-09 13:43 adx * contrib/win32build.bat: - remove objects in case of failed compile too 2005-09-09 13:38 adx * contrib/win32build.bat: - don't compile example_module.c 2005-09-09 13:37 adx * contrib/example_module.c, contrib/m_change.c, contrib/m_ltrace.c, contrib/setup-win32.h, contrib/spy_admin_notice.c, contrib/spy_info_notice.c, contrib/spy_links_notice.c, contrib/spy_motd_notice.c, contrib/spy_stats_notice.c, contrib/spy_stats_p_notice.c, contrib/spy_trace_notice.c, contrib/spy_whois_notice.c, contrib/win32build.bat, src/modules.c: - fixed contrib build with STATIC_MODULES - allow win32 lusers to compile contrib 2005-09-09 13:10 adx * contrib/ip_cloaking.c, src/modules.c: - fixed compile error in contrib/ip_cloaking.c when IPv6 is not available - fixed compile with STATIC_MODULES && BUILD_CONTRIB 2005-09-09 00:40 adx * src/hash.c: - random() generator is never initialized, use rand() 2005-09-09 00:10 adx * src/ircd.c: - minor simplification 2005-09-08 22:40 adx * contrib/win32build.bat: - do not copy files like .cvsignore, Makefile.in 2005-09-08 22:34 adx * contrib/win32build.bat: - fixed a stupid 2005-09-08 21:12 knight * src/csvlib.c: - Removed the last HAVE_REGEX_H from the code. 2005-09-08 20:50 knight * include/modules.h, src/modules.c: - Added missing msgtabs for rkline, sid_msgtab and uid_msgtab - Removed #ifdef for REGEX_H - Please review these changes... 2005-09-08 20:07 adx * modules/core/m_message.c: - fixed core on every /msg or /notice when compiled with static modules. 2005-09-08 19:56 adx * modules/core/m_sjoin.c: - fixed 99% cpu problem that was general to hybrid but showed up on win32 because of another platform-specific bug 2005-09-08 15:35 knight * include/send.h: - #ifdef 0 -> #if 0 2005-09-08 13:50 adx * contrib/win32build.bat: - create c:\ircd on install (md seems to do it automatically on winnt) 2005-09-08 13:47 lusky * RELNOTES: another test to force a snapshot 2005-09-08 13:41 lusky * RELNOTES: test commit to force a snapshot 2005-09-08 13:37 lusky * RELNOTES: stub out beta2 in relnotes 2005-09-08 11:44 michael * include/send.h: - Undef AFP for releases 2005-09-08 11:02 adx * contrib/win32build.bat: - improved it a bit 2005-09-08 00:50 adx * modules/m_info.c: - its now runtime information 2005-09-08 00:47 adx * contrib/win32build.bat: - the same about help/ and etc/ 2005-09-08 00:43 adx * src/s_bsd_win32.c: - fixed a stupid causing a core under w32 2005-09-08 00:36 adx * contrib/win32build.bat: - always copy the fresh binary in place 2005-09-08 00:29 adx * contrib/win32build.bat: - removed useless border at the very end 2005-09-08 00:14 adx * modules/m_info.c: - fixed info 2005-09-08 00:02 adx * modules/m_info.c: - fixed core at oper /info 2005-09-07 23:52 metalrock * contrib/m_services.c: - Simplify the help message. 2005-09-07 23:45 lusky * contrib/win32build.bat: - fix setup.h creation - fix build error handling 2005-09-07 23:10 metalrock * src/s_log.c: - Fix compile warning. 2005-09-07 23:02 metalrock * include/s_conf.h, src/ircd.c: - Fix missing declarations. 2005-09-07 22:54 adx * contrib/win32build.bat: - minor change 2005-09-07 22:46 adx * lib/: .cvsignore, pcre/.cvsignore: - added cvsignore entries for pcre Makefile 2005-09-07 22:43 adx * contrib/ip_cloaking.c, src/channel.c, src/csvlib.c, src/fdlist.c, src/s_log.c: - fixed compiler warnings - fixed core in csvlib - fixed a core on unloading ip_cloaking 2005-09-07 22:32 adx * README.PLATFORMS: - sorted README.PLATFORMS - removed some obsolete entry 2005-09-07 22:30 knight * README.PLATFORMS: - Changed SPARC Linux maintainer to myself. 2005-09-07 22:27 adx * configure, configure.ac: - default to -O2 in CFLAGS 2005-09-07 21:34 adx * Hybrid-team, src/version.c: - moved kire down to the contributors list, as I see only two his commits concerning one thing in the ChangeLog. 2005-09-07 21:07 adx * contrib/win32build.bat: - miscellaneous corrections to the build script - after downloading the binaries and installing the compiler, they now have to do 2 things: 1) make sure the compiler BIN directory is in PATH 2) in case of Borland C++ Compiler 5.5, they need to prepare the BCC32.CFG file, if not already created 2005-09-07 20:59 adx * contrib/: setup-win32.h, win32build.bat: - rewritten win32 build script to be more clicker friendly - also rearranged setup-win32.h a bit as it is going to be human edited 2005-09-07 18:51 adx * contrib/win32build.bat: - fix it more 2005-09-07 18:45 adx * contrib/win32build.bat: - build pcre too 2005-09-07 16:44 knight * INSTALL: - Add 'touch' instructions for the extra logging that Diane implemented earlier this week. - Minor cosmetics to fit in a 80 char terminal. 2005-09-07 16:39 knight * INSTALL: - Reversion Merge... 2005-09-07 16:28 knight * INSTALL: - Update Copyright Header - Added missing ./configure knobs to the file - Fix URL to ircd-hybrid forums. - Update installation instructions (log files do not require to be touch'd anymore) 2005-09-07 14:04 adx * src/s_bsd.c: - this should fix the core on squitting a /jupe'd server 2005-09-07 12:59 db * include/s_conf.h, src/ircd_parser.y, src/lex.yy.c, src/y.tab.c, src/y.tab.h: - first part of adding CIDR_BITLEN and NUMBER_PER_CIDR 2005-09-07 12:40 adx * contrib/Makefile.in, include/s_conf.h, src/ircd_lexer.l, src/lex.yy.c: - make it compile again after Diane and Michael broke it ;P 2005-09-07 12:16 knight * doc/: operguide.txt, opermyth.txt: - Removed opermyth.txt and operguide.txt. These are EFNet specific docs that really do not belong in the ircd-hybrid documentation. I already checked and they are on EFNet's website. IRC network specific documentation really does not belong in the documentation distributed with ircd-hybrid in my personal opinion and if I get yelled at for that, so be it. 2005-09-07 11:08 knight * contrib/README: - Removed reference to Unreal IRCd to please Michael =P 2005-09-07 11:03 knight * contrib/README: - Fixed screwed up text caused by VIM's filters *kicks vim*. Also removed /MAP from the README at the request of adx. 2005-09-07 10:57 knight * contrib/README: - Conform to 80 character requirement 2005-09-07 10:51 knight * contrib/README.cloaking: - Updated cloaking NOTICE example to reflect current IP cloaking masking. 2005-09-07 10:49 knight * contrib/README.cloaking: - Spelling and grammar fixes. Adjusted document to conform to 80 character requirement stated in guidelines.txt 2005-09-07 10:20 knight * modules/m_stats.c: - Minor cosmetic change to my code additions. 