HP 3000 Manuals

HP DeskManager Customization

Message Structure within FSC Files 

Within a particular E file (or its incoming message equivalent) all
messages are structured according to the format for which the related
gateway is configured.

There are three message formats that are used.  Each format contains a
message header, followed by some message content.  There is a double
Carriage Return/Linefeed (CRLFCRLF) between each message header and the
corresponding message content because each header ends with CRLF and each
message content starts with CRLF.

If secured files have been requested, the file code of the output file is
the negative of the normal file code.  For example, if the output file is
a text item it will normally have a file code of 1167.  If it is a
secured file, its file code will be -1167.  (The one exception to this
rule is if the normal filecode is 0, the secured file code is -1172.)

The three different formats for outgoing messages are described below:

ARPA Standard Format 

When configured as a Type 2 gateway on the EFT/FSC configuration screen,
messages going out from HP Desk will be in the following sequence within
the E file:

     ...FF <ARPA header> CRLFCRLF <HP Desk message contents>

FF is an 8-bit form-feed character.  It is the character used to delimit
messages, and is required before the first message in the file.  Note
that it is not the same as the ASCII FF character.  In the above case the
Distribution List is actually part of the message contents.  The E file
has a file code of 1168 in this case, which will be seen as having a file
type of MARPA. The effect of Type 2 is to put the entire contents of the
screen, including the Distribution List, into the message.

The ARPA header (which was derived from the original Distribution List)
only contains those addresses which exist on the network accessed through
the specified gateway.  The list incorporated in the message content is
complete and unchanged from the original.

<HP Desk message contents> exists as sequential text lines of ASCII
characters in outgoing messages.  Each text line is separated from the
next by a CRLF sequence and is a maximum of 80 bytes long.  Parts of
messages that cannot be transported as ASCII text are replaced by a
message explaining that such files cannot be transported in this way.

For incoming messages, whatever appears in the <ARPA header> is converted
to a Distribution List (Part 1) and the message contents are put in Part
3.  Part 2 is the MESSAGE HEADER part which is a direct copy of the <ARPA
header>.  If the text lines supplied in the message contents are longer
than 80 bytes, they are broken down into 80 byte lines.

The main advantage of this format is that ARPA Standard is a widely
accepted format and is readily interchangeable.  Thus a nodefile in this
format can be transferred to another mail system quite easily assuming
that an application can be written to convert from ARPA format to the
other mail system's internal format.  HP Desk can accept input in this
format.  Therefore if a message comes form another system, then HP Desk
could be used to send a message around a network of HP 3000s.

ARPA Compressed Format 

Used when the gateway is configured as Type 4 on the EFT/FSC
configuration screen.  The sequence of a message within an outgoing E
file is the same as for ARPA format except that the original, unchanged
Distribution List is eliminated, and the embedded part numbers of a
normal outgoing HP Desk message are also omitted.  For incoming messages
there is no difference from Type 2.  This format is used when it is
important to economize on the size of messages being transferred.

ARPA Reference Format 

Used when the gateway is configured as Type 3 on the EFT/FSC
configuration screen.  The structures used for this format differ
significantly from those of the other two formats.  For outgoing messages
it has the following structure:

     ...FF <ARPA header> CRLFCRLF <HP Desk message contents>

<ARPA header> is in the same form and is derived in the same way as that
in the other two formats.  FF is an 8-bit form-feed character.  It is the
character used to delimit messages, and is required before the first
message in the file.  Note that it is not the same as the ASCII FF
character.

With this format, instead of one E file containing several messages for a
particular mailnode, each message in the E file contains the usual header
information.  However, the <HP Desk message contents> comprises a
sequence of records, each referring to a C file (the file that is
referenced that actually contains the data) which is to be transported to
the addressees in the ARPA header.  The format is:

     <filename 1> CRLF
     <filename 2> CRLF
     .
     <filename n>

When HP Desk creates the file, these names are in the format Cddsssss
where:

C                     is the standard prefix
dd                    is the Julian day of the year modulo 100
sssss                 is the number of seconds since midnight

However, if the filenames are to be created by another application they
can have any format that is required.

In each message the first file contains the complete and unchanged
original HP Desk Distribution List with each part translated into its
corresponding ARPA format field.  The file code of this file is 1169, or
if the message is to be privileged -1169.  Each subsequent C file
referred to may be an MPE file of any type, including program, HP Word or
graphics file.  Each file is separate, so if a message is mailed that
contains various file type, then these file types will be preserved.

An example might be the mailing of an HP Word document to a user.  The
message contains three parts; a Distribution List, a piece of text, and
the HP Word document.  When the message appears ready for transport to
the foreign system there will be four files in HPMAIL.HPOFFICE (or in the
group specified in the configurator screen).  The E file will contain the
message header and the names of three C files.  These C files will
contain a Distribution List in ARPA format with a file code of 1169 (an
MARPD file type), the text (as a plain text file with a file code of 0),
and the HP Word document.

The translation into ARPA is done by HP Desk, the external application
program only has to identify the E file from the Gateway Reference IPC
File and for each message in the E file, transport the ARPA header and
each referenced file in sequence to the foreign system.  The process is
shown in the diagram below:


            

The process for messages coming into HP Desk mirrors the process for
outgoing messages.  The detailed format to be processed by the
application program will be agreed between the two system administrators
involved.  However, the input to HP Desk must consist of an ARPA header
with addressees limited to those resident on the HP Desk network and one
or more contents files.  These files must be placed in MPE files like the
C files for outgoing messages.  The ARPA header and the names of the
files must be written to the equivalent of the outgoing E file.  Finally
the name of that file must be written to the AREFIPC file.  The process
is shown below:


            

MPE/iX 5.0 Documentation