Megarom

home *** CD-ROM | disk | FTP | other *** search

/ Megarom / Megarom Macintosh CD Software (Quantum Leap)(1992).iso / COMPRESS / Utilities / Text Programs / Mac⁄gnuucp / Getting started with e-mail < prev next >

Wrap

Text File | 1991-01-13 | 45.9 KB | 854 lines | [TEXT/MSWD]

Getting started with electronic mail by Ian Feldman, / ianf@random.se (ianf@nada.kth.se)/(ianf@dada.UUCP) (originally written for a friend in Chicago) _______________________________________________________________________90-07-21 What follows is a crash course in personal data-telecommunications, that should provide enough of a guidance to get you started. The emphasis is on supplying the answers to all the questions that someone not previously exposed to data traffic might have, rather than on the nitty-gritties of getting the computers to talk to one another. You already are a daily user of telecommunications, without even thinking about it -- your telephone is just an entry point in a vast, global telecommunications network, that's made up of several smaller, physically intertwinned ones, of which you know nothing, and could care less if you did. Which is as it should be, because it has been designed to function that way, without requiring the users to be knowledgeable about its presence. Taken on that level, a data connection can be said to function just like a "normal" telephone call, save for the difference in equipment used to send and receive the information (and one more basic distincion, non-synchronicity of communication, of which more below). Voice-telephone calls normally require presence of a person at both ends of the wire for the time of the connection. Because such "synchronicity" commands high price we later find all those zeroes before the comma on our quarterly telephone bills. Now, replace the "life" receivers at both ends of a connection with computers and suddenly you don't have to be talking in synchronous-time anymore.... or even to be physically connected all the way, in a straight line to the destination. All of a sudden it becomes possible to establish personal communication between any two points equipped with a computer and a telephone, in a time-delayed, asynchronous (or non- simultaneous), fashion, using any accessible stations along the way to provide temporary starage for one's messages. And that at a fraction of the cost of a normal synchronous (international) telephone call.... tempting? I bet. Nor do the advantages of personal data communications stop at that. Once past the threshold of setting up access to such "electronic mail" it becomes possible to send and receive news articles, to partake in online discussions on chosen topics, to exchange views and ideas with a vast, international group of users allover the (industrialized) world. The statistics of networks' size hard to come by, the audience can though easily be said to reach tens of thousands of interconnected computers and between a half and a million active, daily users. Yes.... you'd have the dubious opportunity of having me at the other end of the wire as well. The theoretical part now safely behind let us concentrate on what it takes to achieve this, to establish a platform for yet another (tele-) communications channel in your life (the radio, TV and telephone already playing a prominent role in it). Apart from all the other advantages it also means being able to communicate freely with other suitably endowed individuals, at a low and universally uniform running cost. Uniform, because the costs are proportional to the length of the received and outgoing messages, transmitted via phone to a local mail-forwarding computer, rather than to the "real" distance to end destinations/ addressees (generally far below costs for comparable quantities of paper mail, while taking at most two days to most destinations in the "civilized" world). Indeed, it can be said that the main costs of personal e-mail communications are those of acquiring the necessary hardware to make the accesses possible. Before we go into that however, let us deal with the ways and means of gaining access to the networks. Generally there are 3 types of e-mail networks that an individual can join: the academic "Internet/Bitnet", the commercial ones (CompuServe, MCIMail, etc) and the "uucp" network. Having an "account" on either one means generally having a possibility to reach users on all of them (save for some exceptions, due to restricted or one-way links etc). Having an account on a computer connected to the Internet and Bitnet depends on one working for or studying at an academic institution that already makes use of the technology. Clearly, it is the best and the cheapest alternative, not in the least because of its usually high-speed transmission links, vast number of connected nodes and people and the ready presence of terminal equipment. The commercial services are in practice limited to North American users, because of the existence of special telephone-entry points to them only in major American cities. Since they are geared mainly towards business, for which they offer added services like the Dow Jones Index, etc, they tend to be fairly expensive to the average "private" user. In America there are also some 70-odd "public access unix systems" which are often private small systems that offer access to the uucp-network from home terminals for a fee. Though dispersed rather unevenly across the USA, they're fairly low-cost and an excellent choice for e-mail access, if one can only manage to find one locally. Finally, there is the global UUCP network, designed