This Update: 12 Jun 1997 .Q00 - Q14 -- Added the Qwk code, fixed several bugs, made the subject the room name. Still bugs to fix... -- Fixed a bug in the Sysop Menu handling... Should display better now. -- Fixed several messsage bugs(some in the code, some in the message file). Posted the changes in Citadel 68K room -- Created a utility to compute the add/change/deleted messages. -- Added code to read and process a file BADPEOPLE.SYS. This file if it exists will allow a sysop to remove messages by that person totaly. This may not be the final implementation, but it does work. It allow a person to be banned from a room, or all rooms. This means that if the message is removed and your networking the room involved, it does not get sent to other systems. This means your Censoring for the net! -- With the help of Jim Dawson(author of AmiQwk), I was able to find the bug and the Qwk packet downloads works fine! The bug was in the file MESSAGES.DAT, it did not have the conference number wrong. Still some cleanup work before a release. -- Fixed the delay in the recognition of ARexx commands. No longer will it take up to a minute for stuff to be processed. Slightly more cpu overhead, but not really all that much. -- Added many new messages for the QWK processing, several new help files. -- Added an option to CONFG. -F, -f, -I, or -i is identical to FirstInit on the command line. -O, or -o is identical to OnlyParams. -R, -r or ReInit will rebuild the ctdltabl.sys file. Anything else will still be assumed to be a ReInit, but will tell you that you made a parameter error. -- Fixed a bug in CONFG that may be the long outstanding bug that people with A3000s and A4000s have reported where they must do a NOFASTMEM before they can run CONFG. There appears to be a bug in the SHORTINT option since it called a routine with a word value when the routine expected a LONG value. I finally got that to happen with the 6.55 compiler. -- Re-built everything with the 6.55 compiler. It helped to find the above CONFG bug and the solution may fix other odd problems that have been happening from time to time... -- Added parameters #TEMPAREA and #BIOAREA to define a temporary and Biography area. CONFG changed. -- Updated a few messages. -- Changed CTDL to use one temporary file area. -- removed some debug messages. -- Added internal Zmodem code. Zmodem will be disabled unless you put a parameter in the CTDLPROT.SYS file. XPR code removed, it was too buggy. This is a replacement, you can still use XPR as an external protocol. -- Updated the AmigaGuide and Text Documentation. -- increased the size of the message header buffer after Hue mentioned that an author's name could be really long(when comming from a foriegn net) formheader() in misc.c header[100] -> header[250]. also fixed several places where the author's name was used and the item it was copied into was citadel's standard 20 characters. -- removed the #TEMPAREA parameter, it was too dangerous if an error occured, it could delete the homearea files(and did in my testing). CTDL uses the existing temporary file area scheme. -- Added a parameter #BANNERAREA -- increased the delay in the hangup() routine, it was not waiting quite long enough(or so it seemed). -- fixed the check in CONFG for -O,o,F,f,I,i options, there was a bug that always caused these options to do a ReInit. Now, -O,o and -F, -f should work correctly. -- Minor fixes to some menu files. -- added the fix to CONFG to zap the last netted message if you re-initializing the message file. Otherwise, you will not send anything you net until the message index gets past the old value... -- change the network session message to come out even if netloging is not turned on... -- fixed a minor bug in CONFG, a directory could have gotten an extra slash on the end if the name had a colon(was a full path) in it. -- fixed a bug in the QWK code, if the protocol you have selected was made invalid(for example you have Zmodem selected, or never selected a protocol, then Zmodem was taken offline for some reason), the QWK code did not do a transfer. It will now assume Xmodem as the default and use Xmodem if nothing valid is selected. -- fixed a bug for an invalid baud code for serial port speeds greater than 32767, Citadel would report that -2713 was invalid during the initialization of the baud rate. -- Added check for error on open in a crashout to prevent an enforcer hit. -- Re-wrote the file comment routines. partly tested -- Found a Userlog trashing bug. Fixed. -- Found a message buffer trashing bug. -- Removed some terrible auditing code that created the two trashing bugs. -- Found there still is a case of an invalid baud rate. -- Added some code to display modem data/results.sys data mismatch. -- Added some code to collect Caller Id info in debug.sys, option +cid -- added some code to display modem text when collecting info +results -- Back in Q02 I