F I D O N E W S -- Vol.12 No.25 (19-Jun-1995) +----------------------------+-----------------------------------------+ | A newsletter of the | ISSN 1198-4589 Published by: | | FidoNet BBS community | "FidoNews" BBS | | _ | +1-519-570-4176 | | / \ | | | /|oo \ | | | (_| /_) | | | _`@/_ \ _ | | | | | \ \\ | Editors: | | | (*) | \ )) | Donald Tees 1:221/192 | | |__U__| / \// | Sylvia 1:221/194 | | _//|| _\ / | | | (_/(_|(____/ | | | (jm) | Newspapers should have no friends. | | | -- JOSEPH PULITZER | +----------------------------+-----------------------------------------+ | Submission address: editors 1:1/23 | +----------------------------------------------------------------------+ | MORE addresses: | | | | submissions=> editor@exlibris.tdkcs.waterloo.on.ca | | Don -- don@exlibris.tdkcs.waterloo.on.ca | | Sylvia max@exlibris.tdkcs.waterloo.on.ca | +----------------------------------------------------------------------+ | For information, copyrights, article submissions, | | obtaining copies of fidonews or the internet gateway faq | | please refer to the end of this file. | +----------------------------------------------------------------------+ ======================================================================== Table of Contents ======================================================================== 1. Editorial..................................................... 2 2. Articles...................................................... 2 Philosofical research of Germers Message.................... 2 As always, I was and remain utterly correct................. 3 DogCollar: making PGP safe for FidoNet...................... 4 Madam Emila confessional v.2................................ 5 RanD@f101.n16.z1.fidonet.org................................ 5 Scary future??.............................................. 7 What Fido IS................................................ 7 Just one question........................................... 9 OS/2 archiver test.......................................... 9 A Solution to the Nodelist Problem.......................... 11 What Is StormNet?........................................... 21 3. Fidonews Information.......................................... 24 FidoNews 12-25 Page: 2 19 Jun 1995 ======================================================================== Editorial ======================================================================== Perhaps every failed policy complaint should require that the filer resign for annoying the person that must process it. After all, if the policy complaint failed, then the complainer was too easily annoyed, by definition. People that cannot behave in a civilized fashion are annoying to most people. The page down key is available to every single person in fidonet. ======================================================================== Articles ======================================================================== Philosofical research of Germers Message Hydiho Editors! I just read the article of Bob Germer in FNews1224. I am currently finishing my essay in my study as philosophical researcher of old captain-logs, government documentents in the 17th century and other human handwriting. This also involves making statements of the writers of criminal letters. You can think of criminals who wrote letters, or even people who waided through newspapers, cut big bright shiney capitals out of it and compose these capitals into sentences and even complete letters without making grammatical errors. These people write these kinds of letters to get $100000 from you for something you really don't need, blackmail you or just because they like to frighten people. Although I cannot see Bobs handwriting, I can see some extraordinary ways of reasoning in his written message. I have studied his paper for over two hours now, and I think I already have deducted a few very interesting aspects. Instead of boring you with all kinds of silly little details and naming all these aspects I wish to make the final statement directly in the following paragraph. The only thing I can extract out of B. Germers message is that this person does not want to frighten anybody, neither does he want money in court, and he doesn't look like a criminal either. He is just a plain fool. The best way to cope with this is to totally ignore him and don't spend any attention at this person in your issues of Fidonews. This should be easy because this person isn't in the Net anyway. I also recommend to withdraw any conference names with "GERMER" in it, just to make certain not to wake this unnecesary thread ever again. Let's close this now and for good. Thank you for your attention, FidoNews 12-25 Page: 3 19 Jun 1995 Greetings from the Netherlands, Prof. Drs. Melle (2:281/705.29) University of Bokkiewokkie. ---------------------------------------------------------------------- As always, I was and remain utterly correct Fredric L. Rice 1:102/890.0 (818) 335-9601 tstream@centcon.com When I first suggested the immediate ejection of any BLOTTED- BLOT who spouts/threatens 'lawsuit' to FidoNet SysOps, the positive response was overwhelming and only one critic stepped forward publically in FidoNews to rebutt the Policy addition. At the time I warned that the spewers of ideological hatred would be the eventual death of FidoNet as a viable, useful hobby. They can't win an argument rationally because their positions of hatred are undefendable. They can't win in the courts but they're not _interested_ in winning: They rely upon the S.L.A.P.P. lawsuit -- intimidation and financial ruin to 'win' their undefendable positions. The spewers of hatred know that the bright, positive, happy, fulfilled individuals they target will rightfully judge the effort of defense and quietly bow out. My fears have certainly been vindicated, haven't they? As usual the religious zealots among us can't be happy with both the freedom and the enjoyment the rest of us experience in our hobbies (FidoNet is but one positive human achievement under attack by superstitious ignorants) and they've got to beshit and befoul the honest, loving, positive, hard-working among us due to their petty, religion-demanding hatred, spite, and resentment of all that's positive. FidoNet mistake number 1 was the vote to disband the IFNA which could have been used to successfully counter the bastards among us. FidoNet mistake number 2 was the abolishment of a yearly dues to be listed in the nodelist payable to the IFNA. These monies could have also been used to successfully counter the bastards among us. FidoNet mistake number 3 is the continued lack of any Policy statement strictly forbidding the threats of legal actions, making the offense grounds for immediate ejection. All SysOps who wish to be listed in the nodelist is already supposed to read Policy 4 and, if they agree with the terms, submit their request. Any threat of legal action against another SysOp with the rule in effect would have been a clear violation FidoNews 12-25 Page: 4 19 Jun 1995 of the agreement which resulted in connectivity. FidoNet mistake number 4 is to continue to allow ejected individuals (and entire systems of individuals) access to FidoNet and even at times to allow ejected individuals back into the network. All decisions should be final. Joining gated 'alternative' nets should not bea loophole for allowing ejected fucks from still participating in FidoNet unhindered. 