Volume 5 Issue #2 July 1999

Packet Radio Seeks Inside the Quarterly Niche Market Packet Radio Seeks Niche Market ...... 1 NEDA Is On A Membership Drive! ...... 1 Packet radio is great for connecting one ham with Editor's Column ...... 2 another, in real time. For situations where the ham 1999 NEDA Officers and Appointees ...... 2 needs to send another ham a pretyped text message Tech Committee Meeting ...... 3 during emergency situations, packet radio networking Board of Directors Meeting ...... 3 is unbeatable. For situations where a ham needs to send NEDA Compliant Node A Definition ...... 4 a short message to another ham at great distance, packet DAMA ...... 5 radio networking is unbeatable. NOS Info ...... 6 But.. what is packet radio most known for? BBS op- FlexNet STAT Command ...... 6 eration? Dx Spotting? Both of these operations can prob- Frequency Coordination ...... 6 ably be done better on the internet. It’s still fun to do FlexNet standards suggestion...... 7 these on ham radio, because it’s ham radio, but we’re For The Node Op ...... 8 not going to see packet radio to the great unwashed us- Looking At FlexNet For The First Time ...... 8 ing ham radio BBSs as an example of excellence! Usefulness Of Packet Radio ...... 9 Continued on page 63 Setting Your TNC's Audio Drive Level ...... 10 Map Report ...... 51 Can you help with Maps? ...... 51 North East Flea Market Dates ...... 53 NEDA Is On A NEDA Constitution ...... 64 Membership Roster...... 68 Membership Drive! Membership Application...... 70 NEDA TheNET X1J Node Spec ...... 71 Our club, for a long time conservative in our mem- Notes Regarding TheNET Parameter Spec ...... 72 bership goals, has decided that with more members we Board Meeting Notice ...... 73 can provide better services. Our club’s goals are to put printed technical information in the hands of hams who can build a general purpose packet radio network. We NEDA Quarterly, Volume 5 Number 2 have, for 10 years, been the best source of information Copyright © 1999 NEDA. Permission to copy in full on TheNET, practically writing the book on the subject or part is granted, if NEDA is credited and NEDA of implementing multi-port TheNET nodes, G8BPQ Quarterly v5.2 is given as the source. based BBS systems, and backbones. Published by Members of NEDA are recently learning about and The North East Digital Association installing FlexNet. Our last three newsletter issues of PO Box 563, Manchester NH 03105 been concerned with this topic. We’ve found no reason Please do what you can to promote our club. We can do better to stop. Next year we’re planning another copy of the looking magazines if we get our numbers up! Also send ques- famous NEDA Annual. This could become a textbook tions, article ideas to the editor or club postal address. We on FlexNet. It’s going to be a fun year! can use anything related to packet radio networking, VHF NEDA station installation, antenna reviews, site photos, etc... N.E.D.A. Quarterly v5.2 page 1 Editor’s Column from page 1 1999 NEDA Officers and Appointees Editor's Column As of May 1999 Board of Directors: bother bother bother bother bother bother bother ** Jim Wzorek K1MEA @K1MEA.ma bother bother bother bother bother bother bother bother ** Ray Feeley K1CSB @K1MEA.ma bother bother bother bother bother bother bother bother **Tadd Torborg KA2DEW @WB1DSW.nh bother bother bother bother bother bother bother bother ++John Driscoll N2MKH @N2UBH.ny bother bother bother bother bother bother bother bother ++Bob Seger WB2QBQ bother ++Dana Jonas WA2WNI ** Term expires first meeting 2000 bother bother bother bother bother bother bother ++ Term expires first meeting 2001 bother bother bother bother bother bother bother bother Board Member Alternates: bother bother bother bother bother bother bother bother Alternate for Ray, K1CSB Joel, N1JEO bother bother bother bother bother bother bother bother Alternate for Dana, WA2WNI Don, N2IRZ bother bother bother bother bother bother bother bother Alternate for Bob, WB2QBQ Bob, WB2DWD bother Alternate for Jim, K1MEA Burt, VE2BMQ Alternate for Tadd, KA2DEW John, WB2CIK bother bother bother bother bother bother bother Alternate for John, N2MKH not assigned bother bother bother bother bother bother bother bother Appointees: bother bother bother bother bother bother bother bother Treasurer: Bob Seger WB2QBQ bother bother bother bother bother bother bother bother Documents: Tadd Torborg KA2DEW bother bother bother bother bother bother bother bother Membership: Bob Seastream WB2DWD bother Sr. Admin. Asst.: Leo-Paul Chauvin KA1QP Archives: Don Rotolo N2IRZ bother bother bother bother bother bother bother Report Editor: Tadd Torborg KA2DEW bother bother bother bother bother bother bother bother Recording Sec'ys: Dana Jonas WA2WNI bother bother bother bother bother bother bother bother Map Coordinator: John Kushneir N2UBH bother bother bother bother bother bother bother bother Regional Mapmakers: bother bother bother bother bother bother bother bother Cape Cod Carl Black W3KI bother Central NY John Kushneir N2UBH bother bother bother bother bother bother bother Maine Jim Ledger N1PGH bother bother bother bother bother bother bother bother Technical Committee: bother bother bother bother bother bother bother bother Chairman: Don Rotolo N2IRZ bother bother bother bother bother bother bother bother Vice Chair: open - bother bother bother bother bother bother bother bother BBS Committee: bother Chairman Jim Wzorek K1MEA TCP/IP Committee: bother bother bother bother bother bother bother Chairman Pete Butler W1UU bother bother bother bother bother bother bother bother Network Volunteer Regional Contacts: bother bother bother bother bother bother bother bother SE Ontario Eric Meth VE3EI @ VE3NUU bother bother bother bother bother bother bother bother Montreal area Burt Lang VE2BMQ@VE2FKB bother bother bother bother bother bother bother bother Northern NY Roger Osterohout KA2JXI@KA2JXI bother Rochester NY area Mark Oliver NM2J@WB2VPH Central NY John Kushneir N2UBH@N2UBH Northern Tier PA Qualified Volunteer NEEDED Southern Tier NY Chris KB2FAF Eastern NY Dana Jonas WA2WNI @WA2WNI Western MA Jim Wzorek K1MEA @ K1MEA Eastern MA Peter Butler W1UU@W1UU Maine Mike Staines WA1PTC@WA1WOK NH & E. MA Cal Stiles W1JFP@W1JFP New Jersey/NYC Don Rotolo N2IRZ@WA2SNA Connecticut Joel Curneal N1JEO@N1JEO