2005-09-07 09:45 michael * include/s_conf.h, modules/m_info.c: - Fixed a bug where we wouldn't show stats_i_oper_only and stats_k_oper_only in /INFO - Fixed the rest of the m_info.c compile errors 2005-09-07 09:08 michael * include/: msg.h, s_conf.h: - Unbroke m_info.c - Removed useless definition in msg.h 2005-09-07 09:05 knight * modules/m_info.c: - OUTPUT_BOOLEAN -> OUTPUT_BOOLEAN_YN for opers_bypass_callerid. 2005-09-07 08:58 knight * contrib/m_services.c: - Changed the behaviour of SVSNICK to more simulate the working version I have on ChatJunkies. However the function still needs a code audit by another pair of eyes. 2005-09-07 08:48 knight * modules/m_stats.c: - More changes to STATS g. Originally I had STATS g just disabled all together if G-Ling voting was not enabled. However it occured to me this would cause confusion so I changed the behaviour of STATS g to where if an operator or administrator issues STATS g to see pending G-Lines but G-Line voting is disabled it will post a NOTICE saying "This server does not support G-Line voting" and then the usual end of stats report. This is more organised than disabling STATS g all together. 2005-09-06 22:21 adx * etc/example.conf, etc/example.conf.quick, etc/example.efnet.conf, include/s_conf.h, src/ircd_lexer.l, src/ircd_parser.y, src/lex.yy.c, src/s_conf.c, src/s_user.c, src/y.tab.c, src/y.tab.h: - added invisible_on_connect = yes/no to general{} - changed boolean int's to char:1 - removed useless if ypass == 2 checks in ircd_parser.y 2005-09-06 20:50 knight * modules/m_stats.c: - Forgot to encompass the prototype and function for stats_pending_glines in the #define for GLINE_VOTING knob :x 2005-09-06 13:35 adx * src/s_bsd_win32.c: - ifdef'd PROCESSOR_ARCHITECTURE_AMD64, it is not defined on older compilers 2005-09-06 11:51 adx * contrib/setup-win32.h, contrib/win32build.bat, modules/m_stats.c, src/client.c, src/s_log.c, src/s_stats.c: - fixed win32 build 2005-09-06 11:30 adx * src/s_bsd_win32.c: - shut down the daemon gracefully on win32 2005-09-06 10:39 db * src/lex.yy.c: - update lex.yy.c 2005-09-06 10:36 db * src/ircd_lexer.l: - I reversed the test 2005-09-06 09:08 michael * src/: dynlink.c, modules.c: - Go rid of num_mods by simply using dlink_list_length 2005-09-06 07:54 db * src/s_log.c: - Don't try to do any logging if fb is NULL i.e. not readable. 2005-09-06 07:35 michael * modules/m_gline.c: - That one should be an #ifdef. Spotted by lee 2005-09-05 22:25 db * src/: s_bsd.c, s_log.c: - report to io_errlog notice that log_oper_action() is now a misnomer in this case - Don't try to fill in the non existent nick!user@host in this case 2005-09-05 19:45 adx * configure, configure.ac: - fixed gline-voting knob 2005-09-05 19:38 knight * configure, configure.ac, include/setup.h.in, modules/m_gline.c, modules/m_stats.c: - Added configure option --enable-gline-voting to toggle G-Line voting. it configures and compiles but needs testing. 2005-09-05 15:39 knight * modules/m_info.c: - Added opers_bypass_callerid to the table. 2005-09-05 14:49 db * include/s_log.h, src/s_log.c: - turn struct Client * into const struct Client 8 in log_oper_action() 2005-09-05 13:38 db * src/: ircd_parser.y, lex.yy.c, s_serv.c, y.tab.c, y.tab.h: - adds back the vhost per connect block that we somehow lost going from hybrid-6 to hybrid-7 2005-09-05 11:37 db * include/s_conf.h: - removed two variables that were moved into ServerInfo months ago 2005-09-05 10:53 db * include/s_conf.h, include/s_log.h, src/ircd_lexer.l, src/ircd_parser.y, src/lex.yy.c, src/s_conf.c, src/y.tab.c, src/y.tab.h: - make sure operspylog and ioerrlog directives work, though they are dummied out for now. 2005-09-05 10:16 db * modules/m_challenge.c, modules/m_oper.c, src/s_log.c: - missed the terminating \n for log_oper_action in m_oper.c and m_challenge.c - forgot a () around the +1 , this could have made it overflow 2005-09-05 08:08 db * src/s_log.c: - minor comment tweak, correct input comments to log_oper_action() 2005-09-05 08:03 db * etc/example.conf, include/s_conf.h, include/s_log.h, modules/m_challenge.c, modules/m_gline.c, modules/m_kline.c, modules/m_oper.c, modules/core/m_kill.c, src/csvlib.c, src/ircd_lexer.l, src/ircd_parser.y, src/lex.yy.c, src/s_conf.c, src/s_log.c, src/y.tab.c, src/y.tab.h: - Pretty much revamped the entire logging subsystem for now. added independent file logging for kline, gline, kill added fkilllog, fname_killlog, fklinelog, fname_klinelog, fglinelog, fname_glinelog to the logging block logging{} made log a synonym for logging i.e. log {} is now synonymous with logging {}. Made synonyms for fuserlog == fname_userlog foperlog == fname_operlog. Generalised logging function reworked log_oper and log_failed_oper to use log_oper_action all additional logging uses log_oper_action Now, I must have more sleep. 2005-09-05 07:19 db * lib/pcre/Makefile.in: - add dummy install: to make make install not complain for now 2005-09-05 05:45 db * contrib/m_services.c: - *FINALLY* have proper credits direct from 'bane' himself 2005-09-04 15:13 db * contrib/m_services.c: - Considering kre didn't write GetString in the first place, no sense insulting him. - Apparently, sidewndr (the original coder of hybserv) wrote GetString - The total file came via bane not kre, or something. The hybserv guys ripped banes nick/name out of the code for some reason. I hope I got the story right this time. 2005-09-04 10:48 adx * modules/core/m_sjoin.c: - fixed possible core / protocol corruption on ts5 links Spotted by: anfl 2005-09-04 08:08 michael * include/s_conf.h, modules/m_stats.c, src/s_conf.c: - Fixed incorrect STATS k|K output with temporary rklines 2005-09-04 07:49 michael * contrib/setup-win32.h: - Update setup-win32.h 2005-09-04 05:17 michael * src/balloc.c: - Fixed an invalid assert in BlockHeapFree() which would -always- be TRUE Only data pointers to our working adress space are linked into the used_list, not a data pointer to a MemBlock itself. BTW: the used_list can be removed in the future. We don't really need it. I leave it for now for debugging purpose 2005-09-04 04:41 michael * include/control.h, include/io.h, include/servlink.h, servlink/Makefile.in, servlink/control.h, servlink/io.h, servlink/servlink.h: - Moved servlink headerfiles back into their servlink folder where they actually belong to. servlink is a standalone program and the headers have nothing todo in include/ 2005-09-03 16:14 michael * contrib/win32build.bat: - Updated win32build.bat 2005-09-03 15:57 michael * modules/m_stats.c, src/hostmask.c, src/messages.tab, src/s_conf.c: - Updated RPL_STATSKLINE and RPL_STATSXLINE as discussed on irc - Style corrections 2005-09-03 11:23 michael * help/opers/rxline: - Fixed typo 2005-09-03 11:21 michael * help/opers/: rkline, rxline: - Updated rxline and rkline help files 2005-09-03 08:30 michael * configure, configure.ac, include/setup.h.in, src/s_conf.c, src/s_user.c: - Fixed an older bug where we wouldn't test new rxlines against active connections - Removed regex.h configure check 2005-09-03 07:39 michael * lib/pcre/Makefile.in, lib/pcre/pcre_internal.h, lib/pcre/pcre_xclass.c, src/irc_string.c: - We now don't longer enforce caSelEsS matching. People can now specify the ?i option within the mask if they whish to have a caseless match. Just to give them more freedom .. 