to allow exchange of information between free-standing "unix- type" computers on a "each node bears its own telephone costs" basis. Because of that, it is probably the only alternative for those not in Academia and that is why this crash course in personal data tele- communications is in reality about implementing the 'uupc' solution. There are 2 main hardware elements, 2 main software elements and several small obstacles to overcome on the road towards establishment of such a ready-to-run e-mail platform [where do all my metaphors come from, and where are they heading, I wonder]. As to the hardware -- you'll be needing a "real" computer and a suitable modem (a modulator-demodulator, that receives the computer's electronic signals and converts them to encoded sound signals, for sending over telephone wires.... yes, you can actually listen to a modem-to-modem communication, though it ain't music to my ears, honey). Regarding the "real" computer -- what you need is some machine that 1) you are comfortable with and can use for your ordinary writing/ printing out (and other word/ data processing) tasks 2) is of a type for which there already exists a 'uupc' software package (more of that later) 3) is preferably totally silent, so that you can have it switched on at all times, thus enabling you to write without any startup delays (also allowing for later possible permanent connection to your voice telephone line, to answer incoming data calls at predestined hours, when your e-mail host has incoming mail for you, that you've chosen to have delivered automatically to your station, rather that have to remember to make the call yourself. Automation of everyday chores, that's what computers are about...) 4) agrees with your pocket and present and future computing needs There are several computers for which there already exists a 'uupc' package, but your present "word-processor" model is most assuredly not among them. The ones I know of are: Macintosh, IBM PC & clones, Atari and the Amiga. Of the latter two implementations I know nothing. I personally now run it from my Mac SE at home and would probably be able to assist you (advice via snail mail, subject to any delays) with getting hold of the software and setting it up on either a Mac or an MS-DOS machine. As to which actual computer that you _ought_ to have, that's not for me to decide. Unfortunately the only Macintosh model that I could recommend is of the US$ 3500+ variety (the MacPortable with a hard drive and a built-in modem), which I can't afford myself. Still, should you be able to get hold of a used Mac Plus (original Mac Plus, not a field- upgraded model that may be lacking some hardware options - it should say right so on the back, that it is a "Macintosh Plus 1 MByte RAM, One 800K disk drive" computer) for between US$ 500-900 then it may not be such a bad affair. It is certainly capable of running the 'uupc' software, and, besides, I could provide you with on-going support and utility programs once we've established the e-mail connection. It is also silent in its original state (no fan), and has fairly good writing tools available for it. On the negative side -- it is basically the original 1983-vintage model, in an industry that is hardly twice as old (1975 or thereabouts), and with 7 years being much in computing it really shows.... that's why it is being phased out and people are parceling them out to their younger brothers, if not straight out into the attics. It has also a known problem with power supplies, that are prone to overheating and even burning. That's a US$ 150.- replacement cost. An alternative, in many ways superior to the used Mac Plus solution would be for you to buy a Toshiba T1000SE (observe the model's name; there are more Toshibas and not all of them are created equal), that is a very nice (in my opinion, though I haven't written on it much) really portable machine (less than 6 pounds in weight), equipped with a clear, readable screen (it has to be; this computer begun its life as a machine for internal-Japanese market, later widened in scope, equipped with Latin letter capabilities and sold allover the world), a good typist-keyboard, a built-in "operating system" chip and running off rechargeable batteries to boot. It is certainly a machine to have if your only needs would be writing letters, normal- sized stories (not above several hundred pages at a throw, that is ;-)) and running the uupc e-mailaccess program in between. It is totally silent, easily transportable (hmmm... you and cameras, I think that I'll abstain from any further comments) and costs approximately US$ 1150- new from reputable mail order houses in California. Or you could probably get it with a student's discount from the UoCh on-campus computer store.... since it is fairly new, a year or so, it may be hard to find used. The modem: it has to be a good-quality 'Hayes-compatible' modem, capable of running at the speed of (at least) 2400 baud (number of transmitted information bits per second -- ie, there are no 'baud/ second' because seconds are already a part of it). The model I use is called 'Supra Modem 2400' Model 50-2400-0 from the Supra Corporation in Oregon, which is available for approx. US$ 129.