This directory contains some programs having to do with VMS MAIL which are new since the last SIG tape. They work on VMS V5.3 even though the interface isn't officially supported until V5.4 (where they also work). SENT_MAIL_MESSAGES.MAR is a program which lets one user see what has happened to messages which were sent to one other specific user. ALL_MESSAGES.MAR is a program which lists information about all of the mail messages for a specific user. >>> READ THE DISCLAIMER IN THE SOURCE CODE BEFORE USING <<< The remainder of the programs use the old network protocol. Normally this isn't adviseable anymore, but I was faced with a situation where mail was comming in from a particular source with text which very long lines and no carriage control and didn't print or edit properly. The purpose of the programs is to look like the normal MAIL network receiver, intercept the incomming mail, do special processing for one source, and pass everything on to the 'real' mail. This is done by defining the MAIL object to go to WRAP_MAIL.COM (as would be done with DEC's MAIL.COM) rather than the object, and this runs the WRAP_MAIL program. It's value to others, if any, is if you have to do this kind of thing. If you are originating your own mail use the new call interface. NONTRANSP.C is an expansion of the above program. I had to try to find out how the protocol works with CC: lines, which isn't really documented anywhere. To do that, I wrote this program to send mail: the idea being that once I figured out how to send mail I would know enough about the protocol to write a program which receives mail. These programs normally read from SYS$NET, which is a logical name that establishes the network connection. In order to test the programs, I made sure the logical name is NOT defined, and that file SYS$NET.; exists: that way the program just reads from the file. It currenly is set up to include CC: lines, otherwise to test WRAP_MAIL leave out the CC: line. Also be aware that there are two lines which look blank but are actually one byte long containing a NULL. This is necessary for the protocol to work. Bart Z. Lederman System Resources Corp. Burlington, MA