2005-09-03 06:04 michael * lib/pcre/: Makefile.in, config.h, pcre.h, pcre_compile.c, pcre_globals.c, pcre_internal.h: - Another round of pcre updates 2005-09-03 04:57 michael * lib/pcre/Makefile.in, lib/pcre/pcre.h, lib/pcre/pcre_compile.c, lib/pcre/pcre_config.c, lib/pcre/pcre_exec.c, lib/pcre/pcre_internal.h, lib/pcre/pcre_ord2utf8.c, lib/pcre/pcre_tables.c, lib/pcre/pcre_ucp_findchar.c, lib/pcre/pcre_valid_utf8.c, lib/pcre/pcre_xclass.c, lib/pcre/ucp.h, lib/pcre/ucp_findchar.c, lib/pcre/ucpinternal.h, lib/pcre/ucptable.c, modules/m_rkline.c, modules/m_rxline.c, src/csvlib.c: - Must use strcmp() when looking up RKLINES or RXLINES for addition/deletion in order to play nice with metacharacters. - Slimmed down pcrelib some more 2005-09-03 02:34 michael * lib/pcre/: Makefile.in, pcre.h, pcre_compile.c, pcre_dfa_exec.c, pcre_get.c, pcre_internal.h, pcre_printint.c, pcre_refcount.c, pcre_version.c: - Removed files not required for ircd 2005-09-03 02:05 michael * Makefile.in, configure, configure.ac, help/opers/rxline, include/irc_string.h, include/s_conf.h, include/stdinc.h, lib/Makefile.in, lib/pcre/AUTHORS, lib/pcre/LICENCE, lib/pcre/Makefile.in, lib/pcre/README, lib/pcre/config.h, lib/pcre/pcre.h, lib/pcre/pcre_chartables.c, lib/pcre/pcre_compile.c, lib/pcre/pcre_config.c, lib/pcre/pcre_dfa_exec.c, lib/pcre/pcre_exec.c, lib/pcre/pcre_fullinfo.c, lib/pcre/pcre_get.c, lib/pcre/pcre_globals.c, lib/pcre/pcre_internal.h, lib/pcre/pcre_ord2utf8.c, lib/pcre/pcre_printint.c, lib/pcre/pcre_refcount.c, lib/pcre/pcre_study.c, lib/pcre/pcre_tables.c, lib/pcre/pcre_try_flipped.c, lib/pcre/pcre_ucp_findchar.c, lib/pcre/pcre_valid_utf8.c, lib/pcre/pcre_version.c, lib/pcre/pcre_xclass.c, lib/pcre/ucp.h, lib/pcre/ucp_findchar.c, lib/pcre/ucpinternal.h, lib/pcre/ucptable.c, modules/Makefile.in, modules/m_rkline.c, modules/m_rxline.c, modules/m_xline.c, src/Makefile.in, src/csvlib.c, src/irc_string.c, src/ircd_parser.y, src/s_conf.c, src/s_user.c, src/y.tab.c, src/y.tab.h: - Added PCRE lib 6.3 hacked up for use in ircd-hybrid 2005-09-01 22:54 db * src/s_user.c: - Really William (nenolod), this is really nitpicky. I have known about this "bug" since I wrote it, the original "AAAAAA" just looked *PRETTIER* 2005-09-01 21:37 db * src/channel.c: - moved update of last join time into the if(), Piotr and I are really not sure if it really matters. ;-) 2005-09-01 19:18 db * src/: channel.c, ircd.c: - don't let it core if joinfloodcount is set to 0! 2005-09-01 18:58 db * src/: client.c, ircd.c, ircd_lexer.l, ircd_parser.y, lex.yy.c, y.tab.c, y.tab.h: - add /quote set options for JFLOODTIME, JFLOODCOUNT and REJECTTIME (REJECT_HOLD) - check REJECT_HOLD in unknowns check 2005-08-31 07:47 db * include/client.h, modules/m_set.c, src/client.c, src/s_conf.c, src/s_user.c: - move reject_delay to localClient - change /set REJECTTIME to be in seconds rather than minutes 2005-08-31 07:05 adx * RELNOTES, TODO: - updated RELNOTES and TODO 2005-08-30 22:13 db * include/client.h, include/ircd.h, include/ircd_defs.h, modules/m_set.c, src/client.c, src/s_conf.c, src/s_user.c: - re-added REJECT_HOLD but using capture 2005-08-30 21:11 db * include/channel.h, modules/core/m_message.c, src/channel.c: - replace two ints in a struct Channel with one flag 2005-08-30 17:57 db * src/s_conf.c: - find_user_host was missing a return 1, which meant, no one could kline by nick. 2005-08-30 14:28 adx * contrib/m_clearchan.c, contrib/m_force.c, contrib/m_ojoin.c, contrib/m_operspy.c, include/channel.h, include/ircd.h, modules/m_lljoin.c, modules/m_set.c, modules/core/m_join.c, modules/core/m_sjoin.c, src/channel.c, src/ircd.c: - added join flood notification to opers - code by Dianora and me - needs a bit of testing 2005-08-30 12:20 knight * src/version.c: - Added nenolod to contributors. 2005-08-30 07:42 michael * include/restart.h, modules/core/m_die.c, src/ircd_signal.c, src/restart.c: - Flush and close all connections on signal SIGTERM. Added new function server_die() for this 2005-08-29 23:54 db * modules/core/m_kill.c: - remove a core that happened It is *possible* for an uid *not* to be found in the hash If it is not in the hash table, it should be silently ignored and no history tried on it. 2005-08-29 22:31 metalrock * contrib/Makefile.in: - make depend before make 2005-08-29 17:02 db * Hybrid-team, include/numeric.h, modules/m_whois.c, modules/core/m_kill.c, modules/core/m_message.c, src/client.c, src/messages.tab, src/parse.c: - applied nenolod's notarget patch to remove the NOTARGET numeric the NOTARGET numeric could leak uids and is non TS6 standard - added nenolod to contributors 2005-08-29 14:01 db * src/m_error.c: - back out if (MyConnect()) for now. After discussion with Piotr, servers may wish to give more than one error. 2005-08-29 07:22 adx * include/client.h, src/client.c: - fixed ping_warning 2005-08-29 05:31 adx * RELNOTES, src/messages.tab: - updated RELNOTES - updated 004 to contain +D 2005-08-29 05:14 adx * src/client.c: - expecting a PONG from a pending comm_connect_tcp (IsConnecting) looks silly to me. 2005-08-29 05:10 adx * etc/example.conf, etc/example.conf.quick, etc/example.efnet.conf, include/s_conf.h, src/client.c, src/ircd_lexer.l, src/ircd_parser.y, src/lex.yy.c, src/s_conf.c, src/y.tab.c, src/y.tab.h: - added class{} option, ping_warning. If a server doesn't respond to PING in this amount of time, opers are notified of the lag / 99% CPU etc. - totally untested, as usual. :) 2005-08-27 23:51 db * src/m_error.c: - This is a *very* old bug. If a *server* gives an error, then it has to be exited! A "MyClient()" won't exit a server. This bug is in 7.1.x tree as well. 2005-08-27 21:49 db * modules/core/m_message.c: - added message hooks, useful for adding colour filtering, anti spambot filtering etc. Also could be used for nefarious purposes. oh well. Written by nenolod, William Pitcock Reviewed by Alan and Diane 2005-08-27 10:18 knight * contrib/: m_change.c, m_services.c: - Last minute fixes, CHGHOST, CHGIDENT, CHGNAME and SVSNICK should encap.. 2005-08-26 21:47 knight * contrib/m_change.c: - Moved the CHG* commands from the encapsulation tab to the server tab which caused services not to work with CHGHOST *slaps adx* 2005-08-26 17:23 knight * etc/example.efnet.conf: - Same stale option that existed in example.conf existed here as well. Fixed. 