- from the PC/ MacConnection mailorder house, that advertises in the flashy compurags... There are cheaper models, that may well be capable of the same functionality, and no more error-prone than this one (that broke once and had a chip replaced at the factory; took a month via air-mail both ways)... but I can't recommend you any other in particular.... Just remember, that it is a jungle out there, with people waiting to make fast buck off deer-eyed girls like you, all the while society keeps looking the other way, if not with benign approval (of tax dollars!) in its eye.... Of course, for the Toshiba the only option would be to get the original Toshiba modem that fits into its internal modem-slot.... or, perhaps getting a fax-modem (ie, basically an all-electronic fax, that would enable you to send, and receive faxs, just as if the Toshiba were a free-standing, paper-fax machine -- of course, it would require a dedicated telephone line for incoming-fax purposes) that is available for it.... incoming faxs may be viewed onscreen or printed out in normal fashion, yes. Faxs may also be created directly on screen (text only, no graphics) and queued to be sent off at predestined, low-rate hours of the day.... Now for the software bits of it: the 'uupc' stands for unix-to-unix- personal-computer and is a pun on the 'uucp' program (unix-to-unix-copy) that provides an entirely seamless way to exchange files, documents between computers controlled by the 'unix' operating system. Those computers range from fairly small, pc-sized ones to quite large mainframes, all basically using the same command scheme across the board. In contrast to the 'uucp' (learn to observe the order of the letters and the distinction between these two names now) that requires presence of the 'unix' system on the sending end the 'uupc' is designed to talk to a 'uucp' host, and does not require that your machine should be of the 'unix' type. Ie, a Mac or an MS-DOS Toshiba will do just fine. For an explanation of how unix-to-unix comminications all begun please read about it in the October 1983 issue of the BYTE magazine, that should be available at any library worthy of the name. Also the more recent, May 1989 issue of BYTE, is a must for the uucp-information- hungry, and not above the comprehension level of an average white middle- class suburban yenta. Since what the 'uucp' basically does is to accept commands in the form of "send file this to receiver that at site there" (and that's it!), then it goes without saying that the 'uupc' does the same. Arguably, it is not a program (sometimes a suite of programs that talk to one another without any human intervention whatsoever) to wax romantic over, or admire for any sophisticated displays etc... Usually it just shows its name onscreen, accepts the commands from the keyboard or from a previously saved command file, processes these, adds headers to files so that the receiving end will recognize their ultimate destination, queues the files to be sent in its own directories and either calls up the indicated mail forwarder immediately or goes asleep until some preset time to do that (once a day, or whatever). Then it starts spewing totally unintelligible access-log records, the only use of which are to pinpoint trouble spots during initial/ failed attempts to communicate. Once properly set up it and found working the log may be disregarded entirely if not erased outright. The session to the e-mail host is usually a totally automagic, user- hands-off affair.... The program will automatically log in, supply the host with the proper password, send in any own queued letters and then accept any files that may have been waiting for you there. The actual communication is of the very fast type, sending a near-continguous stream of data packets at the highest possible speed. The uucp/ uupc "streaming protocoll" is cognizant of line conditions at the time of transmission, and varies the length of packets in accord with the frequency of detected data transmission errors. The protocol's normal efficiency is approx. 60% of the theoretically possible transfer rate on a 2400 baud line (meaning an average of 120 characters per second). Needless to say, the uucp/ uupc data exchange protocol is of the fault- correcting type, automatically resending any garbled packets until confirmation has been received from the other end. A successful transmission means that the received information is an exact copy of all the bits at the sending end. In all a very stable, fairly painless and generally interaction-free way of transfering files, mail or documents, to another uucp/ uupc computer. There isn't all that much more to say about the 'uupc' program. The most important thing to add is that it is in the public domain, ie, free to use to anybody, for any purpose, without having to pay any royalties whatsoever to anyone. The original code has been provided around 1985 by several people of the 'uupc project', some of whom have continued to support and develop it further. Here I'll only mention the three 'uupc' implementations that are currently available for the Macintosh -- the original Aztec C code written by Stuart Lynne, the Think 3.0 port of it by Sak Wathanasin and the latest (1989) Think 4.0 port by Alen Shapiro. The latter two supersede the original one. Apart from that there is also Jim O'Dell's 'Mac/gnuucp' port (April 1990), providing the same end-user functionality but based on an entirely different underlaying code from the Free Software Foundation/ GNU project quarters. In all, 4 different implementations of it, just for the Mac. It goes without saying that all the 'uupc' programs, no matter their origin or intended hardware/ software platform, have roughly the same capabilities to access remote unix hosts and pretty much the same price: free. For the MS-DOS there may also exist a commercial implementation of the 'uupc', though it is unknown if it provides any added