0-= I want to see Policy 5 ammended and then voted upon. I want to see George Peace come back and I would like to see everyone petition Mike Fuchs to continue in his work and ask him to not let the perpetuators of hatred win. There's already enough hate-spewers winning in the real world. ---------------------------------------------------------------------- DogCollar: making PGP safe for FidoNet by Frank J. Perricone, 1:325/611.0 Not long ago, a childish prat tried to get me in trouble by impersonating me and sending netmails to various people using my name, my EMSI signature, and even, to the very limited ability he had, my style of writing. You can't imagine, if you haven't been through this, what it's like to know that you can never know what the next thing you'll be accused of is, that it just takes one chemically-imbalanced dweeb and some freeware, and all the reputation you've built over years is gone. The only solution is PGP clearsigning your messages, and making sure everyone knows that you do. But there's a problem. FidoNet does not in general smile upon PGP signatures. Most moderators refuse to allow it. Many of the rest only allow it for people for whom it is already too late, those already under attack. If you cave into these pressures, there is no longer any point to using it ANYWHERE because the forger can always get to you in the places you don't use it. Why is it forbidden? I'm usually told because of bandwidth. I counter by pointing out that receiving 100 PGP-signed messages every day would cost you an additional $0.30 per month, using a v.34 modem. But still people object to the waste of bandwidth. I think the fact that the signature is splayed out in their faces is the more important part -- if they saw how big those SEEN-BYs were, they would have other pennies to pinch, I think. Either way, I have a solution: DogCollar. This is a program which compresses a PGP-clearsigned Fido message, eliminating almost 20% of the PGP overhead; and it hides the remainder behind kluge lines, except one tiny indicator to let the reader know that the message has been "collared". Use it again to FidoNews 12-25 Page: 5 19 Jun 1995 "uncollar" before PGP-verifying the signature. My DogCollar.lha archive contains MS-DOS and Amiga executables, and the source code included is pure ANSI stdio-based C, so recompiling for other platforms should be child's play. At present the only message base format it knows is the venerable FTS-0001, Fido's own standard; but the code was designed to make it easy to integrate support for other message base formats, so if you're a programmer using another format, I need your help. Please FREQ DogCollar.lha from me, distribute it far and wide, and help me expand its domain to all platforms, all message base formats. Maybe using this we can recapture our right to freedom on our Net. ---------------------------------------------------------------------- Madam Emila confessional v.2 Hello! Someone wants to sue me for publishing indiscriminately, so i have to tell you my "real" name. Madam Emilia is Sylvia Maxwell. "Maxwell" is my pen name which i have been using for over ten years and i consider it to be real. I never wanted to use an alias as a scapegoat, and before this lawsuit problem I have never felt compelled to tell anyone my legal name. It is Sylvia "Morscher", and i live at 128 Church Street, Kitchener, Ontario, N2G 2S4, voice phone 519-570-3137 (ask for Sylvia or Max, both of whom are me). Also, i am wondering why i am am still formatting Fidonews to be 72 chars a line or less. Do i still have to do this? ---------------------------------------------------------------------- RanD@f101.n16.z1.fidonet.org Connecticut this week put teeth into a new law designed to stop any harassment on the internet, including fidonet. Three cheers for the yankees! No tolerating any of this stuff! ------------------------------------------------------------- Harassment law enters cyberspace Conn. moves to keep on-line in line HARTFORD (AP)--Connecticut has decided to delve into the high-tech world of cyberspace to outlaw over the computer what has long been illegal over the telephone or old-fashioned "snail mail": ha- rassing or threatening messages. Gov. John G. Rowland has signed a law making it a felony to "harass, annoy, alarm or terrorize" another person over the computer--a law proponents contend merely brings Connecticut law up to date with the latest technology, but critics main- tain is an overreaction born out of ignora,nce. "I don't think there's anything es- pecially sinister" about the new law, said Rep. Patricia Dillon, D- New Haven, a frequent user of the Prodigy on-line service and the bill's FidoNews 12-25 Page: 6 19 Jun 1995 sponsor. "It simply brings existing law in to the 21s tcentury." The new law goes into effect Oct. 1. But even proponents are unsure how it would be enforced and whether it would apply to computer users outside of Connecticut who happen to target harassing messages to state residents. It already is illegal to harass or threaten someone over the tele- phone or by mail, which is known by computer users as "snail mail." Computer users who do the same will be subject to up to three years in prison and a $500 fine. And people already convicted of a felony would face up to five years in prison and a $5,000 fine. The issue came to Dillon's attention during her daily visits to Prodi- gy, where she does research for legislation and discovered strings of messages devoted to a case involving a user known as "Vito," who targeted a woman via E-mail and the system's bulletin board system. The man was posting messages that had a "terrifying, immobilizing effect on the victim" by claiming her son's mental retardation was a result of fetal alcohol syndrome and accusing her of being promiscuous Dillon said. Although the various on-line services monitor messages for profane language or conversations deemed unfit for a wide audience, Dillon said the government needs to go further to try to protect users of computer bulletin boards. "At one point, we were worshiping the Internet ... and now there's been a backlash so it's been demonized," Dillon said. "I don't think it's either one," but it's worth protecting people who use computers. However, civil libertarians and some experts in the field of computers and law believe Connecticut may be going overboard. William Olds, the executive director of