2005-08-26 17:17 knight * etc/example.conf: - Removed hidden_oper = yes; as it caused a parse issue. It's actually an option for flags, not it's own individual option. 2005-08-26 11:03 knight * contrib/m_services.c: - Cosmetic code clean ups. Someone however should talk to kreator, his use of SVSNICK in NickServ still does not work with this module however the OperServ on Anope works with it just fine so I cannot figure it out. 2005-08-26 09:42 db * include/channel_mode.h, src/channel.c, src/channel_mode.c: - cache ban code for quiet on ban cache the result of is_banned() when quiet on ban is set. 2005-08-25 13:39 adx * contrib/ip_cloaking.c: - fixed a nasty logic bug in ip_cloaking module which caused us to reuse a umode flag instead of finding an unused one. 2005-08-25 13:14 db * src/channel.c: - added a missing comment block, may as well use doxify style. 2005-08-25 11:13 adx * src/s_bsd.c: - set ip type of service to 'Minimize Delay' on all ircd sockets 2005-08-25 06:16 michael * Hybrid-team, src/version.c: - Removed extra comma on bane's name - Updated version.c 2005-08-25 05:45 db * Hybrid-team: - updated bane's entry 2005-08-23 14:00 db * include/s_auth.h, src/s_auth.c: - make release_auth_client() public so an auth callback can use it 2005-08-23 10:06 db * src/client.c: - back out my accidental commit adding if ... source_p != &me test it is not needed. 2005-08-22 16:55 db * src/s_conf.c: - removed more duplicate code 2005-08-22 09:56 michael * modules/m_resv.c: - Cosmetic fix to parse_resv(). s/xline/resv/ 2005-08-22 06:48 michael * include/s_gline.h, src/s_misc.c: - Removed unused prototypes in s_gline.h - Fixed compiletime warnings in ssl_get_cipher() 2005-08-21 13:22 knight * etc/example.conf: - Minor format issue :x 2005-08-21 13:19 knight * etc/example.conf, include/s_conf.h, modules/core/m_message.c, src/ircd_lexer.l, src/ircd_parser.y, src/lex.yy.c, src/s_conf.c, src/y.tab.c, src/y.tab.h: - Added configuration option "opers_bypass_callerid" to general {} to allow operators to message anyone even if they are +g. Disabled by default. 2005-08-21 05:33 adx * contrib/: m_change.c, m_services.c, m_spoof.c: - "You have no admin flag" => ERR_NOPRIVS 2005-08-21 03:33 michael * src/: version.c, version.c.SH: - Killed version.c.SH and updated copyright header in version.c 2005-08-20 19:48 adx * include/packet.h, include/send.h, src/ircd.c, src/packet.c, src/send.c: - redid hooking method for io_debug.c - ircd is currently creating the callback and io_debug should attach to it - added init_callbacks() to ircd.c - should probably move more of them there, since they are currently sitting in rather inappropriate places, like init_uid (set_user_mode) etc - Diane, please update io_debug.c to match these changes 2005-08-20 19:05 knight * src/version.c: - Minor update; version.c didn't fully reflect the Hybrid-Team list. 2005-08-20 18:59 knight * Hybrid-team: - Moved myself from contrib to core 2005-08-20 14:12 knight * help/opers/rkline: - Stale text from rxline 2005-08-20 13:52 michael * Makefile.in, configure, configure.ac, include/Makefile.in: - Removed include/Makefile.in 2005-08-20 13:35 michael * contrib/: Makefile.in, m_classlist.c: - Fixed compile warning in m_classlist.c 2005-08-20 13:34 knight * etc/Makefile.in: - Add in condition to touch rkline.conf. 2005-08-20 13:19 adx * contrib/ip_cloaking.c, contrib/spy_admin_notice.c, contrib/spy_info_notice.c, contrib/spy_links_notice.c, contrib/spy_motd_notice.c, contrib/spy_stats_notice.c, contrib/spy_stats_p_notice.c, contrib/spy_trace_notice.c, contrib/spy_whois_notice.c, include/hook.h, src/hook.c: - redefined pass_callback as a vararg routine - therefore fixed the bug where we used a stale va_list pointer with pass_callback() 2005-08-20 12:49 michael * modules/m_rkline.c, src/client.c, src/s_conf.c: - Fixed rklines and some older rxline bugs 2005-08-20 12:17 knight * help/opers/rkline: - Another trivial update because Michael told me something as I did the commit *argh* lol 2005-08-20 12:17 knight * help/opers/rkline: - Trivial Update... 2005-08-20 12:08 knight * help/opers/rkline: - Add missing help page for RKLINE 2005-08-20 12:08 knight * contrib/help/: Makefile.in, users/flags: - Removed user help page for flags and have it symlink from opers 2005-08-20 11:03 michael * modules/m_kline.c, modules/m_rkline.c, modules/m_rxline.c, modules/m_xline.c, src/s_conf.c: - Fixed UNRKLINE not working for admins - Removed redundant remoteban privileges check in m_unkline and m_unxline. parse_aline() already does that job for us. - Fixed bug where we wouldn't remove local x-lines if "UNXLINE ON *mask*" matches our server 2005-08-20 11:00 knight * doc/: Tao-of-IRC.940110, index.txt, kline.txt, messages.txt, server-version-info, whats-new.txt, technical/README.TSora, technical/hostmask.txt, technical/index.txt, technical/rfc1459.txt, technical/rfc2812.txt, technical/rfc2813.txt, technical/ts3.txt: - Severe documentation spring cleaning 2005-08-20 10:00 michael * src/s_conf.c: - parse_aline(): increase user and host buffer sizes to play nice with regexp klines and to allow standard klines like *u*s*e*r*n*a*m*e*@someho.st 2005-08-20 08:02 michael * src/: ircd_parser.y, s_conf.c, y.tab.c, y.tab.h: - Fixed RKLINEs 2005-08-20 01:16 michael * contrib/: m_change.c, m_spoof.c: - Added $Revision CVS kw to m_spoof.c and m_change.c 2005-08-20 01:04 michael * contrib/m_classlist.c: - m_classlist.c: there may be more than one class matching the given pattern so don't return after the first reported class. 2005-08-20 00:43 michael * contrib/m_capture.c: - Misc. CAPTURE fixes 2005-08-19 22:52 knight * contrib/README: - Severe spring cleaning of contrib README. Needed to be updated as it had old or vague information. Probably needs to be proof read before 7.2 release though. 2005-08-19 14:24 knight * contrib/: README, confgen.pgsql, confgen.pl: - Added new version of confgen.pl - Added PgSQL table dump for use with confgen.pl (TODO: Add MySQL dump) - Added confgen.pgsql to README 2005-08-19 14:07 db * src/s_user.c: - trivial addition of "at" in between date and time of RPL_CREATED 2005-08-19 13:24 db * modules/m_gline.c: - simplify mo_gline() I couldn't resist 2005-08-19 13:19 db * src/s_user.c: - restore 003 numeric using built in __DATE__ __TIME__ which appears to be a 'c' standard. 