functionality to justify its price. Finally, there is the other "software element" that is needed for the full 'uupc' communication -- the mail writing/ editing and reading program. The 'uupc' is strictly for establishing of a link and transfer of files, queued at either end of the connection. No mail generation or -addressing is possible from within it. The program simply searches for any properly-named files in a preset directory on a disk, decodes addresses from their bodies and sends them on their way off. In order to write and address mail another program is needed. This may either be an ordinary text writing program, that is capable of generating text-only files (ie, no proprietary word-processors' file formats are acceptable) or a special, often an accompanying mail-program, that knows how to create properly formatted address headers in every file that is written or imported into it. Using an ordinary text-only program to write letters one has to create the header lines by hand, and insert them first in all letters. Of the two 'uupc' programs for the Mac, that were tested by me, both used the former approach, using a special, even if simple, mail- generator shell program. The Mac/gnuucp's version of it is in the from of a HyperCard stack, that also functions as mail reading, replying-to and mail archiving program. The 'uupc' (Think 3.0) version did include a mail-generator program, but not a reader companion. It relied instead on all mail being read using any word processor or other text-handling program. The HC stack is definitely the nicer method of the two! Well then, we've got the necessary hardware, have found the source for the 'uupc' software, received and installed it on the hard disk or diskette. Now what. First, read and print out any accompanying documentation, you may be needing it. Next try to read some book about uucp communications, of which most are unfortunately quite technical in scope. The only one known to me that is fairly general and should be understandable by most future 'uupc' users is the "Using uucp and Usenet", a Nutshell Handbook by O'Reilly & Associates, Inc., that is available from said company at 981 Chestnut Street, Newton, MA 02164, among other places. You may also try to establish contact with another computer owner that is already using the 'uupc' to exchange e-mail. You will need to understand the basic facets of the 'uupc' communication so that you may next start searching for a suitable e-mail host in your vicinity. Finding one should not automatically be equated, however, with getting it to function as your mail forwarder host. It is, of course, impossible to give any precise advice as to how to go about finding a host that'll have you for a mail client. There are few magazines -- Unix Review and the like -- that may be suitable for placing ads, of the "looking for a 'uucp login account' at a site near me" type. Do mention that you'll only be using it for limited amounts of e-mail. A good source of information about potential mail-forwarder sites might be a "system administrator" at the Computer Studies (or similar) department of a local university -- after all, work places with high-powered computers is where their allumni end up. Once suitable potential sites are found it is up to you to inquire/ ask/ convince them to be your e-mail host. This is nothing unusual in the Unix world, with various companies having various policies in regard to outside e-mail clients. Lastly, there are also national Unix-users' associations that will assist you with finding a suitable site in your vicinity. In the USA this organisation is: UUNET Communication Services 3110 Fairview Park Drive, Suite 570 Falls Church, VA 22042 +1 (703) 765 5050 For the really-desperate UUNET also offers direct access (ie, via long- distance telephone, if need be) to their own computer, at a price of US$ 35/ month. That's the high-end of e-mail costs, however. In comparison most sites will charge at most a plain US$ 10.- or 15.- a month, or some low-multiple of that per year. In either case, what you'll be looking for is a 'uucp login account', complete with a personal password and an own directory at the other end, that will allow you to poll the host (call it up with the 'uupc' program from your home) at least once a day during some specified time slot. The usual agreement keeps the host's incoming telephone line(s) open to your calls during non-business hours on weekdays and all days on weekends and holidays. It is all individually selectable by the host's system administrator (SysAdm) though, hours as well as the level of access, and he or she is also the person that will be the sole authority as to the cause of any initial failures to establish communications. Generally you should be wary of presenting the SysAdms with simpleminded access problems, however. They've got their own problems, remember, normally are an overworked bunch, also often of not-so-slight antisocial leanings and may turn outright hostile, if expected to solve problems that do not constitute enough of a challenge. In any case -- don't push your luck... they ARE the Masters of the Universe, of your own, private access-to-e-mail universe. Indirectly this is also why you should at least try to appear knowledgeable about the 'uupc' communications right from the start, since possible host sites' administrators may be easier to convince if they are given to understand that you'll take care of your own problem-analysis, -solving and support. That is not to say that you should abstain from any contact with a SysAdm for all times but do