N.E.D.A. Quarterly v5.2 page 2 Tech Committee Meeting Board of Directors Meeting May 2nd 1999 at 10:05AM. May 2nd 1999 at 1:45PM. The meeting was held at the Mercedes Benz Techni- The meeting was held at the Mercedes Benz Techni- cal Center in Montvale NJ. cal Center in Montvale NJ. Attendance was taken Attendance was taken Bob K2BJG Tadd KA2DEW nbod Burt VE2BMQ Dana WA2WNI nbod John WB2CIK Ray K1CSB nbod Ray K1CSB Jim K1MEA nbod Tadd KA2DEW John N2MKH nbod Don N2IRZ Burt VE2BMQ Bob WA3LWR Don N2IRZ Jim K1MEA John WB2CIK John N2MKH Warren WB2ONA Alex N3NMN Bob K2BJG John N2NSA John N2NSA Warren WB2ONA Don passed around the agenda and chaired the meet- Bill NX2P ing Dana WA2WNI Dana, Jim and Ray arrived just after the meeting Committee Reports started and were in radio contact as the meeting was called to order. Treasurer Don read the agenda. These were the items listed: Tadd read the report from QBQ. • What is a NEDA Compliant Node First Quarter 1/1/99 to 3/31/99 • Update on TCP/IP via FlexNet Starting Balance $2239.96 • Hidden Transmitter Syndrome and DAMA Income $1740.68 • Embedded PCs for network node sites Expenses $1118.96 • Netrix for 6pack Ending Balance $2861.68 • Show and Tell - Laser Comm (ed: this includes $155.29 to cover the meeting at the • Operating Standards - BBSs and stuff Depot ) • Frequency Coordination Tadd said that we’d spent nearly $700 on the Quar- NEDA Compliant Node Definition terly that just came out, which was billed after the end At the previous meeting of the Technical Committee of the first Quarter. Tadd said that this Quarterly print- a question was asked, “Can you use FlexNet to build a ing included over a hundred extras to allow us to have NEDA compliant node?” The general consensus was yes. promotional copies. He expects that to help the mem- The obvious next questions are: 1. What other software bership grow. We have on the order of 160 members can be used to build a NEDA compliant node; and 2. and an annual income of something less than $3000. If What is a NEDA compliant node. we bill at the same rate over the next year Tadd said, we’d still run out of money. (ed: at this point I think A committee of 2 was formed, consisting of Bob, that was incorrect). Tadd said that we won’t be printing K2BJG, and Don, N2IRZ, to set about finding out the answer to the 2nd question. The 1st question was ex- Board Meeting Minutes - Continued on page 54 pected to be answered as a follow-on. Tadd added that 2 pages of each issue of the newslet- ter were dedicated to a totally redundant presentation of TheNet parameters. He suggested that it might be a good idea to include such static information on getting Bob Seger - WB2QBQ is started with FlexNet in addition to, or instead of, NEDA’s treasurer. Bob TheNet. has been an officer of the club for half a decade Don said that the committee had started out trying to and was sysop of the define what it meant to be following NEDA guidelines KNOX node, living just Tech Committee - Continued on page 16 west of Albany, NY.