2005-08-19 13:06 michael * etc/example.conf, etc/example.efnet.conf, src/ircd_parser.y, src/s_conf.c, src/y.tab.c, src/y.tab.h: - Added regex support for kill{} blocks - Fixed bug where we would leave an invalid auth{} item on the auth_item list when there is no user= specified in auth{} 2005-08-19 11:29 db * src/: fdlist.c, s_bsd.c: - fixed two off by one errors, which could cause a core or at least cause an assert() to fail. - Spotted by Piotr 2005-08-19 10:56 michael * include/defaults.h, include/s_conf.h, modules/Makefile.in, modules/m_rkline.c, modules/m_stats.c, src/csvlib.c, src/ircd.c, src/s_conf.c: - Added support for regular expression based klines. Nothing tested at all, but it compiles for now. NOTE: It's an admin-only command since there is no sane way to check for u@h masks being too broad. 2005-08-19 06:54 adx * include/fdlist.h, src/s_bsd.c: - fixed possible core 2005-08-19 00:36 lusky * configure, configure.ac: - flip order of SSL_LIBS to make solaris happy - change tail back to the deprecated form that still works on everything 2005-08-19 00:29 lusky * include/ircd.h, src/hash.c: - fix missing prototype for get_ircd_platform() - fix collision for random_key with older openssl 2005-08-19 00:10 lusky * src/version.c: fix missing include 2005-08-19 00:07 lusky * contrib/Makefile.in: clean & distclean for contrib/help 2005-08-19 00:00 lusky * contrib/setup-win32.h, contrib/win32build.bat, include/ircd.h, include/ircd_defs.h, include/s_bsd.h, modules/m_info.c, src/.cvsignore, src/Makefile.in, src/ircd.c, src/lex.yy.c, src/s_bsd_win32.c, src/s_user.c, src/version.c: - eradicate version.c.SH stuff, replace with uname() data - wrap unistd.h include from lex.yy.c with autoconf ifdefs - add stuff to build with VC7 2005-08-18 17:53 knight * contrib/ircd-hybrid.spec: - ircd-hybrid should install in /usr/local/ircd not /usr/ircd to satisfy LFHS 2005-08-18 17:38 knight * contrib/help/Makefile.in: - Last minute fix, left out important echo statements by mistake :x 2005-08-18 17:31 knight * contrib/help/Makefile.in, help/Makefile.in: - Updated contrib help pages Makefile, no longer clobbers core help pages! - Moved sort commands in both Makefiles to occur *after* the sed stripping. - Thoroughly tested, both work. 2005-08-18 16:53 knight * src/s_bsd.c: - fde_t *fd_next_in_loop = NULL; was missing. 2005-08-18 15:48 adx * src/: fdlist.c, s_bsd.c: - fixed core in comm_check_timeouts. We were traversing the fd hash bucket by something like that: for (hashv = 0; hashv < MAX_HASH; hashv++) for (F = fd_hash[hashv]; F != NULL; F = F->hnext) The problem is, F could get freed in the middle by timeout handlers. This happened for example on ident timeout. - There are probably similar bugs in s_bsd.c 2005-08-18 14:36 adx * src/listener.c: - fixed compile error, spotted by knight- 2005-08-18 14:21 knight * contrib/help/Makefile.in, help/Makefile.in: - Removed stale credits help page that no longer exists. - TODO: For some reason the contrib help Makefile is screwing up. The the index from the core help pages is being removed and re-written and opers help pages are bleeding into the users index. Needs fixing *ASAP* 2005-08-18 13:24 adx * src/s_auth.c: - send REPORT_FAIL_ID if an ident server closes the connection 2005-08-18 13:21 adx * include/s_bsd.h, src/irc_getaddrinfo.c, src/listener.c, src/packet.c, src/s_auth.c, src/s_bsd.c, src/send.c: win32 build fixes: - initialize errno properly when used - s/close/closesocket/ other fixes: - ircd was trying to parse an uninitialized buffer when a hard read error occured in read_auth_reply 2005-08-18 13:03 db * src/s_auth.c: - make this if match the one Piotr just committed. Same deal, critical region but my initial code was for debugging purposes. 2005-08-18 12:26 adx * src/s_auth.c: - fixed a relatively new (cvs only) race condition in auth code. Basically comm_connect_tcp can call comm_connect_callback *immediately*, if an instant connect() failure happens. Then, auth_error would free the auth struct, before DNS reply was ever received. - this should prevent win32 build from crashing 2005-08-18 11:53 knight * contrib/help/: Makefile.in, opers/spoof, opers/uncapture: - Added missing help page for UNCAPTURE. - Revised SPOOF help page to include requirement of .include "spoof.conf" 2005-08-18 11:47 knight * contrib/help/: Makefile.in, opers/flags, opers/svsnick, users/flags: - Added the last of the missing help pages for FLAGS and SVSNICK 2005-08-18 09:58 knight * configure: - Rebuilt configure script for contrib help Makefile 2005-08-18 09:23 knight * contrib/help/: Makefile.in, opers/capture, opers/classlist, opers/ctrace, opers/delspoof, opers/etrace, opers/ltrace, opers/spoof: - Added more missing contrib help pages. We still need one for m_flags.so but I do not know enough about how it works to write one. 2005-08-18 08:56 knight * configure.ac: - contrib help pages Makefile 2005-08-18 08:52 knight * configure, contrib/Makefile.in, contrib/help/Makefile.in: - Added Makefile for contrib help pages. They will now instal when 'make install' is issued inside contrib/. We should probably add the missing contrib help pages now. 2005-08-18 02:37 db * include/fdlist.h, include/fileio.h, include/hook.h, src/packet.c, src/send.c: - update io_debug hooks to use new hook format - add HOOK_V2 and FILEIO_V2 #defines to hook.h and fileio.h to allow io_debug.c to be backward and forward compatible for now. There has to be a better solution than this... 2005-08-18 00:55 db * src/hook.c: - add missing irc_string.h header 2005-08-18 00:45 db * modules/m_whois.c: - rudimentary comment added for now 2005-08-18 00:32 lusky * src/s_bsd_win32.c: s/dlink_add/dlinkAdd 2005-08-18 00:31 lusky * src/hook.c: s/strcasecmp/irccmp to make win32 happy 2005-08-17 23:42 lusky * src/send.c: #ifdef out even more broken io_debug hook code for now, restore later 2005-08-17 23:41 lusky * src/: packet.c, send.c: #ifdef out more broken io_debug hook code for now, restore later 2005-08-17 23:36 lusky * modules/m_whois.c: Another missing va_list 2005-08-17 23:08 db * modules/m_whois.c: - missing va_list args; for STATIC_MODULES 2005-08-17 22:40 db * include/modules.h, src/packet.c: - #ifdef out broken io_debug hook code for now, restore later - add missing prototype for tb_msgtab to modules.h 2005-08-17 21:36 knight * contrib/m_services.c: - Fixed SVSNICK; Fudged up the message tab and put mo_svsnick in the wrong place for servers :X. Tested with HybServ STABLE and CURRENT and Anope STABLE and CURRENT. 2005-08-17 21:08 db * src/s_bsd.c: - Yes, I forgot *this* time. I even compile tested, I even *noticed* on another compile run when I was debugging the exit_client and removed it then. ok ok IT'S DEAD IT'S GONE IT'S GONE TO VARIABLE HEAVEN IT'S DECEASED 2005-08-17 20:52 db * include/client.h, include/ircd.h, modules/m_pass.c, src/ircd.c, src/s_bsd.c, src/s_user.c: - add a connect_id variable and set it in each client in accept_connetion() this gives us an unique ID for every client that connects. The unsigned int will eventually wrap but still unique enough for us. This id will be useful for such uses as iauthd. 