Restricted Review command on Listserv@DB0TUI11 If I ever had known that the main purpose of Listserv is to review lists and only a small side effect of its implementation allows the distribution of mail, I wouldn't have delayed this explanation so long. Sorry, this is very long. You may find that I have written a lot about things which are not related to the Big Question "why is this %#^#@%$ guy at DB0TUI11 fiddling around with list headers of peered lists, especially with this Most Important Review field (MIR, isn't this the russian word for peace? :-) Well, you are right. You will find some nebulous pointers to "a better solution". Better than Listserv? Arrrgh, of course not :-) I think Listserv is an excellent tool to distribute the news of this (network) world. But I will cry if I ever again have to read mailing lists by peeking and poking in my virtual reader. If I ever get this letter written tonight, I will explain this solution in another article. Some preliminary remarks ------------------------ Why have I installed Listserv at this node? (That's what I'm asking myself every other day :-) To provide a service for our local users, and for users in the Berlin area, and to do my part to keep the network working (even if the latter isn't as important as it was, but in the beginning there was no Mail-via Distribute in Listerv). I'm trying to keep as much lists as possible here, really not just for fun, but to give users a high probability to send a subscribe/unsubscribe command to "their" Listserv, instead of looking out for the "right" Listserv (even if there is now an even better solution - at least for local users. But that's another cup of tea) Why am I interested in peered lists (instead of redistributions)? Some people are thinking that peered lists are superfluous due to the Mail-via Distribute feature. I don't think so. It is true that peered lists are not necessary to reduce the network traffic. But they are helping to provide a better service. If all lists were peered, a user could send his/her contribution to "his/her" Listserv. With hierarchical redistribution lists this is a mess. BTW, a similar problem exists for some Bitnet redistributions of Arpa-Lists, where you have to send letters to the Arpa-list to reach all subscribers (and not only the Bitnet part). This could be solved without bothering the end user by such nasty details. An example for such a solution is INFO-M2 (Bitnet list) = Info-Modula-2 (Arpa list) = comp.lang.modula2 (Usenet news group), which are fully bi-directional interconnected (and of course with a Review restriction here :-) Just wondering -------------- I think all owners of peered lists prefer to have the same parameter setting on all peers. Me too. Sometimes there are reasons so this is not possible. If you don't have disk space, you can't keep notebooks. This may be really annoying for users (if another peer HAS archives online, but has them restricted to the subscribers of that peer). Nevertheless, such a technical reason (missing disk space) will usually be accepted. But sometimes there are differences in the list configuration for which no technical reason is apparent. Strange, very strange for us techies, right? :-) Conclusion ---------- This seems to be the wrong place for a conclusion as the main part of this letter follows below. But this is a long letter, and you may wish to have an opinion without reading all this stuff. If enough people are thinking that it isn't tolerable to have a peer at DB0TUI11 that has contrary to other peers a restriction on the Review command, I will be really glad to kick off each and every peered list here. I don't like differences in the set up of peered lists more than you, but there are reasons that seem more importatnt to me. I think it should be possible for a user to perform all mailing list related things at his/her local node, especially as I expect the number of users not familiar with details of e-mail to grow rapidly in the future. And I really have a better use for my time then to explain them why to send this command to Listserv at Anywhere and that letter to Foobar-Request at Arpaland. This is why I've tried to add DB0TUI11 to some peered lists. But as I see a better solution now (at least for users on local machines), these peered lists are providing a service for users at sites which aren't paying me, so why should I care? And why the heck is Review restricted at DB0TUI11????? ------------------------------------------------------ Each and every list here has this restriction. This isn't an explanation, of course. There are several reasons, the ordering of the following explanations doesn't imply anything. Please note that for the following it doesn't matter how confidential ones name and address are. I don't think that privacy of data is a previlege only for CIA and KGB agents. I don't think that anybody has to justify why s/he would like to keep data - how unimportant they may be - confidential. I don't think that anybody has to do anythink to prevent broadcasting of their data. 1) The law ---------- It is my understanding of the German law that "Review= Public" doesn't conform to this law. I may be wrong, it's just my understanding. In general you have to inform a person whose data you store in a file. I think that Listserv's acknowledgement of a subscription fulfils this requirement. You may use this data to get your job done (so the newspaper boy and Listserv may bring you the latest and greatest news just to your door), but you are not allowed to give these data away. You have to ask that person before you do so. It doesn't matter if "everybody knows about the Review command". A simple fix would be a modification of the notifications that Listserv is sending ("Your name and address and the fact that you have subscribed to list FOOBAR-L is now public domain. If you don't like this then #### off"). You guessed it, I don't like this all-or-nothing approach. As said above I may be wrong in my understanding of the law. But if I'm right, than all Listserv maintainers in Germany should check their list headers. 