keep in mind that the less they hear about your access problems the better. After all, unless the host in question is that of a friend's or of some (commercial) public-access-unix type, with readily-available voice telephone support for the initial startup anomalies, there can hardly be any "human resources" standing by, just to solve every Dick's, Jane's and Little Mary's half-hearted not-quite-successful access attempts. Unix is supposed to be a challenge for the challenge-minded, remember, though it is unclear as to where it might have originally acquired that unfortunate label. Highly inappropriate, too, since unix is hardly more cryptic, or less consistent than, say, the MS-DOS, which is supposed to be an "industry standard", and the operating system of choice for such nerdy compu- gliilliteratti as Jerry Pournelle. If nothing else then unix, and all unix-to-unix communications' commands are more consistent, and share the same command/ parameter-forming principles with its other programs. Still, that's neither here, nor there for the moment. Right now we are concerned with trying to find a suitable e-mail host, usually some larger machine at an institution, that is already carrying few "outside" e-mail clients to its next-level node. You should be aware, that the very nature of the 'uucp' networking is that the communication runs along ordinary telephone wires, and the basic principle calls for every node standing for its own access costs. That means directly that your e-mail forwarder has to bear the cost of the (usually local in nature, but sometimes long distance) telephone transfers of mail that is addressed to your own 'uupc' node. Even though this cost might be negligible to them, because of otherwise huge amounts of their own mail you shouldn't expect to get it all for free, but pay them a uniform, monthly, low charge for the privilege of having your own personal mailbox at their site. Your situation -- looking for a mail forwarder -- is nothing new to them and they probably have some basic policy as to who may be admitted to their site and at what price. For that reason they may be more eager to accept someone new to unix, that is only after forwarding of small amounts of e-mail, rather than someone that could potentially be a threat to their site's security. Of course there are different rules and customs for different companies. Should your host be a small organisation, with just one or few incoming telephone lines then you'd have to accept more restrictions (ie, fewer hours, lower max. limit on files etc) than when dealing with a larger one. As long as "your" portion of their download sessions from their host doesn't significantly stand out on transfer logs you shouldn't have anything to worry about. It may not be advisable, however, to request huge files from some archive-serving site, without prior consent from, or at least advance warning to the SysAdm. In any event you ought to try to have the files appear at the host during non-business hours, or weekends, and download them, thus automatically erasing them from your host's disks, before dawn of a new working day. Your host's ordinary users can hardly be expected to tolerate space-grabbing files (that wait for your next poll-call) for more than absolutely necessary. "But I hadn't planned to send or receive any huge files" you say. Oh yes, but you will, since it is one of the aspects of the operation over which you will have relatively little control. In order to better understand the nature and principle of the transfer operation here is a description of a typical uupc-exchange of letters. Let's say someone in Europe sends you mail. Once the message reaches a central "uucp-backbone" computer in that country it will be transfered, usually via a dedicated, high-speed datacomm line, to a "uucp-backbone" computer in your own country. Here, a mail-address-decoding program decides which path to use to send it further. Eventually it winds up at a node, that your forwarder's polls to get mail from. Your message gets transfered along with other mail for your hast's local users. Once downloded, it will be unbatched (decompressed) and decoded as to whom it is addressed to. Since the message is for you, an e-mail client, then it will be prepended with a suitable header, giving details of path and time of its arrival at the site, and then deposited in your own (end-addressee's) directory. This process is automatic and not especially taxing in terms of use of resources (space and processor cycles), provided that we're dealing with limited amount of mail per day. It is difficult to give any exact figures as to what might be considered "too much mail", since conditions and availability of resources will vary between machines, but a total file size of 50000 characters (= 50KB) a day and 100-250 KB occassionally shouldn't raise any eyebrows... Now is the time for your machine to call up the forwarder (poll it), or have the forwarder call you (if that's the mutually agreed procedure, for which you'll be paying extra), to find out if there is any mail for you. If your call falls within the accepted call-up hours, and you're let inside then the 'uupc' and 'uucp' programs will query each other of what one end has to offer to the other. If you have written or replied to previous letters then those files will be transfered first to the host. The control of the link will then pass to the host, which will proceed to send any mail that may have been waiting for you at the site. The received mail files will usually be directed to a "Spool" directory on your machine. When