the Connecticut Civil Liberties Union, said recently that the harassment statutes already are troublesome because they're vague. "What is alarming to one person may not be alarming to another," Olds said. It also begins to restrict a key benefit of the Internet: its unfettered flow of ideas and information, Olds and others said. ------------------------------------------------------------- Will Connecticut set the pace on the I-Way and become it's street sweeper? Could be if they can maintain this pace. FidoNews 12-25 Page: 7 19 Jun 1995 Scary future?? by Fredrik Bennison 2:205/300, fredrik.bennison@rosebay.ct.se What is FidoNet coming to?? After reading FidoNews 1223 and 1224 I am for the first time seriously worried about what is happening in Zone1. Of course, I've heard from time to time that it is a different net over there, that it is a much harsher environment and that you have to watch your back all the time. I didn't believe that it could be so, perhaps I didn't want to believe that, I know of no such troubles here in Zone2. But when I read that people are resigning because of some fellow sysop threatening them with lawsuits I began to see that perhaps there is something definitely wrong within Zone1. And now in FNews 1224 I read that Bob Germer threatens FidoNews with a lawsuit and that an NC has had a sysop in his net arrested for harrassment, claiming that his questions are excessively annoying. For the sake of the future of FidoNet, I hope that this settles down soon and that we won't have to be bullied by some sysop's threats. This network is all about communication, or at least it should be, we should try to solve things through communicating and not through the courts. ---------------------------------------------------------------------- What Fido IS. What Fido IS. Richard Ratledge 1:119/88 -One person's perception. June 95 I have been doing a lot of thinking about what FidoNet really is. Mostly in response to the recent resignation of George Peace from the International Coordinator position and Mike Fuch's decision to stop publishing the EchoList. Here is what I have come up with. If you don't agree, please let me know and we can hash it out. :-) To me there are two ways to look at FidoNet and a lot of the disagreement flying around is because people try to put both views together. FIRST: FidoNet IS the structure defined in Policy 4. That includes the Coordinator structure, the Nodelist, Fidonews and all the individual systems. That is ALL FidoNet is, nothing more, nothing less. But SECOND: Fido is a whole heck of a lot more. It has gobs of 'content.' Content such as the message echo areas and the file areas and some stuff I don't even know about. Many people insist that the second definition is the one that is correct and believe it is what defines FidoNet. I maintain that both views are correct. Just make certain that you stay clear about which part of the definition you are working from when you start making assertions about how things should be handled in "Fidonet." Be careful not to confuse the content part with the structure part. FidoNews 12-25 Page: 8 19 Jun 1995 My opinion is that there is little problem with the primary definition of FidoNet. Policy 4 may need a bit of updating. I just read through it and I see very little that is in need of major overhaul. BUT. When you start working from the second definition of Fido there is a huge amount of structure that is not well defined. The role of the various mechanisms providing the content to Fido is unclear to many of us who are involved in FidoNet. This lack of clarity leads to some MAJOR disagreements. This is what I see happening in Zone 1 (other Zones are vastly different, I am sure). FidoNet and the structure defined in the nodelist, IS FidoNet. Then there are the Backbone, the Moderators of each of the Echos and various message and file Distribution Systems. The Backbone has taken on the responsibility of shepherding the echos along. The Backbone has created a structure that parallels the structure of FidoNet itself. They have a ZEC, REC's and NEC's just like FidoNet has a ZC, RC's and NC's. No wonder it gets confused with FidoNet! Especially since some positions are filled by the same person in both structures. The Backbone has created it's own Policy document, Backbone Operating Policy or BOP. It may be policy but it isn't binding on FidoNet itself since only Policy 4 affects Fidonet (they may be merged someday but that is a different matter). Then there are the individual Moderators of the many separate echos. They also operate independently. They control the content of their echos and who is allowed to access the areas. Within the context of the particular echo the Moderator 'is god.' The Moderator has to decide where they want the echo to go. Usually for a wide-interest subject it is placed on the backbone. Doesn't have to be, but that is where the most exposure within Fido will occur. About Distribution Systems -- there are a multitude of them currently and there will be more to come. All the way from private distribution node-to-node to Planet Connect to the ftphub. The Backbone itself is the mother of all distribution systems. Distribution Systems are a non-issue sort of thing. There is some concern about commercial enterprises making money by distributing Fido content but most of those concerns have been satisfied. I left out individual systems and users but their role is hopefully obvious, although probably the most important in all of the net. That is where we are at. What needs to be done? There are some structural things that should be straightened out. I would like to see more definitions of the role of Moderators and what authority the Backbone has in relation to feed cuts versus the moderators' authority. Making it crystal clear to everyone concerned who orders and enforces these things would be a great boon to the net. There are a million details still to work out and many things that will let all this work smoother, in general though it works the way it is. Recent events have some people greatly concerned. I am not terribly concerned, mostly because I see a lot of insight coming from the people responsible for making decisions. I am sorry to see a few people who are less than friendly and cooperative. I tend to agree with Christopher FidoNews 12-25 Page: 9 19 Jun 1995 Baker that 'friendly and cooperative' should be a requirement to remain in FidoNet. Anyway, that is where I am at in my assessment of FidoNet. Hopefully I am not too far out in left field. Maybe it provides just a glimmer of illumination to those who don't quite understand it. Not really sure I do yet! :-) PS. I don't claim any copyright to this. Do with it what you like. :-) RR ---------------------------------------------------------------------- Just one question Rev. Shawn McMahon 1:3806/0 I just have one question for the Backbone in general, and for Z1EC Bruce Bodger in particular. This is just a question, not an accusation nor an attempt to take sides. How are Z1_WINTERS and it's ilk improper titles for an echo, but OJ_SIMPSON isn't inappropriate? Does the protection against having one's name on an echo extend only to personal friends and people who threaten lawsuits, or can we expect that it'll be applied to all echoes with a person's name in them? ---------------------------------------------------------------------- OS/2 archiver test Rev. Shawn McMahon 1:3806/0 I recently did another test of archivers. Just thought I'd share it with everybody; perhaps it might save somebody some time or money. New set of assumptions this time: 1) Only OS/2 archivers were tested. Although DOS ones run great under OS/2, making the OS/2 session wait on them is a bit of a pain so I didn't bother. 