KERMIT USER GUIDE Seventh Edition Christine Gianone, Editor Columbia University Center for Computing Activities New York, New York 10027 May 26, 1988 Copyright (C) 1981,1988 Trustees of Columbia University in the City of New York Permission is granted to any individual or institution to use, copy, or redistribute this document so long as it is not sold for profit, and provided this copyright notice is retained. PREFACE Page 1 PREFACE Kermit is the name of a protocol for transferring files from one computer to another over ordinary asynchronous terminal connections. Kermit programs have been written for many different computers, and in general any two computers that have Kermit programs can exchange sequential files correctly and completely. This manual gives a brief and general overview of what Kermit is and how to use it, but consists mostly of detailed instructions for use and installation of specific Kermit programs. For a more detailed introduction to Kermit, complete with illustrations, diagrams, and tutorials, consult the book Kermit, A File Transfer Protocol, by Frank da Cruz, Digital Press, Bedford MA (1987), ISBN 0-932376-88-6, DEC order number EY-6705E-DP (phone 1-800-343-8321). The Kermit book describes Kermit in detail, from the points of view of the beginner, the user, the computer professional who must install Kermit programs or support their use, and the programmer who wishes to write new Kermit implementations. Also included are general introductions to computers, data communications, and file organization, plus a detailed troubleshooting guide, bootstrapping hints, and various appendices and tables. The latter half of the book is taken up by a complete description of the Kermit file transfer protocol, with programming examples in the C language, plus some analysis and comparisons of Kermit with other popular protocols such as Xmodem. The seventh edition of the Kermit User Guide (May 1988) includes chapters on new releases of most major Kermit programs, including MS-DOS Kermit 2.30, VAX/VMS Kermit 3.3, Portable IBM Mainframe Kermit 4.0, Unix Kermit 4E, Macintosh Kermit, CP/M-80 Kermit 4.09, plus a new chapter on PDP-11 Kermit. History and Acknowledgements The Kermit file transfer protocol was designed at the Columbia University Center for Computing Activities (CUCCA) in 1981-82 by Bill Catchings and Frank da Cruz. Bill wrote the first two programs, one for the DECSYSTEM-20 and one for a CP/M-80 microcomputer. The initial objective was to allow users of our DEC-20 and IBM 370 timesharing systems to archive their files on microcomputer floppy disks. The design owes much to the ANSI and ISO/OSI models, and some ideas were borrowed from similar projects at Stanford University and the University of Utah. The protocol was designed to accommodate the "sensitive" communications front end of the full-duplex DEC-20 system as well as the peculiarities of half-duplex IBM mainframe linemode communications. The protocol was soon implemented successfully on our IBM mainframe systems under VM/CMS by Daphne Tzoar of CUCCA. Meanwhile it was becoming apparent that Kermit was useful for more than just file archiving; IBM PCs were beginning to appear in the offices and departments, and there arose a general need for file transfer among all our systems, as well as a need to use the IBM PCs as terminals. Daphne soon had prepared an IBM PC implementation. After our initial success with Kermit, we presented it at conferences of user groups like DECUS and SHARE, and began to get requests for it from other sites. Since we had written down a description of the protocol, some sites wrote their own implementations for new computers, or adapted one of our implementations to run on additional systems, and sent back these new versions to us so that we could share them with others. In this way, Kermit has grown to support nearly 300 different machines and operating systems; it has been sent on magnetic tape or diskette from Columbia University to nearly ten thousand sites all over the world, and has reached many thousands more through various user groups and networks. Thanks to the hundreds of individuals and institutions who have contributed to the Kermit storehouse over the years. The Kermit protocol was named after Kermit the Frog, star of the television series THE MUPPET SHOW; the name Kermit is used by permission of Henson Associates, Inc., New York. Page 2 Kermit User Guide: PREFACE () Disclaimer Neither Columbia University, nor the editor, nor the authors of the individual chapters, nor any individual or institution contributing Kermit programs or documentation to the Columbia University Kermit Distribution, acknowledge any liability for any claims arising from use or misuse of Kermit programs or for inaccuracies in the documentation or bugs in the programs. Kermit programs are produced on a voluntary basis and contributed freely for