N.E.D.A. Quarterly v5.2 page 3 NEDA Compliant Node From: “Papson, John” To: “‘Tadd Torborg, KA2DEW’” A Definition Editors note: This article was published on NEDA’s Subject: FYI: KISS vs SMACK web page several years ago. Burt - VE2BMQ - NEDA’s webmaster - reserected it as a compliment to the discus- Date: Tue, 11 May 1999 12:38:39 -0400 sion in the Technical Committee meeting minutes.

A NEDA Compliant Node is one of those nebulous X-Mailer: Internet Mail Service (5.5.2232.9) terms that is difficult to nail down. In the past couple of technical committee sessions, we have agreed that a NEDA Compliant Node has the following features:

It is a node connected to the network using Amateur SMACK - Protocol Description Radio Hidden Transmitter Syndrome Free (HTSF) links. A very controversal issue are links using non- amateur —————————————— radio methods. It is agreed that these should be consid- ered non-compliant and configured as "gateways". Ama- teur packet is a branch of the amateur radio hobby and Version 1.0, Dated 27.02.92 we should strive for independance from non-radio solu- tions. by Jan Schiefer, DL5UE and Dieter Deyke, DK5SG/ The node operating parameters are open for inspec- N0PRA tion wherever possible. For those node systems, (such as G8BPQ switches and NOS systems,) where the pa- rameters are not normally available for inspection, a copy would be made available as a file on a publicly accessable source (BBS) and the location of the files in- 1. Introduction dicated with an info message. The node would allow all users equal access to avail- The end of 1990 was the first time that the able network resources. High volume data sources such Stuttgart as BBSs, dxClusters, internet gateways and other serv- Packet-Radio-Amateurs ers are expected to access the network with their own dedicated links but should not hog the network resources thought about the factual certainty of Data Secu- to the detriment of ordinary users. High volume users rity between TNC and would be expected to help build the network and increase WAMPES(*)- Node Computer. Already with other the network capacity. Packet-Node Systems Data There are other node characteristics that distinguish losses a NEDA Compliant Node, but so far we have not been able to agree on clear defining language to describe them. had appeared, we considered which sort and in –Archives what manner a compatible

development of the KISS-Protocols with a Checksum could be considered possible. From the Result of this consideration came the name SMACK (Stuttgart Modified Amateur radio CRC-KISS. This Document should explain the difference between SMACK and KISS and make possible the implementation N.E.D.A. Quarterly v5.2 page 4

of other Systems. DAMA PC/FlexNet V3.3g Welcome to the WA2SNA Flexnet-Digi FlexNet support of DAMA at Ramapo Mountain, Oakland, Bergen County, NJ [FN21ub] Type A for info on Destination table calls. 1.2. DAMA-Master Type H for Command Summary and Command Explanations in The node now is capable of the DAMA protocol, cur- detail. rently only simplex. On synchronized channels, the Type I for Additional Information DAMA mode is synchron, too. The transmission time Type M to connect to BBSNEJ:WA2SNA-4, Packet Bulletin Board per QSO is currently limited to about. 4 seconds; pa- Type C WA2SNA-1 to connect to IPNEJ:WA2SNA-1, TCP/IP rameters cannot be specified currently. DAMA mode may Gateway be activated by an “m” in the MODE-Command. It is possible to have several independent DAMA masters at =>st the same time, e.g. for user ports on different frequen- cies. On multimode ports, the combination “sm” has to be used (not on RMNC fullmaster). 1.3. DAMA-Slave When a channel hears another DAMA master, e.g. when using the system as user frontend, this channel is set to DAMA slave mode automatically. For network nodes this feature can be turned off.

