- Script file initialization problem fixed. - Checked the clulink Y2K compliance. Everything seems to be OK. - A message can be send to the local users if a cluster logins or logouts. This message is contained in a file located into the script directory. Its name is formed by the cluster callsign (in lower case under linux) WITHOUT SSID and has the extension ".in" for logins and ".out" for logouts. Example : dxnet\script\f5mzn.in sent when f5mzn logins dxnet\script\f5mzn.out sent when f5mzn logouts Tks IS0HHA for the suggestion. - A message is send to the connected user for users login/logout. Tks IS0HHA for the suggestion. - SET/LOGIn (user command) enables users login/logout messages (this is the default mode). - SET/NOLOGin (user commands) disables users login/logout messages. - I rewrotte the code of the link initialization sequence. This should fix one of the major bug in the Pavillon protocol, wich sends a wrong hop count for both PC16 and PC19 protocols. DxNet will search the correct hop count in its own database (you can show it with SH/HOPS). This code is not perfect, and will never be until AK1A fixes this bug in his program himself. The problem is to know if the hop count we receive for either PC16 or PC19 is valid or not. It is possible to have a good idea with the help of the PC50 hop count anyway, assuming this protocol starts with a value of 99. If we get a PC16/19 hop count lower than the PC50 one, it is *perhaps* okay. If it is upper, it is *probably* wrong. If we don't receive any PC50, or if PC50 hop count does not start with a value of 99, we can not know exactly. This feature can be disabled starting DxNet with a '-noinit' switch. * IMPORTANT * : Add the new directory tmp (in lower case for the linux version) into the dxnet one. For example : * DOS mkdir c:\dxnet\tmp * LINUX md /usr/local/dxnet/tmp You will get an error message if this directory is not present. - Null hop count value (H0) was send during the link initialization sequence for PC16 (user login) protocol. Fixed. Tks SM7GVF. - Added the buffer left in SH/TNC_STATUS when the BPQ switch is running. - OH6NJ noticed that the main problem wich causes the system crash when using the g8bpq switch is the bpq buffers. When buffers are low and buffer count gets 0 or below, bpq and the whole system crashes. Added a dxnet switch "-bpql##" where ## is the minimum number of free buffers allowed. If the buffers left is lower than this value, dxnet will exit with the error code 9. This is checked by DxNet every minutes (at second 0). Example : dxnet -bpql10 then DxNet will exit if free buffers is lower than 10. - Added SET/Date yyyymmdd function. Sysop function. - Added SET/Time hhmmss function (ss is optionnal). Sysop function. - Added SHow/Date function. - Added SHow/Time function. - Added a macro command feature. Macro commands can be used to define or redefine your own command set. Macro files are placed in either extcmd or syscmd directory and have the .mac extension. Assuming you want to create a new command SH/ALL wich shows the last DX, WWV and ANNOUNCE spots : ---- macro file c:\dxnet\extcmd\sh\all.mac ---- SH/DX/%1 SH/WWV/%1 SH/A/%1 ----------------------------------------------- %1 is an optionnal variable. If the user enter, say, SH/ALL 1, he will get only one spot. You can use in macro the whole variable set of the message files. There is no recursive problem using macro. The following example, even if it has no interest, works perfectly : ---- macro file c:\dxnet\extcmd\l.mac ---- l ------------------------------------------ - dxnet\system\bin\dirmes.bin file format has been changed. You MUST DELETE the old file wich is NOT compatible with the new one. - VE9PT reported that pavillon's cluster use larger recipient field than dxnet in messages. Recipient field size has been increased up to 8 characters.