2) Test subject was over 800k of Fidonet .PKT files, containing several echoes and a couple of pieces of netmail. 3) Since all I was testing was Fidonet performance, I didn't bother timing things. In general, the smaller the archive, the longer it took to make it. This is especially true for HA and Hpack. I figured that the time to archive them was trivial compared to the time to transfer FidoNews 12-25 Page: 10 19 Jun 1995 them over phone lines. 4) I didn't test them on nodelists or nodediffs; I'm not trying to get anybody to adopt a new standard here, just pointing out what's available for your own echomail. 5) This list, although broad, is only representative. There may be other OS/2 archivers out there that I haven't tested. For instance, I didn't go through the gyrations of playing with ports of Unix COMPRESS and TAR, because I doubt anybody bothers with them for Fidonet mail when ZIP and Hpack are available for both OS/2 and Unix. 6) All tests were conducted using the default compression modes, and with the filenames stated as shown. Archiver versions are listed before each test. 7) Each archiver is copyright by somebody. Names used without permission. This is not intended as a challenge to the rights of the copyright holders. Where possible, the copyright holders are listed. ARC2 6.01P, System Enhancement Associates. TEST.ARC 513107 Note: Using the "version 5 compatibility" switch produced the exact same results. ARJZ/2 .15 Alpha. Copyright owner is listed in Cyrillic, so I don't know his name yet. TEST.ARJ 307718 HA .999beta, Harri Hirvola. OS/2 compile by Craig Morrison. TEST.HA 272772 Hpack .79a0, Peter Gutmann. OS/2 port by John Burnell. TEST.HPK 298370 Lh2 2.22, Peter Fitzsimmons. TEST.LHA 327641 RAR 1.53 beta, Eugene Roshal. TEST.RAR 308449 Info-ZIP 1.0, Mark Adler et. al. TEST.Z10 332142 Note: If anybody has 1.1, I'd appreciate a copy of the OS/2 version. 1.0 is rumored to have problems, and the rumor is borne out by the fact that 1.1 exists. PKZIP 1.01-OS/2, PKWARE, Inc. TEST.ZI1 337334 Note: I haven't the foggiest idea why Info-ZIP 1.0 beats this. Both archives test perfectly. Info-ZIP 2.0.1, Mark Adler et. al. TEST.ZIP 310640 FidoNews 12-25 Page: 11 19 Jun 1995 ZOO 2.1, Rahul Dhesi TEST.ZOO 475962 Some parting notes; ZIP and Hpack are the ones most widely available for different kinds of platforms. LHarc and ZOO are probably second. ARC comes in fifth in my experience, with the rest being fairly platform-dependant. In fact, ARJ and RAR weren't even available for OS/2 until recently. All of these listed have DOS counterparts. I know for certain that ZIP, Hpack, and ZOO have Unix counterparts. I've seen ZIP-alikes for most other common computers, as well as Hpack compiles for everything that's got a ZIP version. If your favorite archiver isn't listed, and you have OS/2 executables, let me know. I'll be glad to FREQ them on my nickle. I'm especially looking for OS/2 versions (especially unpackers) for the more platform-specific archivers, such as the plethora of Mac-specific ones. ---------------------------------------------------------------------- A Solution to the Nodelist Problem by Joel C. Maslak jmaslak@cchs.ccsd.k12.wy.us A Solution to the Nodelist Problem ---------------------------------- A problem which has plagued Fidonet since its very beginning is that of the Nodelist. I have documentation dating from 1988 which indicates problems with the St. Louis Nodelist Format (Baker 12-14). Since 1988, there have been over 65 articles published in the "Fido News" relating to the Nodelist (complete list of works consulted is not published, to save space. Contact me at the address listed at the end of this article for a copy). Other proposals have been submitted, including one from Robert Heller, which support a Internet-style Domain Name Service for Fidonet. Unfortunately, Mr. Heller was ahead of his time. The Fidonet community was not (and is still not) ready for his suggested changes. While I'm not sure what the Fidonet reaction will be to my suggested changes, I am publishing the results of over a year of research in the hope that it is adopted as a standard. This research originally began as an attempt to "give something back" to the Fidonet community. It later developed into a research project which was entered into the International Science and Engineering Fair (Maslak, "Roadblocks...") as well as the Wyoming Junior Science & Humanities Symposium (Maslak, "Distributed..."). What follows is a modified version of my research report. Many FidoNews 12-25 Page: 12 19 Jun 1995 sections which describe the basic workings of Fidonet have been omitted, while other sections, describing more advanced topics, such as the details of the data files, have been added. ================================================================= Nodelist Problems One of Fidonet's biggest problems is its extraordinary rate of growth. Between the months of July 1993 and July 1994, Fidonet grew over 30 percent! In the month of April 1995, a Nodediff was released with a size, compressed, of 519,152 bytes. During the single week ending April 7, 1995, Fidonet grew by 1,600 systems. During the entire month of April, over 2,000 systems were added to the Nodelist. This is extraordinary growth. (Note: Some Nodelist statistics were taken from Mr. Bush's article entitled, "A Review of the Fidonet Nodelist.") ================================================================= Distributed Node Information Database To solve problems with the nodelist, a distributed node information database was developed. This system would allow for Net Information Segments to be stored on systems throughout Fidonet. The Net Information Segments contain nodelist information for systems within a net. The Net Information Server would distribute the segments via the Fidonet FREQ system. The Net Information Server would be filled by the node in the position of Net Coordinator. For example, the segment for net one in zone one could be requested from 1:1/0. The segment for net 10 in zone one can be requested from 1:10/0. Then segment for