Starting with tag: [TAG 2.14.0 quozl@us.netrek.org**20080226232316] [add metaserver delist tool quozl@us.netrek.org**20080301112911] [fix release date quozl@us.netrek.org**20080303214108] [deprecate name.c quozl@us.netrek.org**20080308102816 name.c has no further use, it was once used to identify the source version for creating the tar.gz distribution, but that function is now handled by Makefile logic. ] [fix configure, relied on name.c quozl@us.netrek.org**20080310021145] [post release notes quozl@us.netrek.org**20080317221808] [doxygen, add configuration and build target quozl@us.netrek.org**20080319053953] [doxygen, add documentation quozl@us.netrek.org**20080319054020] [various spelling fixes quozl@us.netrek.org**20080320015932 Contributed by Zach Uram, reviewed on netrek-dev, signed off with changes by James Cameron. ] [newbie.c, 80 column re-indentation quozl@us.netrek.org**20080325055748] [pret.c, 80 column re-indentation quozl@us.netrek.org**20080325080717] [doResources change to pl_reset quozl@us.netrek.org**20080325111155 Move many planet functions into a planet.c source file. For a long time, the reset of the galactic has been copied from the daemon into other pieces of code, resulting in a lot of duplication. This was also the last piece of daemon code that was called from other code by direct linking. planets.h was also a problem, because it contained a global scope array pdata, the planet data before play. This made it difficult to factorise. It has also been moved into planet.c. What remains outside planet.c is daemon planet simulation, planet moves, and INL reset. ] [doxygen, daemon (1) quozl@us.netrek.org**20080325112215] [inl, report who freed a slot, it can be ambiguous quozl@us.netrek.org**20080327023638] [configure, remove X libraries quozl@us.netrek.org**20080402233354] [doxygen, struct.h incomplete documentation quozl@us.netrek.org**20080402233455] [glog, add timestamps jrd@gerdesas.com**20080402235823 Add timestamps for god log entries. ] [daemon, add t-mode duration logging jrd@gerdesas.com**20080403000051 Per requests from James and Karthik added logging for duration of t-mode sessions. Log entries are sent to the god log. ] [pret, win parade, experimental quozl@us.netrek.org**20080403001917 Adds a round win parade to pre-t mode, such that the winning player and planet are highlighted by the usual conquer parade, so as to reinforce that a round was won, to increase enjoyment of pre-t. ] [review, t-mode duration logging quozl@us.netrek.org**20080403002622] [forgotten, move common shmget usage to openmem.c quozl@us.netrek.org**20080404043838 Several supervisory robots had used a common code sequence that tries to detect daemon failure by checking to see if the shared memory segment is still registered. Moved this code to openmem.c ] [is_robot_by_host, enables testing of pret locally quozl@us.netrek.org**20080404053006 Whether to classify a slot as a robot based on the IP address works for production servers, but makes things difficult for local test servers. This sysdef value suppresses the check. Default on. ] [pret, round win parade, tested quozl@us.netrek.org**20080404102309 Tested version of the Pre-T round win parade, which uses the old style conquer parade of a ring of plasma from the planet to the parade ring. Changed the conventional conquer parade to use a ring of plasma moving the opposite direction. ] [doxygen, add robotd John R. Dennison **20080409232044 Added the robotd directory to the INPUT declaration for doxygen. ] [pl_info, better documentation quozl@us.netrek.org**20080415053825] [pret, fix for save_armies_carried() John R. Dennison **20080525213059 At start of pre-t carried armies were returned to homeworld whether the slot was alive or dead; this simple patch skips non-alive slots. ] [REPOSITORIES, update of darcs link John R. Dennison **20080607040853 I have moved my darcs repo to an external host; this patch updates the link for that repo. ] [include call graph quozl@us.netrek.org**20080621024624] [add py-struct to UDP packets for pygame client quozl@us.netrek.org**20080621024650 (not a change to protocol, only a change to documentation in code) ] [note bug in clue game re captaincy on draft quozl@us.netrek.org**20080621024744] [pret, code formatting quozl@us.netrek.org**20080621024944] [SP_BADVERSION, avoid leak, symbolise quozl@us.netrek.org**20080621025910 Rework of the SP_BADVERSION packet which is sent to the client if the CP_SOCKET packet contains incorrect protocol version number, or when the queue has no free slots, or when a player is banned, or when a connection is dropped due to netrekd denied IP list. - add symbolic names for the reason codes, - prepare the packet buffer with zeros to avoid data leaks, - document the current state of client processing of reason codes. This patch does not change the behaviour of the server, apart from starting with a packet buffer which is clean. ] [SP_BADVERSION, renumbering quozl@us.netrek.org**20080621031316 A renumbering of the SP_BADVERSION reason codes, consistent with current Microsoft Windows and Linux clients. Not yet checked against