HOW TO INSTALL VMS KERMIT-32 VAX/VMS Kermit-32 was written by Nick Bush and Bob McQueen of Stevens Institute of Technology in Hoboken, NJ. As of April 1987, development of the program is more or less frozen since Nick and Bob have both changed jobs, but there have been a couple updates since then from Jonathan Welch at the U of Massachusetts, and Burt Johnson of Diversified Computer Systems Inc. Bug reports will be accepted with no guarantee that they will be acted upon. See the file VMSMIT.BWR for a list of known bugs, limitations, or for bug reports that have not been responded to. Meanwhile, the C version of Kermit (C-Kermit) will probably evolve to become the preferred version of Kermit for VMS. VMS Kermit-32 comes in 3 forms: hex, Macro source, and Bliss source. Each can be used as the basis for installation. MAGNETIC TAPE DISTRIBUTION VMS Kermit comes on Columbia University's Kermit Tape B, which you should normally order in ANSI format for reading on VMS 9-track tape drives. Before beginning, make a special directory for VMS Kermit and read the files VMS*.* from the Kermit distribution tape into this directory. Columbia's Kermit tapes are written with blocksize 8192, which is 4 times larger than the default tape blocksize for VMS. You should mount these tapes on the VMS system with the following command: MOUNT/BLOCK=8192/DENSITY=1600 MTA0: KERMIT (or substitute some other tape drive name for MTA0:) Do NOT use the /FOREIGN switch. Once the tape is mounted, you can use normal VMS COPY commands to copy the files from the tape. For instance, if you have defined your Kermit directory to have logical name KER:, you can use the following command to copy the VMS Kermit files into this directory: $ copy mta0:vms*.* ker: You can also get Kermit on TK50 cartridges. These are written in VMS BACKUP format. FILES The VMS Kermit files are the ones whose names start with "VMS". Most sites will not want to bother with the source code, and simply want the executable program, the help files, and the documentation (see METHOD 1 below). The minimum collection of files is: VMSMIT.HEX - The hexadecimal encoding of KERMIT.EXE VMSDEH.MAR - A macro program to convert the VMSMIT.HEX file into KERMIT.EXE VMSMIT.DOC - The VMS Kermit user manual VMSMIT.PS - A version of the user manual for printing on Postscript printers VMSMIT.HLP - VMSSYS.HLP - VMSUSR.HLP - VMSMIT.BWR - "Beware File" listing known or suspected bugs & limitations. METHOD 1: DECODE THE HEX FILE WARNING -- If you are running a pre-4.0 release of VMS, see Method 1.5 below! The easiest way to install VMS Kermit is to "dehexify" the hexadecimal encoded task image. Follow these steps: 1. There should be a file VMSDEH.MAR on your disk. This is a Macro-32 program that decodes the hex file of the VMS Kermit task image. Compile and load this program: $ macro vmsdeh $ link vmsdeh 2. Run the VMSDEH program: $ run vmsdeh Please type the file name: vmsmit.hex The VMSDEH program automatically creates KERMIT.EXE. Or it might create VMSMIT.EXE, which you can rename to KERMIT.EXE. 3. Now you can type "run kermit". Make sure it works. If not, try method 2. NOTE: There are some new files available (as of version 3.3.122) to help with the installation: VMSINS.KIT, VMSMKH.COM, and VMSMKI.COM. If you can use these files (DCL procedures) to install Kermit-32 on your system, then you can skip steps 4 and 5. 4. Install KERMIT.EXE in the appropriate system area. 5. Install the help files, for instance: $ LIBRARY/HELP/DELETE=KERMIT SYS$HELP:HELPLIB.HLB $ LIBRARY/INSERT/HELP SYS$HELP:HELPLIB.HLB VMSSYS.HLP $ LIBRARY/CREATE/HELP SYS$HELP:KERMIT.HLB VMSUSR.HLP See the file VMSINS.HLP for further hints about the help files. Note, the companion program to VMSDEH.MAR is VMSHEX.MAR. It creates hex files from binary files. You can use this program to encode any VMS binary file into printable form, and VMSDEH to decode it back into its original form, with all of its directory information intact. METHOD 1.5: DEHEXIFY THE OLD KERMIT-32 HEX FILE If you are running a version of VMS before 4.0, then the current release of Kermit-32 will not work on your system. If you tried Method 1, and the resulting Kermit program gave an error message like "invalid image header" when you tried to run it, this is probably the cause. In that case, you must use version 3.1 of VMS Kermit until you upgrade your VMS version. To use version 3.1 of VMS Kermit, follow the directions for Method 1, but use VMSDEH to dehexify the file VMSV31.HEX rather than VMSMIT.HEX. The result will be called VMSMIT.EXE rather than KERMIT.EXE. You can rename this file to KERMIT.EXE and install it and the help files normally, but note that the help files apply to the newer releases. METHOD 2: ASSEMBLE THE MACRO FILES Only use this method if method 1 fails for some reason. This method assembles all the Macro-32 source programs into object files, and then links them together into KERMIT.EXE, and then installs KERMIT.EXE and the help files in the system. NOTE: Kermit-32 is NOT written in Macro-32; it is written in Bliss-32, and these Macro files are output by the Bliss compiler. It is not recommended that changes be made to the Macro source. To build from the Macro code, follow these steps: 1. Make sure you have all the VMS*.MAR files on your current disk. 2. There is a DCL procedure to assemble and link all of these files, but unfortunately it assumes the files are called KER*.* instead of VMS*.*. There is another DCL procedure, VMSREN.COM, to rename them: $ @vmsren 3. Now that the files are renamed to KER*.*, you can run the DCL procedure VMSINS.COM to assemble and link them into KERMIT.EXE. $ @vmsins Kermit-32 Installation Procedure Rebuild from sources? (YES or NO) yes Which version? (BLISS or MACRO) macro Install Kermit-32 on the system? yes (note, you have to have write access to the system directories to do this) This may take some time (time passes) Kermit-32 installation is complete. METHOD 3: COMPILE THE BLISS SOURCES You can use this method only if you have a Bliss-32 compiler. Even then, there's no reason to do this unless you want to change the sources in some way. Follow the steps in Method 2, except reply "bliss" instead of "macro" to the "Which version?" question.