all messages have been transferred, or if there was no waiting mail to begin with, the connection will terminate automatically. One important thing to keep in mind is that files that have been transfered from either site are by default also deleted from the relevant "spool" directory at the originating end. It should be also remembered that the 'uupc' program doesn't distinguish between sending of mail from one site to another or strictly within just one site. Thus it is possible and, indeed, recommendable to send a copy of all outgoing messages to yourself at the same time, so as to keep a record of what has been sent out and to whom.... Such addressed copies never leave your machine anyway, being simply recognized as mail pieces that are addressed to you, decoded and deposited in the incoming-mail directory. So now we have a brand-new letter from Europe, waiting to be processed in the "Spool" directory on your disk.... which is what the 'uupc' program does next, after completing the call. After that it's reading time, folks, using either the supplied mail-reading/ replying program or any word processor that's capable of reading plain-text (ASCII-) files. Should you decide to write a reply to it, the first one of these programs should generally be capable of decoding the sender's e-mail address and its "subject" line. It will then use these to generate suitably-formated header lines while allowing you to write the message by hand, or to import some previously saved text for the reply message's body. Upon termination of input the message will automatically be saved to a predefined directory on your computer's (hard) disk. Your mail is now waiting to be sent off. After launching, the uupc program will look in the same directory to see if there is anything there to send. Finding a file, it will read its beginning few (header) lines and parse (decode) the address in these. It will then create 2 or 3 (depending on version of uupc used) temporary files in the "Spool" directory (or folder on a Macintosh) that tell it what to do with the message -- transfer to the host or treat as mail sent to another addressee on your own machine. After that the uupc program will generally expect further commands to poll the host or to quit. A typical example of the first one of those is "-x1 -SmyHost", where the "x1" instructs it to select the "least verbal" level of error messages onscreen (otherwise it tends to overflow with meaningless information) while the "SmyHost" tells it to dial the number to the host called myHost (because there may be more hosts than one in your dial-out file). From now on it becomes an automatic, look-Ma-no-hands operation, ie no further manual input will be required. The host will reply and query your site for the password, which your end will supply from a previously set up file. Once the call has been accepted (ie, came within the hours during which the host will allow you to block their telephone line) your 'uupc' program will signal to the host that it has mail to upload to it.... etc, the cycle repeating itself. That's it. Get cracking. <3030> ------------------------------------------------------------------------------- [A postscriptum, written 90-12-09] Now for the US$ 64K question: how do you get hold of a uupc program for your computer? Basically there are several channels through which one can obtain one of the many implementations of uucp for various operating-system platforms (ie, the different computer models). However, the uupc being free there is no "normal" distribution of it via the retail channels. However, the same site that will be your mail-forwarder (ie the "branch for your leaf" node) will usually advise and help you get hold of it from some academic program archive (ie a [larger] computer somewhere, connected to an academic network [most often then not], that also functions as an "archive site" for freely-copyable programs and documents. Such documents (also our uupc program is just another electronic document on the network) are accesible via e-mail and from the academic sites usually also via an interactive connection method called FTP (for 'file transfer protocol'). Using the former a command requesting the uupc program file has to be sent to an e-mail archive- server site, which I'm not about to describe in detail because the exact syntax for requesting this may vary from site to site. Usually sending a "help" command will result in the archive-server in question sending out a detailed instruction on how to continue. Instead, here is just a sample 'conversation' with an FTP-archive site (in California, USA, accessible from any site anywhere in world, that's connected to Internet) which contains, among others the uupc for MS-DOS machines (one of several implementations for it; not tested by me). First, however, how do we find out which archive site contains the program that we need? How do we find out the name for some service which we're not previously familiar with? We look in the yellow pages, of course, under more-or-less specific headings. The 'yellow pages' for the FTP-able world are the periodically republished (electronically of course) listings of FTP-sites. Any site worth name should have one; ask the guru -- he or she may keep it if for no other reason than to be able to request files by her/himself. The uupc package that I personally use (on the Macintosh under the Macintosh OS) is called "Mac/gnuucp", currently [Dec. 1990] version 3.5, may be requested directly for Jim O'Dell, <jim@fpr.com> or gotten from the info-mac@sumex-aim.stanford.edu archive site (an earlier version of it is there right now; updated now and then). This particular port of Free Software Foundation's gnuucp code to the Mac is the only one currently being supported actively by "the person in charge" ;-)) There is also now a commercial program called UMail, written by Tim Endres of ICE Engineering, Inc., that also communicates using the uucp protocoll, and offers several other bells and whistles, although at a (sizeable) price. In the end this might be the more suitable solution for those of you that feel _very_ unsure about starting some kind of computer-related datacomm activity without conventional means of ongoing support from publisher of a program. Contact ICE Engineering at <info@ice.com> or, via "normal channels" at > Tim Endres | time@ice.com > ICE Engineering | uunet!ice.com!time > 8840 Main Street | > Whitmore Lake MI. 48189 | (313) 449 8288 At the very end I've included a short article from usenet that gives one user's opinion of the newly-released UMail program. Haven't seen it myself yet though. For the rest of you, the non-Macintosh-but-IBM-PC owners here is another way of getting hold of the uupc for your machine. Please observe that there are several more (independently adapted) implementations of it, with various specific names, though I've not used any of them. However, obtaining one from an archive site may give you some assurance that the program has been found more-or-less working (or it'd have been removed after complaints from those that tried to install it & failed make it work). I know from personal experience that many academic sites in Finland (of all places ;-)) have well-stocked archives (partial listing follows). When everything else fails -- try one of these: funet.fi 128.214.1.1 msdos, minix, FUNET/NORDUNET info, hydra.helsinki.fi 128.214.4.29 misc, TeX, X, comp.sources.misc, jyu.fi 128.214.7.5 unix, atari, amiga, mac, etherprint, kampi.hut.fi 130.233.224.2 DES routines (unrestricted), GNU kolvi.hut.fi 130.233.160.32 kermit stuff, radio amateur stuff lut.fi 128.214.25.8 GIF, PD sources modified for hp-ux, sauna.hut.fi 130.233.251.253 GNU, unix, amiga, athena docs, some tolsun.oulu.fi 128.214.5.6 amiga, atari, c64, msdos, mac, irc tut.fi 128.214.1.2 Images, lots of misc. unix uwasa.fi 128.214.12.3 mac, pc, suntools, unix, vms, os2 vega.hut.fi 130.233.200.42 msdos, mac, Kermit, fusion docs, However, back to our quest for an uupc for the MS-DOS platform. I start by finding out if any site in my own "ftpsites.txt" file has a listing explicitly for the 'uupc' (and also for the 'uucp' because those two are easily taken for one another although they describe program suites for different platforms -- the 'uucp' one meaning "unix-to-unix-copy"). All emphasis via "^^^^^^^^^^^^^^^^^^^^^^^" and comments with "<-----" arrows added by me to he output to easier describe what is happening. % fgrep uupc ftpsites.txt <--- looking for 'uupc' in 'ftpsites.txt' j.cc.purdue.edu comp.sources.amiga, elm, uupc, NeXT, van-bc.wimsey.bc.ca 128.189.233.155 interim (uupc), comp.archives ^^^^^^^^^^^^^^^^^^^ [Incidentially, the "van-bc" site in Canada is the original "home" of the uupc, ie the site where one of the major authors of it, Stuart Lynne, worked at the time] % fgrep van-bc.wimsey.bc.ca ftpsites.tex <--- repeating the search for more occurences of 'van-bc...' in the same 'ftpsites.txt' file [Alas, ther 'purdue' site did not contain any uupc or uucp programs, or, perhaps, I couldn't find them and there was no FTP-connection to 'van-bc' when I was writing this, so I continued looking, this time for the uucp <---cp!! keyword] % fgrep uucp ftpsites.tex <--- looking for 'uucp' bulldog.cs.yale.edu 128.36.0.3 dict, uucp paths, kafka handies.ucar.edu 128.117.64.4 uucp maps, src, feminism archive polyslo.calpoly.edu nethack, nntp, uucp, hosts, TR, ^^^^^^^^^^^^^^^^^^^ ^^^^ [Now, the 'polyslo' is a well-known archive site, so I tried that. First one more search in the file, this time for the 'polyslo' keyword] % fgrep polyslo.calpoly.edu ftpsites.txt <--- below's the full listing polyslo.calpoly.edu 129.65.17.1 xtrek, top 2.0, spaceout, cnews, polyslo.calpoly.edu nethack, nntp, uucp, hosts, TR, polyslo.calpoly.edu RFCs, Conquer Docs, usenix tape polyslo.calpoly.edu files, short usenix files, polyslo.calpoly.edu Hitchhikers guide to the Internet, polyslo.calpoly.edu Internet Email list [Now, let's connect to the site and find out if does have any uupc instead of the uucp package!] % ftp polyslo.calpoly.edu <--- connecting Connected to polyslo.calpoly.edu. ---> the response's instanteneous 220 polyslo FTP server (Version 5.52 Sun Apr 1 18:17:32 PDT 1990) ready. Name (polyslo.calpoly.edu:f87-sir): anonymous <-- must be "anonymous" 331 Guest login ok, send ident as password. Password: <-- any password will do, no echo 230 Guest login ok, access restrictions apply. ftp> dir pub <-- list the pub(lic) directory 200 PORT command successful. 150 Opening ASCII mode data connection for /bin/ls. total 1304 -r--r--r-- 1 root root 227151 Oct 4 1989 Internet.E-Mail.list.Z -r--r--r-- 1 root root 878 Nov 14 15:36 README drwxrwxr-x 2 root root 2048 Feb 6 1989 TR -r--r--r-- 1 root root 95233 Apr 12 1990 byteunix.tar.Z -r--r--r-- 1 root root 33697 Aug 2 1989 conqdoc4.3.sh.Z -r--r--r-- 1 root root 37502 Aug 2 1989 conqdoc4.3.tar.Z -r--r--r-- 1 root root 26496 May 27 1990 distribute.shar -r--r--r-- 1 root root 379708 Sep 30 15:32 fsuu11r5.zip <--- uupc ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ dr-xr-xr-x 2 root root 2048 Aug 17 1989 hgi -r--r--r-- 1 root root 275599 Aug 4 1989 hosts drwxr-xr-x 2 root root 2048 Sep 20 14:41 rfc -r--r--r-- 1 root root 6403 Aug 7 1989 short_usenix_files.Z -r--r--r-- 1 root root 13228 Nov 14 15:36 spaceout.tar.Z -r--r--r-- 1 root root 24881 Jul 19 11:12 spider.tar.Z -r--r--r-- 1 root root 32257 Mar 7 1990 sxi-1.0.tar.Z -r--r--r-- 1 root root 42444 Nov 14 15:36 top2.0.tar.Z -r--r--r-- 1 root root 53955 Aug 2 1989 usenix_tape_files.Z 226 Transfer complete. remote: pub 1167 bytes received in 0.42 seconds (2.7 Kbytes/s) ftp> cd pub <--- change directory 250 CWD command successful. ftp> get README <--- download the info file 200 PORT command successful. 