Aucbvax.2878 fa.unix-wizards utzoo!decvax!ucbvax!unix-wizards Mon Aug 31 22:12:23 1981 UUCP >From Lauren@UCLA-SECURITY Mon Aug 31 22:05:06 1981 Let me say this about that! For some time now, I have been working at building a CP/M // MARC version of UUCP (written in BDS C). This is an independent implementation completely separate from any Bell code -- nobody in their right mind would WANT to use the real UUCP code for anything anyway, as many of you have no doubt discovered when you tried reading it. When I first started this project, I was told that I was crazy (well, people tell me that all the time anyway), and that it would never fit on an 8080/Z80 system. Hogwash. It will work and it is almost finished (well, the packet protocol part of it anyway). The tricky part was learning how the hell the thing worked. I spent alot of time roaming around Usenix (what a stupid name!) conferences asking people if they knew how UUCP worked inside. They all laughed at me. "How it works, INSIDE? You MUST be kidding!" **Sigh**. So, I decided to figure it out for myself. This involved direct contact with some helpful Bell Labs people who were willing to push the release of certain documents through the Western Electric labyrinth so that they could be released to me. After a number of months, I actually found myself with a copy of the packet protocol documentation. It was interesting reading, but still was not nearly enough. Not only were a number of protocol situations not discussed, but there were a number of rather unclear elements as well. Also, the document made no mention of the complexities in protocol negotiation and the trickiness of the higher level protocols that operate through the packet driver. Needless to say, the UUCP documentation included with the system is also completely insufficient in this regard. So began a long series of messages and calls between a variety of Bell Labs (and other) people and myself. Little by little, the haze began to lift from the Truth, and at this time, I *claim* that I fully understand all aspects of the protocol. However: 1) It is not clear if I can release the original document for general distribution -- hassles about such things I do not need! 2) The document is not NEARLY enough. Alot of the nitty-gritty details that are essential to understanding the protocol only exist in fragmented messages I have buried away. Some of the other details are (horrors!) inside my head and not written down anywhere at this time. There are apparently no other documents that discuss UUCP internals. If I had the time, I could write up a nice paper that completely explains all this stuff and would enable anyone (well, almost anyone) to build a UUCP interface for almost any computer or operating system. But I don't have the time right now. There is a good chance that I will give some sort of UUCP seminar/talk/gab session at the January Usenix meeting here in L.A. ("Everything you always wanted to know about UUCP but were afraid to ask?") Short of that, I am not too sure what to say. If anyone has any specific questions, I would be glad to try help, but please note that I am not talking about problems with the Bell UUCP code itself -- only about the internal protocols and operations of a system-independent UUCP implementation. By the way... have you ever wondered WHY the Bell UUCP code looks so, uh, complex? One reason is that it is FULL of hooks for experiments and protocols that were never implemented, or at least were never released outside the Bell System. Another important fact is that the various elements (packet driver, higher level protocols, etc.) were written by completely independent people, some of whom did not necessarily treat the other people's work as other than a "black box". One thing's for sure, I was unable to find a single person at the Labs who understood all aspects of the various protocol levels -- that's why it took me so long to figure out what was going on. In any case, my micro-UUCP implementation is proceeding, and when it's complete you will no doubt hear about it, since a number of groups have expressed interest in the result. I have no idea how or on what basis (or IF) I will distribute it, but time will tell... --Lauren-- ------- ----------------------------------------------------------------- gopher://quux.org/ conversion by John Goerzen of http://communication.ucsd.edu/A-News/ This Usenet Oldnews Archive article may be copied and distributed freely, provided: 1. There is no money collected for the text(s) of the articles. 2. The following notice remains appended to each copy: The Usenet Oldnews Archive: Compilation Copyright (C) 1981, 1996 Bruce Jones, Henry Spencer, David Wiseman.