zone three, net 103 in zone one can be requested from 3:103/0. A program was then written to scan for outbound mail. Should this program find a message with an unknown destination address, it requests the appropriate net information segment from the net coordinator of the net for which information is desired. Upon receipt of the segment, the required node entry is merged into the local nodelist, which is used by the Fidonet software to establish contacts with other Fidonet systems. This is accomplished by using two programs, as shown below. These programs, together with the actual project research, took a total of 12 months to complete, and, for that reason, are relatively complex. Complete listings can be found in the research notebook. Contact the author via Internet E-mail at jmaslak@cchs.ccsd.k12.wy.us, or, via Fidonet CRASH MAIL (routed mail will not reach this network), Joel Maslak@1:316/23. Although the programs which were written contact the appropriate NC, as listed above, the test procedures did not exploit this capability. Instead, the testing procedures were modeled after the "transitional mode." In the transitional mode, some systems would still keep the entire nodelist, allowing all net segments to be requested from just one system. Because of financial concerns, and because such a system would have to exist until all nets implemented a FidoNews 12-25 Page: 13 19 Jun 1995 Net Information Server, this was thought to be a valid test. ================================================================= Net Information Segment Format The Net Info Segment is very similar to the traditional nodelist, except all lines relating to systems outside of the network have been deleted. Note that the Net Info Segment for a region includes only regional independents, not members of individual nets. Note that the segment for a zone includes zone independents as well as both Regional and Network Coordinators. Neither Regional Independents nor nodes belonging to local nets are listed in this file. Files are named according to the following: xxxxyyyy.nl xxxx - Hex number representing ZONE (it is this large to allow for non-Fido zones to use this standard). yyyy - Hex number representing NET Example: 00010068.NL (net 1:104) :::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: ; Net Info Server Segment List, Generated by NETAX ; Zone,1,North_America,Surrey_BC,Bob_Satti,1-604-589-8562,9600,CM,XA, H16,V32b,V42b,V34,VFC,V32T ; Region,15,REGION15_COORDINATOR,AZ_CO_NM_UT_WY,Marv_Carson,1-602-894-87 62,9600,CM,HST,V32,V42b,XA ; Host,104,Denver_Area_Net,Denver_CO,Tom_Johannsen,1-303-455-0507,9600, CM,XA,HST,V32B,V42B,V34,VFC,V32T ,1,Co-Op_Distribution_System,Denver_CO,John_Kaufman,1-303-343-0691, 9600,V32B,V42B,CM,HST,V34,VFC ,2,Net_104_UFGate,Aurora_CO,N104_UFGate,1-303-429-2713,9600,HST,CM,XX, UGTI ,3,Net_FILES_Coordinator,Littleton_CO,Bob_Simpson,1-303-770-4969,9600, CM,XA,V32B,V42B,MNP,VFC ,4,Net_Echomail_Coordinator,Denver_CO,John_Kaufman,1-303-343-0693,9600 ,V32B,V42B,VFC,CM,XA ; :::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: Note that the segment itself would function as a nodelist! ================================================================= NODELIST.400 Format Nodelist.400 is a current list of Fidonet local nets, as well as all nodes within the local net. It is merged with info from Net Info FidoNews 12-25 Page: 14 19 Jun 1995 Segments to form a 'personal phonebook' of systems which the node contacts. For more info, contact me. ================================================================= Format of LIST.NIR It is a comma-deliminated file listing nodes for which net info segments have been requested. Example: 1,23,162 1,42,252 2,2,4 In the example, info segments for systems 1:23/162, 1:42/252, and 2:2/4 were requested. Note that since this file is only stored on the local system, it's format is not relevent. It is presented here for the sole purpose of documenting the test site's software. ================================================================= Netmail Scanner Program: 1. - Outgoing Netmail is scanned - If a message exists go to 2 - Else go to 7 2. - Read the first message - Go to 3 3. - If destination address is unknown go to 4 - Else go to 5 4. - Request Net Information Segment for destination net from appropriate net coordinator - Append destination address to list of Net Information Requests (LIST.NIR) - Go to 5 5. - If another message exists go to 6 - Else go to 7 6. - Read the next message - Go to 3 7. - End ================================================================= Nodelist Merger Program 1. - Received files are scanned - If a net information segment (*.NL) file exists go to 2 - Else go to 9 2. - Set ADDR variable to value of the net pointed to by first *.NL file - Go to 3 3. - Open list of net Information Requests (LIST.NIR) - If an address is present go to 4 - Else go to 8 FidoNews 12-25 Page: 15 19 Jun 1995 4. - Read first address from LIST.NIR - Go to 5 5. - If LIST.NIR entry's zone and net correspond to ADDR file go to 6 - Else go to 7 6. - Grab node information from net information segment (*.NL) pointed to by ADDR - Add to master nodelist (NODELIST.DAT) - Go to 7 7. - If another address exists in LIST.NIR, grab it and go to 5 - Else go to 8 8. - If another *.NL file exists, set ADDR variable to next *.NL file and go to 3 - else go to 9 9. - End ================================================================= Hypothesis It was hypothesized that the distributed nodelist would be as reliable as a non-distributed nodelist for the purpose of establishing a connection with a remote Fidonet system. Thus, as it provides a significantly smaller list of Fidonet systems, it should be adopted by the citizens of the Fidonet network. ================================================================= Phase I: Test Procedure This test used the distributed node information database system outlined previously. The purpose of this test is to determine the feasibility of the distributed node information database. Two groups of 20 nodes were chosen randomly, using a custom computer program. These nodes were from the population of Fidonet nodes in the continental US. Thus a total of 40 nodes were picked, 20 in the control group, and 20 in the experimental group. The nodes in the control group were sent a Fidonet file request. The nodes were then either placed in the 'successful' or 'unsuccessful' group depending upon the success of the initial contact. The control group used the normal Fidonet nodelist to establish the connection. The nodes in the experimental group were also sent a Fidonet Freq. The nodes were then either placed in the 'successful' or 'unsuccessful' groups, as described above. A distributed nodelist was used to establish the connection for this group. For this feasibility study, all net information segments were stored on 1:316/19 (note: This system is no longer accepting incoming