2005-08-17 19:15 adx * src/client.c: - fixed core in recurse_send_quits 2005-08-17 19:08 metalrock * contrib/m_services.c: - Wrap long line 2005-08-17 18:36 metalrock * contrib/m_flags.c, contrib/m_services.c, contrib/m_spoof.c, include/numeric.h, messages/ircd-standard.example.lang, src/messages.tab, src/s_user.c: - Numerified services down notice. - parv[0] -> source_p->name - Updated the need_id = yes to missing flag system. 2005-08-17 12:02 michael * contrib/m_flags.c, contrib/m_killhost.c, contrib/m_operspy.c, contrib/m_services.c, include/common.h, include/ircd_defs.h, modules/m_accept.c, modules/m_encap.c, modules/m_ison.c, modules/m_map.c, modules/m_names.c, modules/m_restart.c, modules/m_rxline.c, modules/m_userhost.c, modules/m_whois.c, modules/core/m_kill.c, modules/core/m_message.c, modules/core/m_mode.c, modules/core/m_sjoin.c, src/channel.c, src/channel_mode.c, src/csvlib.c, src/ircd_parser.y, src/numeric.c, src/s_conf.c, src/s_log.c, src/s_serv.c, src/s_user.c, src/send.c, src/y.tab.c: - Removed BUFSIZE define. Everything should be using IRCD_BUFSIZE - Fixed memory leak in m_killhost() - Style corrections here and there - m_services.c: /me shoots knight 2005-08-17 11:27 knight * contrib/README: - Removed stale m_tburst.c entry - Added IP cloaking and Services modules entries with descriptions 2005-08-17 10:56 knight * contrib/m_services.c, include/modules.h, src/modules.c: - Removed MAXLINE - Added BotServ and HostServ Support for Anope Services 2005-08-17 10:38 knight * contrib/m_services.c, include/modules.h, src/modules.c: - Removed SVSMODE, deemed useless. - Added *secure* messaging (Thanks Michael) - Source code clean up. Everything Works! 2005-08-17 09:45 michael * contrib/m_services.c: - yet another deliver_services_msg() fix 2005-08-17 06:33 michael * contrib/m_services.c: - Another fix to m_services.c 2005-08-17 06:22 michael * contrib/m_services.c: - m_services.c was full of bugs and didn't even compile - Fixed. 2005-08-16 23:05 metalrock * contrib/m_services.c: - Fixed a bunch of compilation errors. - This still doesn't compile, please test compile before committing! 2005-08-16 22:59 metalrock * contrib/ip_cloaking.c: - Fix missing header. 2005-08-16 22:29 db * modules/core/m_server.c: - sid was not being used in ms_sid() at all, so removed. - name was being used in some cases, SID_NAME was being used in others name was consistently replaced with SID_NAME in all cases. 2005-08-16 22:22 db * modules/m_pass.c: - remove the id given to a connecting client, if it happens to be a server with its own SID, then replace the id with the given SID 2005-08-16 21:25 db * include/s_auth.h, src/s_auth.c: - Fixed a critical region bug Basically, start_auth() is not guaranteed to be atomic. This means that an auth response can occur before SetDNSPending happens, thus incorrectly trying to free the auth twice. 2005-08-16 21:13 knight * contrib/m_services.c, include/modules.h, src/modules.c: - Added SVSMODE support; Needs testing and possibly to support channel modes. - TODO: Will add SVSTOPIC support for setting topics through SVSTOPIC with services that support the command. - Moved macro generation again to a proper place, still needs testing. 2005-08-16 20:47 knight * contrib/: ip_cloaking.c, m_services.c: - Some CRC32 hash changes made to IP cloaking to avoid making the hostnames to long. - Moved the SERV_FUNC marcos to the end of the file to avoid 'funny' generations. 2005-08-16 16:59 knight * contrib/m_services.c: - Fixed minor typo; IsServ to IsServer 2005-08-16 16:42 knight * contrib/Makefile.in, contrib/m_services.c, include/modules.h, src/modules.c: - Initial Commit; Added services support for hybrid. - Added SVSNICK and it's been tested and works - Added services aliases using a port of hybserv's contrib module - Everything needs testing and possible additions for other daemons! - Added appropriate changes into modules.h and modules.c 2005-08-16 14:34 knight * modules/m_rehash.c: - Minor #ifndef clean up, caused GCC 4 to have a severe coniption fit by claiming there were orphaned brackets. 2005-08-16 11:14 knight * include/hook.h, src/hook.c: - Fixed nasty segmental fault from MODRESTART caused by cb->name 2005-08-16 10:14 knight * contrib/README.cloaking: - Another modules typo *sigh* *grrr* 2005-08-16 10:12 knight * contrib/README.cloaking: - Minor syntax typo concerning modules {}. 2005-08-16 10:06 knight * contrib/README.cloaking: - Revised README.cloaking to appropriately adjust for the fact IP cloaking is now a module instead of a patch. 2005-08-16 06:11 adx * src/hook.c: - updated copyright notice. spotted by Michael. 2005-08-16 06:03 adx * contrib/: 4000_ip-cloaking.patch, Makefile.in, README.cloaking, ip_cloaking.c: - modularised ip cloaking patch 2005-08-16 05:27 adx * contrib/: m_ctrace.c, m_etrace.c, m_ltrace.c, spy_admin_notice.c, spy_info_notice.c, spy_links_notice.c, spy_motd_notice.c, spy_stats_notice.c, spy_stats_p_notice.c, spy_trace_notice.c, spy_whois_notice.c: - fixed contrib to compile with the new hooking system 2005-08-16 04:01 adx * include/s_user.h, src/s_user.c: - added hooks to allow special handling of certain usermodes 2005-08-16 03:33 adx * modules/m_topic.c: - allow halfops change channel topic 2005-08-15 21:02 lusky * RELNOTES: test commit 2005-08-15 20:42 knight * Hybrid-team, include/modules.h, src/modules.c: - Added missing contrib modules to modules.h and modules.c - Removed m_tburst entry as that contrib module was moved into main src as m_tb.c - Moved my own entry in Hybrid-team to be in alphabetical order 2005-08-15 17:20 db * src/: client.c, s_serv.c: - send_tb is not called from a hook at all anymore. fix comments, remove now uneeded hook.h include. 2005-08-15 16:50 adx * include/hook.h, include/list.h, include/s_auth.h, include/s_conf.h, modules/m_admin.c, modules/m_info.c, modules/m_links.c, modules/m_motd.c, modules/m_stats.c, modules/m_trace.c, modules/m_whois.c, src/hook.c, src/ircd.c, src/s_auth.c, src/s_bsd.c, src/s_conf.c, src/s_user.c: - imported callback/hook code from AISP in place of inefficient and not very flexible earlier one 2005-08-15 15:41 knight * contrib/README.cloaking: - Complete re-write of README.cloaking to be more specific and detailed since IP cloaking is rarely used outside of UnrealIRCd and ircu2. This should hopefully answer everyone's questions. 