public use in the hope that they will be useful, but without any kind of warranty or guarantee, or any commitment to address or fix problems. In practice, Kermit programs and documentation are contributed in good faith, and will be supported on a best-effort basis, time and other commitments permitting. Customizing This Manual Although the Kermit User Guide was produced at Columbia University, all attempts have been made to keep it free of site-specific information. However, due to the large number of Kermit implementations, descriptions of each one would make the manual prohibitively thick. Therefore, the manual is sent from Columbia with specific documentation about a selection of systems. Some of these descriptions may not be of interest at your site, while others that are may be lacking. Each site, upon receiving a Kermit tape, may decide which versions of Kermit are important to it, and include the appropriate documentation in this manual. This is most conveniently done if your site has the Scribe text formatting system (from UNILOGIC Ltd in Pittsburgh PA, USA), with which this manual was produced. Scribe runs on a wide variety of systems. There are also Scribe subsets, such as Perfect Writer and Final Word, that run on various microcomputers. Many have asked why Scribe is used for Kermit manuals instead of TeX. The answer is simply that TeX can only produce output for typesetters, not plain-text ASCII files, which are necessary for online documentation. The system-specific parts of the Kermit User Guide are included with "@INCLUDE" statements at the end of the Scribe source file for this manual, whose filename is KUSER.MSS. You may add or delete @INCLUDE statements to suit your needs, and run the result through the text formatter to produce a customized manual. If you do this, you should include an indication on the title page that the manual has been customized for your site. Not all system-specific documentation is provided in .MSS (Scribe input) format, since some Kermit contributors do not have Scribe at their sites. In that case, you will either have to add Scribe formatting commands, or else enclose the whole subfile in @BEGIN(VERBATIM)...@END(VERBATIM) brackets (and replace all atsigns (@) in the text with double atsigns (@@)). If you do not have SCRIBE, you may still use an editor to delete or add sections to the finished documentation file, though the results will not be as satisfactory -- the table of contents, index, cross references, and page numbers will not be automatically adjusted. If you are running a version of Kermit for which adequate documentation has not been provided (after all, this is a distributed, volunteer effort!), please feel free to write some, preferably in Scribe input format, and send it back to Columbia so that others may benefit from it. Likewise if you produce a new implementation of Kermit. If you don't know Scribe, you can use one of the existing chapters as a model. How To Get Kermit Page 3 How To Get Kermit The Kermit software is free and available to all, source code and documentation included, from a variety of sources. For example, most universities are connected to academic computer networks from which the Kermit files at Columbia University can be reached. The Kermit files are also available from many user groups, dialup information or bulletin board services, diskette reproduction services, and private volunteers. Kermit software is not in the public domain. The Kermit manuals and most Kermit programs bear copyright notices to protect Columbia University and the various contributors from having their work taken by others and sold as a product, for profit. This is not to say that the Kermit file transfer protocol can never be included as a feature of a commercial product; the conditions under which this may be done are spelled out in a flyer POLICY ON COMMERCIAL USE AND DISTRIBUTION OF KERMIT. Columbia University distributes Kermit programs by mail order on various magnetic media (primarily 9-track reel-to-reel tape and certain kinds of diskettes), charging a distribution fee to defray costs for media, printing, postage, materials, labor, and computing resources. This is not a software license fee; no license is required. To receive a current list of Kermit implementations, the statement on commercial policy, and a Kermit order form, write to: Kermit Distribution Columbia University Center for Computing Activities 612 West 115th Street New York, NY 10025 Everyone is free to copy and redistribute Kermit programs and documentation, and is encouraged to do so, with the following stipulations: Kermit programs should not be sold for commercial gain; credit should be given where it is due; and new material should be sent back to Columbia University at the address above so that we can maintain a definitive and comprehensive set of Kermit implementations for further distribution.