Release notes for MindTerm 3.1 * Can now print screen to printer There are two new entries in the "File" menu which lets the user print the current screen, or the entire scrollback buffer, to a printer. * Added virtual keyboard It is now possible to define a virtual keyboard in the configuration file. This will be shown as a bunch of buttons at the bottom of the MindTerm window. This is not intended to be able to replace the entire keyboard but rather as a place to define special and shortcut keys. The virtual keys can generate any number of characters each. * Added the ability to remap keys It is now also possible to remap any of the keyboard keys to produce any string. This has to be configured in teh configuration file. * Added 'resizable' terminal window option This option, which defaults to true, controls if the user can resize the terminal window. * Added support for host-based authentication MindTerm can now use host-base public key authentication. * Can now disable keep-alive It is now possible to disable the sending of keep-alive packets via the settings dialog. * Fixed VT220 & VT320 key mappings The function key mappings for VT220 and VT320 were wrong. * Added subsystem example Added an example showing how to implement teh client part of a new subsystem. * Fixed a number of lockup-bugs MindTerm 3.1 fixes a number of bugs where the entire client could lock up. It also introduces timeouts during the connection and version handshake phases. * Improved detection of buggy SSH-servers The code which detects various buggy SSH-servers has been improved. * Terminal window problems fixed The terminal windows did not properly release all resources when they were closed. * Terminal window now ignores null-bytes Added an option 'ignore-null', default true, which causes the terminal window to ignore received null-bytes. * Telnet protocol fixes The telnet plugin had a number of problems with end-of-line representations. This made it more or less impossible to talk to a non-telnet server (like SMTP). Also added and 'crlf' option which allows users to force the end-of-line behavior. * SCP problems fixed There were problems with SCP when talking with sime Cisco devices with "interesting" behavior. * Less agressive focus behavior Do not request focus every time we are deiconified. This was changed since some virtual window managers generated an deiconify event whenever teh user switched to the workspace containing the window. * Fixed handling of --p option A bug made the "save" and "load" arguments of the --p option meaningless. They both acted just like "both".