Kantronics Supports DAMA Ms. Mary Steidler, Amateur Sales, Kantronics wrote to tell me that any Kantronics unit with firmware 7.0 or higher has support for DAMA slave mode. You could get an 8.2 upgrade for the KPC-3, KPC-3*plus*, KPC-9612, KPC-9612 Plus, and KAM Plus. The KPC-2 and KPC-4 can no longer be upgraded. All of Kantronics’ current amateur TNCs (KAM Plus, KAM ’98, KPC-3 Plus, KPC-9612 Plus) support the DAMA protocol (slave mode). Our customers rarely ask us about this, so I’m glad you’re curious. Thanks for writing.

N.E.D.A. Quarterly v5.2 page 5 Tadd:

More good news:

Bob K2BJG now has DX Cluster running on top of FlexNet, using the TFEMU NOS Info interface, at N2BIM in Western NJ.. As you will read in the minutes, the board of directors had a discussion regarding methods for doing TCP/IP FlexNet simply replaced G8BPQ for the networking across FlexNet. The WAMPES driver is a module that program. may be incorporated into NOS to make the NOS station into a part of the FlexNet network. For internet users, Les Kalmus, W2LK, may also have had it running here is a URL for the WAMPES Reference Manual: experimentally at times http://www.ife.ee.ethz.ch/~sailer/ham/wampes.html#H1. earlier from Manhattan NY..

WAMPES is used in two versions of NOS. They are: 73

WNOS John WB2CIK ftp://ftp.ucsd.edu/hamradio/packet/tcpip/wnos/ Frequency Coordination HRL-Mini-NOS Frequency Coordination has a few purposes. These (NOS w/ WAMPES module, and w/o NETROM mod- are: ule): • Register a single fixed station user on a frequency ftp://ftp.ucsd.edu/hamradio/packet/tcpip/hrlnos/ for a given piece of the Earth; —WB2CIK • Resolve conflicts between stations attempting to use FlexNet STAT Command the same frequencies in any overlapping space. The FlexNet Infobox command article in the last is- • Document use of the Amateur Radio bands so those sue left out the STAT command. This is the results from bands may be best protected by our ARRL a stats command. representitives to the US Government agencies that uptime: 38d, 3h; total 200k, max 44k, used 10k give or take radio spectrum; po device version txframes rxframes terr rerr rberr ioerr 0 6PACK 0.9d 189604 199016 0 0 0 4 • Create lists of public frequencies that should be listed 1 6PACK 0.9d 514321 813163 0 0 0 143 in guide books for public consumption (like repeater 2 6PACK 0.9d 758436 686103 0 0 0 19 directories). 3 6PACK 0.9d 1231066 1985558 1 0 0 26 4 6PACK 0.9d 1047495 890890 1 0 0 17 What you can expect 5 KISS 1.5a 818463 828987 0 0 0 0 How your local council deals with you and how help- 6 6PACK 0.9d 719076 791156 0 0 0 0 ful they are is often dependant on what you offer or ask 7 6PACK 0.9d 742910 631420 0 0 0 2 of them. 15 SHELL 3.3g:2 0 0 0 0 0 0 –K2BJG In practice I think what you can expect is that if you have a plan in mind that is rational and legal that you will get How long will it take them to get back to you? What should you do in the mean time? I think the biggest problem the coordinators have is with spectrum that is offered and then never used. The only reason I can think of that would make some- one using an amateur frequency on a long term basis to not want to register with a coordinator is one of trust. Since the coordinators are volunteer Amateur Radio lic- ensees, it is quite possible that they will be driven hon- est ethical people. It is also possible that they will not be. The easy way to assure they will be, is to volunteer a stint in the local council. So, given no rumors to the affect that the local council is dishonest, it behooves us to register our user and back-

N.E.D.A. Quarterly v5.2 page 6 FlexNet standards suggestion I have a suggestion. Every FlexNet node should have vantage of this information to automatically plot the the following as a connect text: nodes on the map. The information up to the text “Flex- Net Node” will be incorporated into the map exactly as gridsquare:callsign:location:FlexNet Node:New users it appears on the node’s connect text. type I for more information This system also leaves a much shorter connect text For instance for any ham that is exploring the network. Network FN32ft:KA2DEW:Amherst New Hampshire:FlexNet Node:New exploration is the one thing we hams can do on the ham users type I for more information network that we cannot do on the internet network. Let’s What this does is makes it possible to do automatic consider this. mapping of the network. Shortly we’ll have a system –KA2DEW- NEDA editor for creating maps for this newsletter that can take ad-