Fidonet Netmail). In actual practice, the net information segments would be stored on various boards throughout the network. FidoNews 12-25 Page: 16 19 Jun 1995 Results were then compiled, using the BinkleyTerm log. All connections which were successful were grouped into the 'SUCCESSFUL CONNECTION' category, while others were placed in the 'UNSUCCESSFUL CONNECTION' category. Connections were determined to be successful if a session handshake took place. ================================================================= Phase I: Test Results CONTROL GROUP - 15 Successful connections - 5 Unsuccessful connections = 75% Success Rate EXPERIMENTAL GROUP - 16 Successful connections - 4 Unsuccessful connections = 80% Success Rate ================================================================= Phase I: Discussion This project did not test the distributed nodelist system fully. The distributed nodelist standard currently does not allow the Net Information Server to be a system other than the Net Coordinator. This poses a problem because many net coordinators run busy systems. It may be difficult to connect to such a system, delaying the transmission of the net information segment. This, in turn, would delay delivery of Netmail. Thus, in large nets, it may be appropriate to have dedicated Net Information Servers. The problem with this is that, currently, there is no method of signifying a net information server in the nodelist. It is believed that the small difference between the control and experimental group is not due to the method of how the nodelist is stored, but that it is due to the variation of systems in Fidonet. From this research, it can be determined that this project would be technically feasible. One interesting fact which was discovered during the process of testing is that the Fidonet's perception of the nodelist is quite different that the reality of the nodelist. Fidonet believes that the nodelist is updated frequently, and very few systems in the nodelist are represented incorrectly. In actuality, many systems listed in the nodelist could not be contacted. This indicates a problem with the nodelist, as it may be too large to be easily managed. A distributed nodelist may be easier to maintain, as a dedicated position would be created - the Net Information Server. This server's only responsibility would be to maintain the nodelist. Since that system would not be the net coordinator, it would not be required to settle disputes, test to see if new nodes comply with Fidonet technical specifications, nor to manage the network hubs. The system's only responsibility would be to manage the net information segment. A person who wishes to maintain high integrity of the nodelist should be FidoNews 12-25 Page: 17 19 Jun 1995 appointed/elected (I don't want to write policy). ================================================================= Phase II: Test Procedure This test used the distributed node information database system outlined previously. The purpose of this test, like phase I, is to determine the feasibility of the distributed node information database. This experiment is a further extension of the research previously conducted in phase I. Unlike phase I, only one group, consisting of 100 nodes, was chosen. These nodes were each contacted once using the St. Louis Nodelist and once using the Distributed Nodelist Standard. The nodes in the control group were sent a Fidonet file request (for a file that does not exist). The nodes were then either placed in the 'successful' or 'unsuccessful' group depending upon the success of the mail transfer, unlike phase I which determined success based upon the negotiation of a session handshake immediately after a connection is established. The control group used the normal Fidonet nodelist to establish the connection. The nodes in the experimental group were also sent a Fidonet file request. They were then placed in either the 'successful' or 'unsuccessful' group, as described above. A distributed nodelist was used to establish the connection for this group. Like phase I, all net information segments were stored on 1:316/19 (Note: This system no longer accepts incoming Fidonet Netmail). In actual practice, the net information segments would be stored on various boards throughout the network. Results were then compiled, using a custom program which determined if the individual mail packets were sent successfully to the receiving system. ================================================================= Phase II: Results Control Group - Success Rate: 89% Experimental Group - Success Rate: 86% ================================================================= Phase II: Discussion The results of this test were analyzed with a two-tailed T-test. The results of the T-test appear to indicate that there was no significant difference between the results of the control group and those of the experimental group. This would indicate that the Distributed Nodelist Standard is a feasible alternative to the Nodelist, based upon technological concerns. FidoNews 12-25 Page: 18 19 Jun 1995 The two groups, control and experimental, are a random sample of the US Fidonet population, and representative of that population. Since 100 systems were tested, a large base of data was collected. This significantly reduces the possibility of these results being attributed to chance, and this, indicates that the Distributed Nodelist Standard was successful. ================================================================= Phase I Compared to Phase II Several significant differences exist between Phase I and Phase II. First, Phase I used a control group which was independent of the experimental group, which allows the possibility of variation between groups affecting the results. Phase II used only one sample of systems, polled once to form the control group, and then pulled again, using the Distributed standard instead of the St. Louis Standard, to form the experimental group. This is one of the reasons that phase II was conducted. Phase I consisted of only 20 systems in each group. Phase II used a larger sample size of 100 systems. The larger sample size limited the possibility of randomly choosing a sample group which did not represent the master population. Two types of mailers were used in this project. In Phase I, BinkleyTerm 2.59 beta was chosen. To demonstrate the ability of the Distributed Nodelist Standard to function properly with a wide variety of Fidonet systems. The success of both phases serves to indicate that the Distributed Nodelist Standard would, in fact, function properly with various Fidonet mailers. The method of analyzing the results was also different. In phase I, success was determined by analyzing session handshakes, which are transmitted early in a Fidonet session. Phase II determined success based upon the transmission of a file request (FREQ). If the request was successfully