discovered that the compiler was not passing a long where memcpy was expecting one. I found the bug again in another place in the code and did a global search for the same situation. It killed my system for two days until I found the cause because it was in net recovery routine and prevented the bbs from coming up. It should be fixed once and for all... -- Fixed Fast Transfers! There were several bugs in the process, they are all fixed. Cit 68K will not do it with Zmodem. There is still a filename disconnect with external protocols, nothing I can do with that at this time. -- Fixed a bug in the Readbps routine, it did not handle all possible baud rates on the command line. -- adjusted the code for readme messages. .Q15 -- Fixed the zip file access(it also messed all the file extraction code). .Q16 -- Fixed a typo in the strcmpu routine for an error message. -- Fixed small bug in Create Floor routine. Trash in floor file. -- Added code to remove unused floor entries to CONFG. -- Added code to reset floor moderator entry if invalid. -- Added code to the verify utility to support dumping of the ctdlflr.sys file with the -f flag. -- Added vortex checking to MSGADD plus fixed the %y & %m bug in archiving. .Q17 -- fixed a couple of small typos -- confg - displayed text on rooms, users, and node scans 6J12 -- fixed a bug with private room creation (per Hue, Jr). -- fixed a bug in msgadd(actually two of them) with long node names -- added a feature to CONFG to check the number of rooms you have shared versus the #SHAREDROOMS parameter. -- Added MR to the "official list of utilities build with Citadel, it lists the domain area message headers. -- Added updates for better formatting to NETLIST. -- Updated all the utilities to have 6J12 as version ID. -- Renamed all the Citadel Archives to clarify what they are. -- Updated the headers on CTDL and CONFG to include 6J12 6J24 -- re-wrote the vortex routines to keep more data. It now creates a sub-directory off the #NEWAREA called vortex where it stores all its information. The contents will be files called "NODEID.ROOM". There could be a fair amount of files in this directory, I may do a further sub-directory and named by the Node Id and just have files in it with the name of the room. The vortex routines will have up to 20 message entries instead of the last one. -- restructured how the utilitites are built, eliminated some duplicate code so that there is actually less source in the over all package, but nothing changes size. -- removed a bogus message about problems with mail. 6K05 -- modified the vortex routines to keep more data in each file and have less files. Aproximately same space requirements, just have one file per room. This way the files should open quicker and not take any noticable longer time to read and write. -- removed the setting of MBORIG field with host node id in MSGADD and MSGOUT. -- fixed incorrect messages in the vortex routine. 6L01 -- adjusted the routine to get baud rate from modem, GetModem(), it was not returning the lockport rate if there were any errors. If the rate is locked and any problems occur, the locked rate is returned. This will speed up connections for networking. -- found that the Vortex checking with msgadd could mark a message as received when the room was not shared formally. This could create false vortex reports later when the actual message is received the correct way. Also finished the implementation of the force flag so you can "force" a message to be added even if the room is not shared with the other system. 7A01 -- Fixed some net debug messages so they did not come out if net logging was not turned on. -- Fixed a nasty bug in msgadd's version of PutMessage(), it overwrote the stack on certain longer values of the mbaddr field. -- Minor adjustments to baud rate detection code. 7C07 -- removed some dead assignments spotted while generating a 68030/optimized version of CTDL and CONFG. -- Fixed a baud rate mismatch that caused a 16 bit int to overflow for baudrates 38.4K and higher. There was a long to int to long movement of data, now it is all longs. -- Rewrote some code for password checks, it was pretty poor and I ran across it while looking at the dead assignments. The current executable is a little smaller. -- Fixed several compiler warnings about missing prototypes and data type mismatches where it did not matter. Fixed them anyway. 7D07 -- Fixed a bug that would allow screwy dial strings to be generated if the sysbaud was greater than 6 and the baudrate of the node was greater than 6. -- moved some data definitions into a single file. 7F12 -- fixed a bug in the vortex checker that made rooms with similar names ( only differing by a special character or whitespace ) be tracked to the same vortex file. -- fixed a bug where vortex problems were not reported in the Aide/Net Aide room. -- Added preliminary support for routes.sys operations.