2) My personal opinion ---------------------- It's one of the rare occurences where my opinion conforms to the law (or to my understanding of the law, considering that I'm not completely sure) - at least partially. I think we all have to learn that information is a precious thing. And we have to learn damned fast. But why is this so hard? It is obvious that information has a value, there are a lot people who earn big money this way. And as there is enough commercial interest in this business of storing/processing/transferring/retrieving information, the society is willing to kick out some bucks to pay us. Or did you think you are paid for your beautiful smile? Immediately you will reply that nobody gives a damn whether Joe User has subscribed to FOOBAR-L or not. Right, and if nobody gives a damn, then why anybody should know about Joe User's subscription? The crucial question is to whom does information belong, whose property is it? The question is not, whether your name and address is such a secret that it has to be protected. The question is, whether I am allowed to give away data you gave me to get the FOOBAR-L mailings. Maybe you are interested in everybody being able to retrieve your name and address, maybe not. As it is now, the user has no choice, and therefore I'm protecting the data by the Review restriction, even on the risk that this is against the intention of the huge majority of subscribers at this Listserv. You may think "ouuugh, these Germans, always seeing a problem". I don't think that I'm a typical German (and most of the typical Germans will share this opinion :-) but you are right, I'm thinking that it is an fundamental question. Today the majority of mailing lists are dedicated to network related issues or are computer related at least. The average subscriber is a more or less experienced user of computers and e-mail. In the future electronic communication will be used by computer illiterates more and more. The defaults we (techies) choose today will influence this future operation. Even now, where we all are friends and where is no need to hide our names by such a Review restriction, it is accepted that the AIDS list is running with such a restriction (at least that's my impression from some discussions). What's bad with being interested in AIDS related discussions? You will probably find a quick answer why it makes sense to protect such an address list in a world which let you think you are in Middle Ages. But then, why should I (a postmaster) have the right to decide which list should be better protected, and whose data has to be public domain? If one has to deceide, then it is up to the subscriber him/herself. 3) Technical aspects -------------------- The nasty version: I anywhere in this world some KBytes of data are accumulated, somebody comes away and says "Hey, what wonderful data. As they are already collected, we could make some use of it" The friendly version: The interest in reviewing the list of subscribers is caused by the absence of decent name servers/user directories. I have no problems with name servers (except when they are based on data that was collected for another purpose). It makes a difference if the user may control whether s/he is added to such a data base. It makes a difference if s/he may describe the areas of interest instead of deriving this information from a list subscription. As long as reasonable name servers do not exist (or if they already exist, as long as they are not in widespread use) I have some sympathy with people who are trying to find this information by searching thru meaningless lists of subscribers. But what do you expect to find on these lists of subscribers? You will find some regular subscribers, but you will find an increasing number of anonymous servers. These are either non-peered redistribution lists or something like NETNEWS or VMNETNEW. For Listserv-based redistribution lists you could try to gather a complete list (a recursive technique which is useful for compilers but boring otherwise :-) but you will be out of luck for non-Listserv mail forwarders or any kind of news server or bulletin board systems. For several lists the huge majority of readers is on the Internet-side of the list or isn't subscribed anyway (Usenet news system doesn't require such registrations). Some random remarks ------------------- I don't expect a massive abuse of data stored in Listserv's subscriber lists, that's not the point. However I could imagine some things that I wouldn't like to support by running a Listserv. Even if I share the opinion that the subscriber lists are not "suuuuuuuch a National Defense Classified Confidential secret", they may be of commercial interest. An theoretical example: you have written a a FAL (IBM TCP/IP) device driver for the NECU/K200 control unit. You may be interested to get a feeling for the potential customer base for such a product. Knet sites may be willing to convert to FAL if there is a driver for their hardware. Same may be true for some Wiscnet sites. FAL sites may be interested to replace their DACU by a NECU/K200 box. Furthermore you may be interested to find sites which may be ready to install some TCP/IP software on their IBM. Recommended lists to be reviewed: KNET-L, WISCNET, IBMTCP-L, TCPIP-L. Even with the above mentioned restricted usability of subscriber lists, this may be a good starting point. You may find two contradictions here. 1) I say the output of the Review command isn't very useful and then I give an example for its commercial evaluation. 2) I say that name servers are a good thing, but they could be abused in the same way as the review output (or even better). To find a solution for this contradiction is left as an exercise to the reader :-) I'm really searching for examples why the review command is suuuuuuuch a useful tool. I have heard "I don't know whether my colleague has subscribed to this list and whether I should forward interesting stuff to him/her". Of course it is much better to look into the output of a machine. Don't try any form of personal communication. Don't let your colleague know that you are in the same boat, if s/he really like to know, s/he could find your name in the output of a review command. But maybe you don't tell the truth? Maybe you are a boss who would like to know what your programmers are reading during their work time. And maybe you are a programmer who would like to know what your boss is reading, as you have severe doubts that he can read anything more complicated than his pay check. ----------------------------------------- Well, it's 6am now, and I'm getting tired. I'm probably have forgotten some important points. But I recognize that I'm now in a mood where I'm not able to write serious letters. As this letter should be long enough now, so that all honourable networkers have stopped reading, it shouldn't matter if anything is missing. It's your turn now to explain why this restriction of the review command is such a terrible thing. Thomas