MacTrek. ] [SP_BADVERSION, use for eject, ban, free quozl@us.netrek.org**20080621060325 Use the SP_BADVERSION response for eject, ban, and free actions, and ensure it is processed in any of the three socket wait states; login (getname), outfit (getentry), and in flight. Add a "setship N disconnect M" syntax for testing. Note: clean rebuild may be required, struct player changed. ] [multiplexor design idea 2007-10 quozl@us.netrek.org**20080621231931] [SP_BADVERSION, compile fixes quozl@us.netrek.org**20080621235120 Move BADVERSION codes to defs.h since packets.h is not included everywhere. Fix prototype in ntscmds.c. ] [SP_SHIP_CAP, add py-struct documentation quozl@us.netrek.org**20080623081453] [CP_SOCKET denial of service, security update quozl@us.netrek.org**20080626093114 Reduces the effect of a particular style of denial of service attack, and provides a control to compensate during an attack. ] [pl_pick_home, pathological case fix quozl@us.netrek.org**20080629000922] [rcd, bring in changes from netrek-client-cow quozl@us.netrek.org**20080705054426 merge the dist_type enum and distmacro definitions from the current version of the Linux client. ] [SP_GENERIC_32, fix structure definition quozl@us.netrek.org**20080708010459 Fix the definition of the packet structure, as it was flawed following a previous patch. This should have no impact on clients since the packet was being sent with the correct size already. ] [RCD bomb, show planet name regardless quozl@us.netrek.org**20080708014039] [SP_GENERIC_32, only send when it changes quozl@us.netrek.org**20080708074803 The SP_GENERIC_32 packet was being sent every update, even though there was nothing changed about it, so now we send it only when it has changed. ] [fix missing end of line quozl@us.netrek.org**20080711102321] [util, s2du quozl@us.netrek.org**20080711102424 Add utility function for converting a duration to described units. ] [packets, remove unnecessary ifdef lines and blanks quozl@us.netrek.org**20080711102535] [remove old fixme now done quozl@us.netrek.org**20080711102622] [features, begin to use arg1 quozl@us.netrek.org**20080711102823 Add code to support the optional feature arguments, and create an argument value for the SP_GENERIC_32 feature, to be used in a later more contentious patch. ] [inl, fix for 50 fps, add context-> fields quozl@us.netrek.org**20080711103448 1. fix for the INL robot on 50 frames per second servers, the timing was being done per daemon frame, without reference to the fps value, the change is to define PERSEC in terms of fps. 2. addition of context->inl_* fields so that daemon and ntserv may see INL timing values. ] [SP_GENERIC_32 version 'b' packet support quozl@us.netrek.org**20080711103810] [TAG 2008-07-11 quozl@us.netrek.org**20080711110133] [pl_reset, sanity of loop quozl@us.netrek.org**20080713093510 Discovered a daemon looping, backtrace showed it to be in pl_reset, trying to place fuel on a planet, yet all planets had fuel. Add code to exit the loop if too many iterations occur. ] [SP_GENERIC_32, version 'b', fix byte swapping quozl@us.netrek.org**20080714001014] [pret, conquer parade honors pre-t idlers John R. Dennison **20080714064353 Pre-t conquer parade will honor pre-t idling ships and not kill them. ] [fix comment on SP_GENERIC_32, add bug quozl@us.netrek.org**20080716021739] [TAG 2008-07-16 quozl@us.netrek.org**20080716021816] [p_disconnect support for xtkill F quozl@us.netrek.org**20080723111256] [trimscores, add trim by zero stats billbalcerski@gmail.com**20080524125843 This patch adds a new option to trimscores (harsh value 100) that trims players with rank, tkills, tlosses, tarmsbomb, and tplanets equal to zero, with tticks equal to 1. In other words, players who made their character then never played. Useful for trimming player logins generated by DoS attacks. This patch contains the following changes: M ./Vanilla/tools/trimscores.c +24 ] [trivial packet doc fix quozl@us.netrek.org**20080726102330] [experimental, WHY_DEAD_2 fix quozl@us.netrek.org**20080726121527 The WHY_DEAD_2 feature was added in a way that prevented it from working, since features are not functional in the context of the daemon. This is experimental rework of the feature, untested. ] [experimental, clue game substitution quozl@us.netrek.org**20080726121619 Provide a structured way for players in a clue game to be substituted with observers, under their own control. May require client support. ] [TAG 2008-07-26 quozl@us.netrek.org**20080726123321] [RSA does not mean trusted, reported by zelchenko quozl@us.netrek.org**20080727034634] [openmem, test program quozl@us.netrek.org**20080727035758] [xtkill, add (f)latten John R. Dennison **20080802024406 Adds an option to flatten armies and kills on the given slot. ] [is_idle, not if torps in flight quozl@us.netrek.org**20080803020136 Changes the definition of safe-idle to exclude ships that have torps in flight, as it was possible to throw torps from the home world and yet be safe from attack. Adds GU_UNSAFE, which allows the client to see whether the ship is safe idle, and as the bit used to be GU_GAMEOK the meaning is inverted. ]