2005-08-15 14:08 adx * include/client.h, src/s_user.c: - removed ALL_UMODES macro - we are ready for umode modules. 2005-08-15 14:02 knight * contrib/README: - Removed the entry for halfops.diff. This is a stale entry as we no longer have a diff for that. 2005-08-15 13:54 adx * include/s_user.h, modules/core/m_nick.c, src/s_user.c: - remove user_modes array - rename user_modes_from_c_bitmask -> user_modes (in preparation for modular usermodes) 2005-08-15 13:40 knight * contrib/README.cloaking: - Added instructions and recommendations to alter the KEY entries in cloak.c. 2005-08-15 13:16 knight * Hybrid-team: Submitted by: Reviewed by: Approved by: Obtained from: Added myself as a contributor *only* to the Hybrid-Team list 2005-08-15 13:07 knight * contrib/: 4000_ip-cloaking.patch, README, README.cloaking: Added IP cloaking ability to ircd-hybrid as a contributive patch. Might still need some testing but has proved relatively stable. Enjoy! 2005-08-15 07:45 michael * modules/m_stats.c, src/client.c, src/s_log.c, src/s_stats.c: - Fixed compiletime warnings on 64bit architectures where 'uint64_t' is defined with an 'unsigned long' instead of an 'unsigned long long'. And no, there is no 64bit architecture where a 'long long' is 128 bit wide, even if some people might think so. 2005-08-15 04:00 metalrock * contrib/vprintf_fix.patch: - Updated vprintf_fix.patch 2005-08-15 03:50 metalrock * include/s_user.h, src/s_user.c: - Removed shadowed declaration 2005-08-14 03:26 michael * etc/example.conf, etc/example.efnet.conf, include/resv.h, modules/m_resv.c, modules/core/m_join.c, src/channel.c, src/ircd_parser.y, src/match.c, src/resv.c, src/y.tab.c, src/y.tab.h: - A colon ':' needs to be treated as a wildcard character in order to get valid_wild_card() working with ipv6 based klines. - Use find_exact_name_conf() instead of find_matching_name_conf() for removing RESVs. - Fixed server matching bug in ms_unresv() - Got rid of clean_resv_nick(). Actually we allow people to set invalid RESVs. e.g. beginning with a '-' or [0-9] etc. Such RESVs won't take effect on a real nick anyway. Wildcard sanity check is now done by valid_wild_card_simple() - Updated example configuration files - # is now a wildcard character, so disallow non-admins to add a RESV include a '#'. 2005-08-13 11:38 db * include/s_conf.h, modules/m_kline.c, modules/m_resv.c, modules/m_rxline.c, modules/m_xline.c, src/s_conf.c: - A compromise for now with Michaels suggestion Add a parse_flag to parse_aline, there is only one parse option available at the moment. Check for wild cards or not. Unkline type functions should not test for wild cards. 2005-08-13 05:29 michael * etc/: example.conf, example.efnet.conf: - Updated example configuration files 2005-08-13 01:15 db * include/hook.h, include/s_user.h, src/hook.c, src/s_bsd.c, src/s_user.c: - comment the hook.c functions. - make sure all the hook events are initialised in init_hooks() - add new hook event start_auth - after uid is generated, start_auth event happens, if a local auth hook is in place, the ircd can auth through that hook instead of using the default ircd start_auth code, or use both. (TBD a bit more actually) - a little untested but works without a hook on start_auth for now 2005-08-12 21:50 db * src/: s_auth.c, s_bsd.c, s_conf.c, s_user.c: - generate an UID in add_connection() early enough to be used for start_auth i.e. for asynchronous authentication. 2005-08-12 15:53 michael * include/irc_string.h, src/channel_mode.c, src/irc_string.c, src/s_conf.c: - Got rid of pretty_mask as suggested by diane - Removed "restricted" leftovers - Cleaned up chm_ban, chm_except and chm_invex a bit 2005-08-12 02:31 db * src/s_conf.c: - grrr changed interface forgot to change comment 2005-08-12 02:25 db * src/: s_conf.c, s_user.c: - added flags_to_ascii() seemed obvious to me. Also simplified stats U - added more comments 2005-08-11 23:07 lusky * RELNOTES: another test commit 2005-08-11 22:23 lusky * RELNOTES: Test commit 2005-08-11 09:32 db * include/s_conf.h, modules/m_kline.c, modules/m_locops.c, modules/m_resv.c, modules/m_rxline.c, modules/m_xline.c, src/ircd_parser.y, src/s_conf.c, src/y.tab.c, src/y.tab.h: - removed redundant CLUSTER_ flags as they were identical to SHARED_ flags moreover, with two sets of flags, it is easier to have a typo (as I did) with duplicate flag bits. - removed an unused find_matching_name_conf() for OPER_TYPE 2005-08-11 08:12 db * src/: ircd_parser.y, y.tab.c, y.tab.h: - Take out obsolete code not needed with split_nuh() 2005-08-10 21:58 db * src/s_conf.c: - oops, simple formatting fix 2005-08-10 21:54 db * src/s_conf.c: - commented nick NULL usage for split_nuh() 2005-08-10 21:50 db * etc/example.conf, etc/example.efnet.conf, include/s_conf.h, src/ircd_lexer.l, src/ircd_parser.y, src/y.tab.c, src/y.tab.h: - added tkline txline tresv flags for share and cluster code 2005-08-10 21:06 db * include/channel_mode.h, include/s_conf.h, src/channel_mode.c, src/ircd_parser.y, src/s_conf.c, src/y.tab.c, src/y.tab.h: - get rid of split_user_host replace with split_nuh completely - modify split_nuh() to put hostname into *host if nick pointer is NULL originally split_nuh would parse that as a nick. 2005-08-10 18:39 db * Hybrid-team, src/channel_mode.c: - move older non active members down the list in Hybrid-team - make split_nuh do the string allocations, this greatly simplifies add_id and del_id, i.e. no more strcpy in the worst case fewer DupString() in the best case 2005-08-10 13:16 db * include/modules.h: - commited patch fix from Alan LeVee alan.levee@prometheus-designs.net (knight-) 2005-08-10 11:48 adx * src/s_bsd.c: - fixed client ssl 2005-08-10 11:25 db * src/s_conf.c: - fixed comments which were no longer agreeing with function/prototype (order change) 2005-08-10 11:04 michael * src/s_conf.c: - Fixed bug in parse_aline() where opers were able to set a remoteban with invalid parameters, i.e. a NUL servername "KLINE *@someho.st ON :" - Another cosmetic fix to parse_aline() 2005-08-10 00:34 db * src/s_conf.c: - use upper case letters for perm, lower case for temp 2005-08-09 23:57 db * include/resv.h, modules/m_resv.c, modules/m_xline.c, src/resv.c, src/s_conf.c: - added temporary resvs - tested interoperability 2005-08-09 15:45 db * include/s_conf.h, src/ircd_parser.y, src/s_conf.c, src/y.tab.c, src/y.tab.h: - cluster conf items do not need to include a MatchItem no use was ever made of the user/host portions 2005-08-09 11:37 michael * etc/example.conf, etc/example.efnet.conf, include/s_conf.h, modules/m_kline.c, modules/m_locops.c, modules/m_resv.c, modules/m_rxline.c, modules/m_xline.c, src/ircd_parser.y, src/s_conf.c, src/y.tab.c, src/y.tab.h: - Fixed clustering once and for all. Changes need to get backported to 7.1.2 since clustering is broken there, too. 2005-08-09 07:26 db * src/s_conf.c: - fixed bogus comments on cluster_a_line() - renamed variable to make it a bit clearer action_type -> cluster_type 2005-08-09 07:04 db * src/s_conf.c: - comment the NULL reason pointer usage in parse_a_line() - move reason checking after user and host checking 2005-08-09 06:58 db * modules/m_kline.c, modules/m_resv.c, modules/m_xline.c, src/s_conf.c: - parse_a_line does not need a reason pointer if it's for an unkline etc. So allow a NULL reason pointer and ignore it if present instead of coring. - Allow an ON * (or anything that matches local server) to apply line locally i.e. kline something on * will kline locally as well. i.e. this is putting back in the match() (noted by Michael) - added more comments! 