sent, the trial was categorized as successful. To compare the phase I control group to the phase II control group, the results from phase I were re-analyzed using the methods used in phase II. The phase I control had a success rate of 55%, while the phase II control group had a success rate of 89%. These results, when analyzed with the T-test, show that there was a significant difference between the phase I and phase II control groups. The difference is attributed to the types of mailers used, as this was the only variable to change between the control groups. Thus, it appears that BinkleyTerm 2.59 beta may be less reliable than Frontdoor 2.02. Further research is warranted, though, as phase I consisted of only 20 trials. ================================================================= Conclusion It can be concluded that, based on the success of this experiment, that further experimentation is warranted. The standard tested in this project appears to have worked successfully. A system FidoNews 12-25 Page: 19 19 Jun 1995 similar to this would not require Fidonet software to be rewritten. This would ease the transition from the St. Louis nodelist format to the distributed nodelist format. As the software used for the testing procedures has been donated to the public domain, others can freely use it without fear of legal action. This project also discovered, by accident--not design--that the current state of nodelist management is not successful, when compared to commonly held beliefs. Also by accident, this project determined that a flaw may exist in the BinkleyTerm software, a very popular Fidonet mailer. Further research should be conducted to test this hypothesis. ================================================================= Statistical Analysis The test statistic of phase I was -.36972. The absolute value of this statistic is less than 1.684, the value for the 90% level of confidence. Thus, the difference observed in phase I was not significant. The test statistic of phase II was .63887. Since .63887 is less than 1.645, no statistically significant difference, within a 90% level of confidence, occurred. This indicates that the difference observed in phase II was not significant. The comparison of phase I and phase II's control group yielded a test statistic of -3.3669. Since the absolute value of this statistic is greater than 1.960, the hypothesized value for 95% confidence, the difference was significant. This indicates that some event, which occurred between phase I and phase II, may have caused a significant difference in the chance of a successful connection. ================================================================= Acknowledgements All custom programs are of my own creation. Ms. Mona Mitzel assisted with scientific research controls and analysis using statistics. Without her, I would not have the interest in the sciences that I currently have. She is an inspiration, and a model of what today's teachers are doing right. She caries with herself a highly contagious thirst for knowledge. Mr. Al Griffin, Net Coordinator (1:316), Former NC (1:105), assisted by providing this project a Fidonet node number. Lets all hope he gets well soon! (Al Griffin @1:316/23 ---------------------------------------------------------------------- -------------------------------------------------------------------- ÖÄÄÄÄÄ¿ ÖÄÄ· Â º º º ³ ÓÄÄÄÄÄ¿ ÖÄÒÄ¿ ÖÄÄ¿ ÒÄÄ¿ ÖÄÒÄ¿ º º ³ ÒÄÄ¿ ÖÄÒÄ¿ ³ º º ³ ÇÄÂÙ º º ³ º º ³ ÇÄ º ÓÄÄÄÄÄÙ Ð ÓÄÄÙ Ð Á Ð Ð Á Ð ÓÄÄÙ ÐÄÄÙ Ð Check out our WWW Page! http://www.storm.net/ --------------------------------------------------------------------- What Is StormNet? StormNet is an alternative network for use with FidoNet compatible software. We pass messages back and forth both in netmail and in echomail conferences. In StormNet, each node is welcomed and assisted in many ways by other StormNet members. We have active echomail areas and a growing file echo selection. Our echomail traffic is growing every day, and are now pulling more mail ever than before. StormNet has a friendlier atmosphere than can be found in many other networks. StormNet has been in existence for over two years, and in that time, has grown significantly. Our membership has changed from an inexperienced group of local nodes to a more mature group of people from all over the United States, Canada, Europe, and Asia. Currently, we have over 600 nodes in StormNet, and are expanding every week! * Why is StormNet here? StormNet was started for a few reasons. When we created it, we wanted to serve teens, adults, and others worldwide with a quality alternative network that is relatively cheap to pull in. Most of our high speed transfers take less than a minute. Although we prefer it, you don't have to poll every day, we are flexible and will allow you to poll whenever you like. We want to serve you with the finest quality echomail and files for you and your users. We are considerably smaller than FidoNet, and therefore do not have the overflow of mail/files often seen in its conferences. The average cost per month for one who pulls StormNet and polls daily is around $6.00-7.00 within the US. FidoNews 12-25 Page: 22 19 Jun 1995 * What are the rules like? StormNet? Rules? You've got to be kidding me. Well, it's not like we don't have any rules; all of the rules in our policy statement basically stem from one basic principle - "Be nice and use common sense." The policy's specifics were written to outline some problems which may potentially arise when people aren't nice and don't use common sense. The SNAC (StormNet Advisory Council) helps to ensure the smooth running of StormNet affairs. StormNet does not discriminate because of age, sex, national origin, religion, sexual orientation, beliefs, taste in food, opinion on world politics, or favorite color. We welcome anyone who is interested in joining a fun network to try out StormNet. We also don't allow "bashing", spindling or other forms of mutilation of groups in our newsletters, or any of our echos (excluding WAR_ZONE), and other parts of our network. * What are the echos like? We have a variety of conferences to suit most needs. If you are a user of StormNet, or a node, you can request an echo if you feel it would be active. We have echos on many subjects, A to Z (as we say SN_A to SN_Z). We have a talented staff of moderators and co-moderators, and combined with the efforts of our international echomail coordinator, keep the network running smoothly. * Internet... Yes, Internet! StormNet is now registered in the domain listings as "storm.net". All StormNet nodes can use our UUCP host to send messages anywhere on Internet at no additional cost to you. * Vanity Internet addresses The StormNet International Internet Coordinator has just recently allowed a "vanity" name that you can apply for once you join StormNet. This can be just about anything, and saves typing. For example, "John.Doe@johnsbbs.storm.net" is