Open Hub
Open Hub

 Settings |  Report Duplicate

0
I Use This!
Activity Not Available

Commits : Listings

Analyzed 8 months ago. based on code collected over 1 year ago.
Oct 07, 2022 — Oct 07, 2023
Commit Message Contributor Files Modified Lines Added Lines Removed Code Location Date
- Fixed critical bugs with new 'doubles' support.
akari
More... over 22 years ago
- Removed all set/remove restrictions for the LIGHT flag. - Made it so that connections of type INBOUND are not logged/logwalled when disconnected. Since this are handled in MUF, it is assumed that the MUF code will handle any logging needed. - Changed all FLOAT precision to DOUBLE instead. [FB6 change] - Made it so that the 'Program not compilable' message reports the dbref. - Made it so that the notify_descriptor* prims don't abort when trying to notify to an invalid descriptor. They simply don't do anything in such a case. - Added MUF EVENT pid info to GETPIDINFO prim. [FB6 change] - Changed all FLOAT precision to DOUBLE instead. [FB6 change]
akari
More... over 22 years ago
- Attempted a fix to the NBSOCKRECV bool being more accurate on whether the connection was killed or not. Might work better now, will test it for awhile. - Made it so that socket events send an ID of "SOCKET.READ.#" where # is the socket number. Now that I know that event_waitfor uses smatch routines to match event IDs.
akari
More... over 22 years ago
- Made it so that things set VEHICLE will be treated just the same as a player when moving them via MOVETO. Things set ZOMBIE already were handled this way. This basically affects triggering _depart and _arrive propqueues.
akari
More... over 22 years ago
- Changing to 1.80 beta 3.
akari
More... over 22 years ago
- Fixed the 'occasional leading space' bug with descriptor user names. - Fixed a bug introduced in beta2 with all of the *NOTIFY_EXLUDE prims.
akari
More... over 22 years ago
Fixed a small bug in COPYPROPS involving propdirs with values.
foxsteve
More... over 22 years ago
Added COPYPROPS primitive.
foxsteve
More... over 22 years ago
- Added the following in-server $defs to indicate various MUF related compile time options: HAVE_SQL - If $ifdef'd, means SQL support was compiled in. HAVE_FILE_PRIMS - If $ifdef'd, means File prims are compiled in. HAVE_SOCKET_PRIMS - If $ifdef'd, means Socket prims are compiled in. HAVE_MUF_EDIT_PRIMS - If $ifdef'd, means the MUF editing prims are compiled. - Fixed up a lot of things wrong with the GETDIR prim. Also made it so that it now returns 0 if the file exists but is not a directory. (Instead of returning an empty array like it did before.) - Made it so that array_delitem no longer aborts if the index is out of range. The array is simply returned to the stack unchanged. - Added a "OUTPUTQUEUE" field to the GETDESCRINFO array. Indicates the number of lines waiting to be sent in that descriptor's output queue. - Added a .debug/crashpid prop to the .debug/ directory. Indicates the last PID it was that actually errored out.
akari
More... over 22 years ago
- Fixed a big with GETDIR corrupting strings if there were subdirectories.
akari
More... over 22 years ago
- Fixed a bug with the new string prefixing function. [Proto only bug]
akari
More... over 22 years ago
- Made $pubdef abort rather than mess with @ or ~ props. [FB6 change] - Added $libdef function-name. Adds the definition needed to call that function to the _defs/ directory. Similiar to the $pubdef directive, except that it just takes the function name as an argument and does the rest. $pubdef is still needed for definitions that require any special handling other than the standard _defs/ call. [FB6 idea] - Fixed a bug with players not being unblocked when a MUF in event_waitfor was dequeued. - Add an invalid name check to the copyobj prim. [FB6 Change] - Added logging to the PROG_SETLINES prim. [FB6 idea] - Made it so that REMOVE_PROP doesn't abort when props end with the '/' mark. [FB6 change] - Added PARSEPROPEX. [FB6 prim] - Fixed a bug where programs made via NEWPROGRAM were considered in-use. [FB6 fix] - Fixed a potential Buffer overrun in ABORT_MPI and RETURN MPI macros. [FB6 fix] - Made it so that @restrict indicates the current status when used with no arguments. [FB6 change] - Added @chown_lock and @force_lock aliases to @chlock and @flock. [FB6 change] - Updated internal telnet negotiations. [FB6 change] - Started adding some of the DB accessing/setting macros that FB6 uses. [FB Change] - Made it so that @lock used with no argument will behave like @unlock. [FB6 change] - Added @relink as an in-server command. [FB6 change] - Added @tune max_wiz_preempt_count. If it isn't 0, then it indicates the maximum number of instructions that Wiz level PREEMPT programs can run as a way to prevent a w-bitted PREEMPT program infinite looping and locking up the MUCK. [FB6 idea] - Added REGEXP and REGSUB prims. [FB6 prims] - Fixed a memory leak from @toading players currently online. [FB6 fix]
akari
More... over 22 years ago
- Fixed QUEUE to return the correct PID instead of always being 1 shy. This may break some programs that were already correcting for the error in the MUF. In such cases, adding: $def queue /queue 1 - to the top should fix it.
akari
More... over 22 years ago