150 Opening ASCII mode data connection for README (878 bytes). 226 Transfer complete. local: README remote: README 901 bytes received in 0.27 seconds (3.2 Kbytes/s) ftp> !less README <--- now read it, using a polyslo.calpoly.edu anonymous FTP local program 'less' (could be 'cat' or 'more) Contents: fsuu11r5 -- An implementation of UUCP for PC/MS-DOS machines. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ hosts -- SRI's host file TR -- Various Tech Reports rfc -- Internet Requests For Comments conqdoc4.3 -- tbl/*roff formatted Conquer documentation. usenix_tape_files -- Files on the 1989 Usenix Source Tape short_usenix_files -- Directories on the 1989 Usenix Source Tape (short form) hgi -- The Hitchhiker's Guide to the Internet Internet.E-Mail.list -- The Internet E-mail list form Ohio State. Use at your own risk. sxi-1.0 -- Simple X Interface (Xlib UI toolkit, ANSI C) byteunix -- Byte unix benchmarks distribute -- Mailing list software spider -- spiderbot; a robot MUD player top2.0 -- Top version 2.0 spaceout-- The master distribution for spaceout for X11R3 (program that gives the appearance of flying through a star field). set proper type for transfer 200 Type set to I. <--- response may vary ("type set to binary") ftp> status <--- check it to be sure (it's important!) Connected to polyslo.calpoly.edu. No proxy connection. Mode: stream; Type: binary; Form: non-print; Structure: file ^^^^^^^^^^^^^ Verbose: on; Bell: off; Prompting: on; Globbing: on Store unique: off; Receive unique: off Case: off; CR stripping: on Ntrans: off Nmap: off Hash mark printing: off; Use of PORT cmds: on ftp> get fsu11r5.zip <--- now download it to your end 200 PORT command successful. 150 Opening BINARY mode data connection for fsuu11r5.zip (379708 bytes). 226 Transfer complete. local: fsuu11r5.zip remote: fsuu11r5.zip 379708 bytes received in 103.2 seconds (3.6 Kbytes/s) ftp> quit <--- this is it. Now download 221 Goodbye. the file to you terminal % <-- back to my 'native' prompt and unpack using any 'zip' decompresser. Good luck. --------------- this is it; below an opinion of UMail ------------------------ > From: wallich@iti.org (Ken Wallich) > Newsgroups: comp.sys.mac.misc,comp.sys.mac.wanted > Subject: Re: UUCP/mail/news for Mac? > Date: 8 Dec 90 02:49:55 GMT time@tbomb.ice.com (Tim Endres) writes: # ICE Engineering sells "uAccess" a full UUCP/Mail/News implementation # for the Macintosh. Full interface for news and mail and administration. # Disclaimer: I work for ICE Engineering, and thus I am biased. I don't work for ICE, but am a customer. I installed uAccess Monday, and only 1 hour later, threw out uupc and gnuucp (well, I archived them :-). It is very easy to configure, compared to UNIX, but can be confusing if you have never done it before. The step by step instructions in the manual are quite good, but can be a little confusing at times. It packs a lot of power into a very simple interface, and provides a lot of functionality. I have encountered less than half a dozen things I'd like enhanced, which is remarkable considering how picky I am about my interfaces. I've only encountered one bug, and that is a problem with the Communications Toolbox. Occasionally, uAccess with complain about having problems talking to something in the CT, and I have to reboot the Mac to clear up the problem. No crashes or anything nasty though, so it isn't an insurmountable problem. I'll be doing my full blown configuration this weekend, adding additional mail clients, in addition to my main mail and news servers. Something that was simply too hard to bother with with the public domain packages. Also, I haven't started getting my full-blown newsfeeds transferred to the Mac yet, so I'm not certain how well it will deal with the volume :-(, or how well the interface will support wading through everything. My main complaint with the interface at this point is that it remembers the configuration of some windows (i.e. placement and size), but doesn't remember others. For instance, the Mailboxes window comes up right where I left it last time, but it doesn't put the individual message windows where I last put them, or have a concept of a default message window size and location. This will be a problem with news, for me at least, since I have a 19" monitor, and would like my messages/articles to all come up large in a place on the screen that isn't where uAccess wants to put it by default. I may try to tweek this with ResEdit, but it is something that I hope to see taken care of in the next version. If you have a little 9" Mac monitor, I think the default interface probably would be ideal. All in all, this package is even better than it was advertised to be. =============================================================================== Good luck! Once you've set it up send me a "hello" to <ianf@random.se> Stockholm, Sweden, 90-12-09 --Ian "the missing link between Artificial Intelligence and Natural Stupidity" Feldman / ianf@random.se /(ianf@nada.kth.se)/