Path: utzoo!mnetor!uunet!husc6!cmcl2!rutgers!sdcsvax!ucbvax!decvax!virgin!zinn!mem From: mem@zinn.UUCP (Mark E. Mallett) Newsgroups: comp.os.vms Subject: Re: Store & forward for VMS mail Message-ID: <204@zinn.UUCP> Date: 1 Jan 88 04:32:38 GMT References: <8712250028.AA23885@jade.berkeley.edu> Organization: Zinn Computer Co., Litchfield NH Lines: 45 In article <8712250028.AA23885@jade.berkeley.edu>, WARNER..NAGY@FNALB.BITNET (Frank.J.Nagy@jade.Berkeley.EDU, writes: > Leo Geoffrion writes: > > > Does anyone have recommendations for an inexpensive store & forward extension > > for VMS Mail? Since we have a homogeneous VAX/micros environment, we are > > not really interested in the elaborate packages that do conversion to EMAIL > > PROFS, or other foreign protocols. > I didn't see the original of this-- One possibility, if you already have a lot of VAXes and happen to have one or more that runs Ultrix, is to send all the mail through the ultrix systems which DO have store-and-forward mail. Some years ago I wrote a program that ran as a VMS DECnet mail object and handled SMTP mail traffic between VMS and TOPS20 machines. The program queued up incoming mail; a background job was run to dequeue it and deliver it by talking to the appropriate mail object (SMTP if the destination was a DEC20, VMSMail if the destination was a VAX; common destination sites were of course optimized, failures were requeued and returned to sender after some parameterized length of time). Mail was sent the other way by using the old (new then) trick of sending to a dedicated account and putting the real destination in a comment, viz: to: MAILBOX!user@DEC20 (I also provided other ways of specifying information in a pseudo-header within the VMS mail body). A background process was responsible for taking mail out of the MAILBOX mailbox and putting it in the above-mentioned mail queue. A fallout of this is that the process magically provided VMS-to-VMS store-and-forward capability. Unfortunately the MAILBOX!user addressing scheme made the REPLY command not work, but this was acceptable. This was a fairly quick hack but it worked very well. For simply doing VMS store and forward, I would consider writing a new VMS mail object that would handle the VMS mail negotiations through a store-and-forward queue, and either attempt local delivery itself or talk to the "real" VMS mail program assigned to another DECnet object number and let IT deliver the mail. -- Mark E. Mallett Home phone: 603-424-8129 uucp: mem@zinn.UUCP (...decvax!elrond!zinn!mem or ...sii!zinn!mem) BIX: mmallett