alot easier to remember than "John.Doe@f104.n192.usa.storm.net." * Support StormNet has support sites for different types of software. We currently have technical support nodes/echos for BinkleyTerm, EzyCom, Maximus, RemoteAccess, Storm Watch Software, T.A.G., and TurboSoft. New technical support nodes for other software are permitted and encouraged. When you apply for technical support, you will get a flag in the nodelist designating you as a support site. You also get your name added to the list of technical support nodes at the bottom of the nodelist. And finally, you have the option to add file and message echos software you support. * What we're looking for FidoNews 12-25 Page: 23 19 Jun 1995 StormNet is looking for nodes who want to pull in high quality echomail at a relatively cheap price. Ones who think they would like StormNet to be on their BBSes or computers. We also are looking for new zones to create all the time. StormNet is not just limited to North America, other continents and countries are already members, but we would like more. We'd like to expand StormNet worldwide, and we think that everyone would benefit from this and it would strengthen relations and understanding between other nations. It already has. With European and Asian nodes, StormNet has a very unique atmosphere where you will regularly chat with people not in the same country, or continent as you. We are willing to work with you to install a new zone in/on your country/continent. We are looking for Zone Hubs in Asia and other continents that we don't already have. Zone Hubs that have direct Internet connections can get their mail sent to them via the Internet. Offer only good for New Zones and Zone Hubs at this time, however. * Why should I consider StormNet? We respect each and every StormNet member and his/her rights. We offer our services to all. We have a great network setup. We want all to join and have a good time in the network that we have created. We're proud of our network. We offer a unique and quaint atmosphere not found on any other network. We guarantee it. If you have any questions, comments, or problems, feel free to contact any of the nodes listed below. We'd be glad to help you. Alan Jurison StormNet Int'l EchoMail Coordinator You may F'req or download "STORMNET" (or STORMNET.ARJ) from these nodes: BBS : Ham-Net BBS Position : StormNet Int'l HQ / IEC / Mid Atlantic RC SysOp : Alan Jurison Location : Manlius, New York, USA Data Phone : 1-315-682-1824 Voice Phone: 1-315-682-9411 (14:00-23:00 *EST/EDT*) FidoNet : (1:260/375) StormNet : (181:181/1) InterNet : ajurison@terminus.storm.net WWW : http://www.storm.net/=ajurison/ Speed(s) : 33,600 VFC/V34/V34+ BBS : *A*R*T*H*U*R* BBS Position : StormNet Nodelist Coordinator / Detroit Area NC SysOp : Victor Capton Location : Troy, Michigan, USA Data Phone : 1-810-740-8764 FidoNet : (1:120/120) StormNet : (181:181/2) InterNet : Victor.Capton@inc.storm.net FidoNews 12-25 Page: 24 19 Jun 1995 Speed(s) : 33,600 VFC/V34/V34+ BBS : ASA CompuHelp Positions : StormNet United States ZC / Ohio Valley RC SysOp : Jeff Binkley Location : Gahana, Ohio, USA Data Phone : 1-614-476-3723 FidoNet : (1:226/600) StormNet : (181:181/0) InterNet : Jeff.Binkley@zc.usa.storm.net Speed(s) : 28,800 VFC/V34 BBS : New Era BBS Positions : StormNet Canada ZC SysOp : Mauro Incrocci Location : Kelowna, British Coloumbia, Canada Data Phone : 1-604-762-6239 FidoNet : (1:353/210) StormNet : (182:182/0) InterNet : Mauro.Incrocci@zc.can.storm.net Speed(s) : 28,800 VFC/V34 BBS : The Rising Sun BBS Positions : StormNet Europe ZC / North Central RC SysOp : Jens Fendler Location : Destedt, Germany Data Phone : +49-5306-7824 FidoNet : (2:241/510) StormNet : (183:183/0) InterNet : Jens.Fendler@zc.eur.storm.net Speed(s) : 28,800 VFC BBS : Closet Land ][ Positions : StormNet Asia ZC / Korea RC SysOp : Greg Wilburn Location : Munsan (Camp Howze), Korea Data Phone : +82-348-940-5677 FidoNet : (6:760/22) StormNet : (184:184/0) InterNet : Greg.Wilburn@zc.asia.storm.net Speed(s) : 14,400 v32b Thank you for your time! We hope to see you soon! ---------------------------------------------------------------------- ======================================================================== Fidonews Information ======================================================================== FidoNews 12-25 Page: 25 19 Jun 1995 ------- FIDONEWS MASTHEAD AND CONTACT INFORMATION ---------------- Editors: Donald Tees, Sylvia Maxwell Editors Emeritii: Thom Henderson, Dale Lovell, Vince Perriello, Tim Pozar Tom Jennings "FidoNews" BBS FidoNet 1:1/23 BBS +1-519-570-4176, 300/1200/2400/14400/V.32bis/HST(DS) more addresses: Don -- 1:221/192, don@exlibris.tdkcs.waterloo.on.ca Sylvia- 1:221/194, max@exlibris.tdkcs.waterloo.on.ca (Postal Service mailing address) FidoNews 128 Church St. Kitchener, Ontario Canada N2H 2S4 voice: (519) 570-3137 Fidonews is published weekly by and for the members of the FIDONET INTERNATIONAL AMATEUR ELECTRONIC MAIL system. It is a compilation of individual articles contributed by their authors or their authorized agents. The contribution of articles to this compilation does not diminish the rights of the authors. Opinions expressed in these articles are those of the authors and not necessarily those of FidoNews. Authors retain copyright on individual works; otherwise FidoNews is Copyright 1995 Donald Tees. All rights reserved. Duplication and/or distribution permitted for noncommercial purposes only. For use in other circumstances, please contact the original authors, or the eds. OBTAINING COPIES: The most recent issue of FidoNews in electronic form may be obtained from the FidoNews BBS via manual download or Wazoo FileRequest, or from various sites in the FidoNet and Internet. PRINTED COPIES may be obtained by sending SASE to the above paper-mail address. INTERNET USERS: FidoNews is available via FTP from ftp.fidonet.org, in directory ~ftp/pub/fidonet/fidonews. Anyone interested in getting a copy of the INTERNET GATEWAY FAQ may freq GISFAQ.ZIP from 1:133/411.0, or send an internet message to fidofaq@gisatl.fidonet.org. No message or text or subject is necessary. The address is a keyword that will trigger the automated response. People wishing to send inquiries directly to David Deitch should now mail to fidonet@gisatl.fidonet.org rather than the previously listed address. SUBMISSIONS: You are encouraged to submit articles for publication in FidoNews 12-25 Page: 26 19 Jun 1995 FidoNews. Article submission requirements are contained in the file ARTSPEC.DOC, available from the FidoNews BBS, or Wazoo filerequestable from 1:1/23 as file "ARTSPEC.DOC". Please read it. "Fido", "FidoNet" and the dog-with-diskette are U.S. registered trademarks of Tom Jennings, and are used with permission. "the pulse of the cursor is the heartbeat of fidonet"... -- END ----------------------------------------------------------------------