- Fixed it so that MUF socket events aren't stalled by the select() call in interface.c. - Added SOCKTODESCR prim. This prim pushes a MUF socket into the in-server descriptor list as an Inbound descriptor. It will show up on WHO ! as [Inbound]. The following should be kept in mind: The [Inbound] descriptor will never be idled off by the inserver idle routines. While the [Inbound] descriptor can be written to via the various descriptor notification routines, it cannot be read from at all. The original MUF socket must be used with NBSOCKRECV to read in anything on that connection. The descriptor will be unconditionally closed once the MUF socket it was made from no longer exists. The purpose of this prim was to allow me to write a MUF only webserver that maintained backwards compatability with already existing web-mufs. Most people will probably never find much use for this. - Fixed GETPIDS to catch programs in EVENT_WAITFOR states as well. [FB6 fix] - Fixed the listening MUF sockets to be non-blocking. - Fixed a really hard to find memory leak with scoped variables and the FORK prim. Scoped variables will now pass correctly to child processes as well. Passing the fix on to FB6. - Fixed a bug that allowed socket events to get to a frame that is already cleared. - Fixed a infinite looping bug when trying to queue more than 2 socket events. - Fixed a bad malloc() call relating to MUF events.
akari
More... over 22 years ago
- Added {parse_ansi:<type>, <string>} function that works exactly like the PARSE_ANSI MUF prim. - Removed the internal definition for staff? Too many programs use 'staff?' as a function name, and this in-server $def will override program-specific 'staff?' functions. - Added in-server $defines for use with SET_SOCKOPT: NOQUEUE(0), SIMPLEQUEUE(1), TELNETQUEUE(2), HOMEINSTANCE(5). - Added a HOMEINSTANCE option to SET_SOCKOPT. When used, it will assign the socket to that instance, making it so that that is the instance that gets notifed of SOCKET events on that socket. By default, Socket Events are only notifed to the program instance where the MUF Socket was created. Using the HOMEINSTANCE argument for SET_SOCKOPT, the socket is reassigned to that particular instance.
akari
More... over 22 years ago
- Removing debugging statements... again.
akari
More... over 22 years ago
- Added @tune compatible_muf_perms. Turning this on makes it so that newly created MUF programs have the QUELL and HARDUID and M3 flags on them, causing their MUF permissions levels to function the same way as Glow and Fuzzball. Defaults off. - Made it so that when a MUF Socket event happens on a listening socket, the Event ID is "SOCKET.LISTEN" instead of "SOCKET.READ". - Made it so that the SOCKET? prim will return 1 for normal sockets and -1 for listening sockets, and 0 for non-socket data.
akari
More... over 22 years ago
- Fixed a crasher dealing with making and removing objects.
akari
More... over 22 years ago
- Updating sysparms.txt.
akari
More... over 22 years ago
- Removed some debugging strings that slipped in there.
akari
More... over 22 years ago
- Made it so that the PLAYER_CREATE @power allows: @pcreate, @newpassword, and the ability to @name other players. - Added PLAYER_PURGE @power that grants access to @purge and @toad/@frob.
akari
More... over 22 years ago
- Just format cleaning.
akari
More... over 22 years ago
Fixed crasher in FORITER due to stack overflow
foxsteve
More... over 22 years ago
- Cleaned up the messy shovechars() function in interface.c. Considering changing IDLE flag behavior, and needed to make sure the function was clean before trying to work with it. - Fixed an initialization error with NBSOCKRECV. This change MAY result in breaking some MUF programs using this prim to determine disconnects. The 'result' value wasn't getting initialized to 0, so often times a large number was pushed onto the stack when it should've been zero. NBSOCKRECV can't determine disconnects anyway, so the practice of checking its return for lost connection is now considered a 'bug'. - Fixed some stack overflow errors in a couple of the socket prims. - Finished implementing MUF Socket events. Will be testing further.
akari
More... over 22 years ago
- Added socket_events @tune. Turns on the socket_events support. Defaults to on, but should be turned off if any problems are encountered while using socket prims. - First phase of MUF SOCKET events implemented. Working, not thoroughly tested yet.
akari
More... over 22 years ago
- Fixed SOCKDESCR prim. Was returning garbage to the stack.
akari
More... over 22 years ago
- Fixed it so that the wiz-only mode message will not show up on the MUF port connections.
akari
More... over 22 years ago
- Removed a debugging message that slipped back in. :P
akari
More... over 22 years ago
- Fixed TREAD to return the empty string and 0 to the stack when it times out like it was supposed to. Had been returning nothing. - Removed the enable_sockqueue @tune. Changed this control to a SET_SOCKOPT prim. - Fixed it so that one can never @force a wiz that's higher level than them. This was fixed way back in 1.00, but the change got lost somewhere along the way before we got to CVS. - Finished the MUF SOCKET->player stuff. Seems to work good. SOCKET_SETUSER works just like DESCR_SETUSER, except the descr is replaced by a MUF socket. Tested extensively. - Added SET_SOCKOPT prim for controling what queue mode NBSOCKRECV might use. This prim may be used in the future for other socket options, if I can think of any. - Altered the NBSOCRECV queue support some. Took out the @tune that was for toggling it on or off. Now there's two modes: 0 SET_SOCKOPT = Traditional behavior. No queueing, so telnet input is recieved as it's being typed. 1 SET_SOCKOPT = 'Dumb' queue. Simply queues all valid ASCII until it hits an End of Line, or the buffer is full. 2 SET_SOCKOPT = 'Smart' queue. Works like a normal telnet connection. Only printable ASCII is queued in the buffer, non-printable ASCII is ignored, except for delete and backspace, which will remove characters from the queue. Changing the mode from 1 or 2 to 0 will cause any data in the buffer to be lost.
akari
More... over 22 years ago
- Removed a debugging line I'd missed.
akari
More... over 22 years ago