2005-08-09 06:21 db * include/s_conf.h, modules/m_kline.c, modules/m_resv.c, modules/m_rxline.c, modules/m_xline.c, src/s_conf.c: - changed parameter order to parse_aline() put parc,parv before user,host 2005-08-09 06:02 db * modules/m_kline.c, modules/m_locops.c, modules/m_resv.c, modules/m_rxline.c, modules/m_xline.c, src/s_conf.c: - more cleanups to the cluster code - use parse_a_line() when possible - modify parse_a_line so it can take a NULL pointer to a temp line parm - rxline cannot cluster, #if 0 it for now - locops, kline, resv, xline, rxline all tested. 2005-08-09 00:29 db * include/cluster.h, include/s_conf.h, modules/m_kline.c, modules/m_locops.c, modules/m_resv.c, modules/m_rxline.c, modules/m_xline.c, src/Makefile.in, src/cluster.c, src/ircd_parser.y, src/s_conf.c, src/y.tab.c, src/y.tab.h: - removed cluster.c and cluster.h, moved cluster_a_line() into s_conf.c for now. It might not belong there either, but it's just one function. 2005-08-08 23:55 db * modules/m_kline.c, modules/m_locops.c, modules/m_resv.c, modules/m_rxline.c, modules/m_xline.c, src/cluster.c: - I had reversed CAP_ and CLUSTER_ flags in most places. - Fixed argument number for m_xline for both ENCAP and non ENCAP cases 2005-08-08 17:28 db * include/cluster.h, modules/m_kline.c, modules/m_locops.c, modules/m_resv.c, modules/m_rxline.c, modules/m_xline.c, src/cluster.c: - First pass at cleaning up the cluster mess replace all the cluster_this_and_that() with one cluster_a_line() Next pass will clean this up even more, but this is a good start. Notably, cluster_a_line() is simple enough to be just moved into s_conf.c 2005-08-07 17:26 adx * include/numeric.h, src/messages.tab: - change RPL_ISCAPTURED and RPL_ISUNCAPTURED from 707-708 to 727-728 to avoid numeric conflicts with ratbox. Confusing clients is not what we want to achieve. They should now tell us about their new numerics, and we should do so (via hybrid ml) as well. 2005-08-07 06:02 michael * doc/server-version-info, modules/m_version.c: - Updated server-version-info - Fixed small cosmetic bug in confopts() wich had an impact on /VERSION output. We always used to use a lowercase 'g' for NO_FAKE_GLINES and an uppercase 'G' for GLINES, not vice versa. 2005-08-07 04:48 michael * etc/example.conf, etc/example.efnet.conf, include/s_conf.h, modules/m_oper.c, src/ircd_parser.y, src/lex.yy.c, src/s_conf.c, src/s_user.c, src/y.tab.c, src/y.tab.h: - Added support for encrypted auth{} passwords 2005-08-06 18:15 michael * contrib/m_spoof.c, etc/example.conf, etc/example.efnet.conf: - s/have_ident/need_ident/ 2005-08-06 12:40 michael * etc/example.efnet.conf: - hide_server_ips was originally set to 'yes' in example.efnet.conf 2005-08-06 12:30 michael * etc/example.conf, etc/example.efnet.conf, src/ircd_lexer.l, src/lex.yy.c, src/y.tab.c, src/y.tab.h: - ircd_lexer.l: It's maybe a better idea to leave it an optional whitespace so we can allow .include and/or .include"file" - Copied example.conf to example.efnet.conf and made efnet specific changes, e.g. removed SSL options etc. so we're again more or less in sync with the base configuration file. 2005-08-06 08:57 db * src/s_user.c: - add a few more comments 2005-08-06 07:26 michael * etc/example.conf: - Updated example.conf 2005-08-06 07:01 michael * include/memory.h, src/balloc.c, src/ircd_lexer.l, src/ircd_parser.y, src/lex.yy.c, src/memory.c, src/y.tab.c, src/y.tab.h: - Fixed invalid whitspace input pattern - Added support for c++ comments - Added better error reporting in case we cannot open an include file 2005-08-04 20:04 metalrock * src/ircd.c: - Add comment about what RXPATH is. 2005-08-04 19:53 metalrock * modules/m_kline.c, modules/m_stats.c, modules/m_testline.c, modules/core/m_kill.c, modules/core/m_squit.c, src/csvlib.c, src/ircd_parser.y, src/y.tab.c, src/y.tab.h: - Made all the "no reasons" consistent. 2005-08-04 06:14 michael * src/s_serv.c: - Fixed broken topicburst, i.e. added missing colon 2005-08-04 01:50 metalrock * src/: ircd_parser.y, y.tab.c, y.tab.h: - Added missing remoteban = yes in ircd.conf. - Spotted by zerocide. 2005-08-02 18:30 metalrock * src/s_user.c: - Make the ircd compile again. 2005-08-02 18:18 adx * src/channel_mode.c: - fixed b0rked halfop support 2005-08-02 13:06 db * doc/technical/send.txt: - corrected send_message and added send_message_remote 2005-08-02 10:42 michael * include/s_user.h, modules/m_set.c, src/s_conf.c, src/s_user.c: - Fixed bug where we wouldn't remove INVEX and/or EXCEPTS from isupport after it's been disabled in the config 2005-08-02 10:19 michael * src/s_user.c: - Misc. cleanups and fixes to isupport 2005-08-02 02:26 db * doc/technical/send.txt: - corrected doc a bit more, removed obsolete functions 2005-08-02 02:20 adx * src/: ircd.c, s_log.c: - don't reuse fd 2, always leave it as NULL when running in background 2005-08-02 02:12 adx * include/fdlist.h, include/s_bsd.h, src/fdlist.c, src/s_bsd.c: - move close_standard_fds() implementation to fdlist.c 2005-08-02 02:02 adx * doc/technical/send.txt: - added info about my sendto_* changes 2005-08-01 19:01 adx * messages/: Makefile.in, custom.lang: - added customized numerics 2005-08-01 18:46 db * src/send.c: - serial number checking is not necessary here either 2005-08-01 18:10 db * src/s_log.c: - if reopen on ircd.log fails, open it on /dev/null to try and keep fd 2 stderr from being grabbed for something. This is a corner case I am not too worried about. 2005-08-01 18:06 db * src/send.c: - serial check is a waste of time 2005-08-01 10:18 db * src/: fileio.c, s_log.c: - Trying to close NULL fb's is a bad idea, be a little anal about it 2005-07-31 14:01 db * modules/Makefile.in: - change Makefile to reflect repo move of m_tburst.c to m_tb.c 2005-07-31 10:39 db * help/opers/: kline, rxline, xline: - changed