18-Jul-86 14:14:08-EDT,3613;000000000011 Received: from CUVMA by CU20B with HASP; 18 Jul 86 14:14:01 EDT Received: from SEGUC21(GUCJS) by CUVMA (Mailer X1.23a) id 4003; Fri, 18 Jul 86 14:13:02 EDT Date: Wed, 16 Jul 86 14:29 -0200 From: Jonathan Scott Subject: KERMIT is now too good ... To: Frank da Cruz Comment: cc Lars R?kaeus , Stefan Lundberg We are having a problem at the Gothenburg Universities Computer Centre because of the high quality of the terminal emulator software provided as part of the excellent new KERMIT-MS V2.29 level for the IBM PC. The problem is that if we give it away effectively for free as suggested in the accompanying materials this will make it very difficult for us to continue selling other terminal emulator packages. These other packages have various advantages such as providing graphics support or slightly better facilities in full screen mode, but it is difficult to compete when there is a very good program available for no cost at all. We have also developed a package of key definitions to make KERMIT-MS behave almost exactly like our own locally-written terminal emulator TERM-S1, and it is difficult to see how we should charge users for this command file (although admittedly it only took a few hours to set up). Another task that we have also undertaken to help our KERMIT users is that we have recently written a KERMIT command procedure in our MVS GUTS command language (with a certain amount of support in GUTS itself). This provides basic KERMIT support plus a limited number of server and host command functions, but unlike the previous modified version of MVS TSO KERMIT it allows direct access to the GUTS library rather than working with OS data sets. This took two days to implement and debug. (Does anyone else have an interpreter-based KERMIT implementation?) To enable KERMIT and graphics procedures to operate through a Series/1 or IBM 7171 control units, we have changed GUTS so that it is possible to switch between line by line mode and full screen mode in the middle of a session (or even the middle of a procedure), and all programs and procedures executed in line by line mode think that they are running on a normal ASCII line. All of these things mean that we find it very frustrating that we should be expected to pass on KERMIT-MS almost free in unlimited quantities. I am considering the following solutions to this problem. Could you please comment on what you think of these according to your principles for redistribution of KERMIT, and tell me if you have any suggestions of your own: 1. Give away KERMIT-MS plus the key definitions free with any other PC terminal emulator sold here (and also perhaps with other software and services). This would probably be the easiest solution for us. 2. Treat KERMIT-MS together with our local key definition package as a new "product" to be sold for a price cheaper than any other terminal emulator but significantly more than the package cost. 3. Forget KERMIT-MS and integrate KERMIT protocol support into our own terminal emulator, so that we can sell the terminal emulator for the same price with KERMIT protocol support as an added feature. This would be the best solution for our revenue, but it involves a lot more work and will probably produce a poor KERMIT within a good terminal emulator. Thank you for your time. Please feel free to circulate this letter more widely if you like. Jonathan Scott (GUTS development manager)