N.E.D.A. Quarterly v5.2 page 7 For The Node Op Looking At FlexNet For The First Time Question: route tables, and etc.. The destination list (nodes list) The ham who runs the node just west of me is con- includes all of the nodes within a certain amount of time, verting to FlexNet. He wants me to turn digipeating on and shows those stations along with the estimated time in the TheNET TNC that is my backbone port facing to and from the node. The latency on each backbone is him. Is this reasonable? Tell me about FlexNet. Is it a measured constantly (unless there is no activity in which good idea to be replacing a TheNET with FlexNet? What case it is only measured every few minutes). Backbone are the features and problems of doing this? latency is available with the “P” command. You can connect to stations in the TheNET fashion, to Answer: the node, to the destination node, to the user, or you can There is much to say about FlexNet. I am enthused issue a digipeat connect. ROSE style, as C dest v about it. What they are asking you is standard fair for node1,node2 adjacent FlexNet node ops. In order for a FlexNet node to make a connect to a non-FlexNet station, user sta- There are four problems with PCs in FlexNet node tion included, the station has to allow a digipeat off the sites. Some of these are crucial to a node op, some are station. In addition, two FlexNet nodes may use an in- not. termediate TheNet node as a stupid digipeater (or pair These are: of them) in order to pass FlexNet information across the 1 PCs are large TheNet site. This is only important if you are surrounded 2 PCs have moving parts by FlexNets. 3 PCs are expensive The basics of FlexNet-PC: 4 PCs are power hogs (65 watts typical) The solutions to these problems depend on which com- It runs on a PC much the same way G8BPQ does (as bination needs to be fixed. a TSR). It uses TNC2s as slaves. The software that runs in the TNC2 is called 6pack and fits in a small • We can get free 386 PCs with HDs. EPROM. It gives the PC total control over the TNC • We can get solid state hard disks that are flash (not like KISS) and also allows the PC to control mul- ROM for under $100. tiple TNCs (up to 8) over one RS232 COM port (the TNCs • We can get small PCs that have imbedded Flash HDs are daisy chained, TX -> RX:Tx -> RX ...). The only and RAM for under $300. These are 300mA @ 5v limitation on the 6pack TNC loop is that the RS232 baud and are very small. rate (same on all TNCs on the loop) has to be twice the We haven’t found anything that solves all four prob- total RF baud rates. lems at once. The PC side has the smarts. The FlexNet node does The FlexNet node needs something that looks like a everything TheNet does in terms of having a command disk drive for temporary data store. It saves all sorts of prompt and access to routes and whatnot. It measures information dozens of times per minute. Further, when the performance of user and backbone ports and auto- you make changes to the configuration at setup time, or matically handles frames, backoff, and whatnot to bal- on the fly, it saves the route entries and whatnot on the ance latency and throughput while maximising channel disk drive. You can’t use a Flash disk for data store utilization. because in much less than a year you exceed the manufacturar’s specifications for number of writes. So... The FlexNet command prompt (the “infobox”) gives we use a RAM disk for storing parameters and then use access to routing information and statistics, as well as a DOS application to copy the data from the RAM disk letting the user connect to other stations, enter chat/ to the flash ROM.

N.E.D.A. Quarterly v5.2 page 8 Usefulness Of Packet Radio This article attempts to address these issues: that if an appliation needs autoforwarding, it shouldn’t • What good is Amateur Radio packet? be on ham radio. • How do we best promote it? Real time data gathering, i.e. weather and proximity The Internet has knocked the bottom out of packet as information. One thing APRS is doing that is wierd is regards the general ham population. People who, ten doing general broadcasts, overflowing regions to hit the years ago, would have set up a little station and used random targets. How about if a station could connect the BBSes, sent email, etc., are now on the ‘net which is across the network, in several directions at once, to look faster, offers many more abilities than packet, requires at the APRS frequency from various different cities. In- *less* equipment than packet and is practically free. stead of looking just at the local picture. Sort of like APRS HF/VHF gateways, this would be APRS network/ So, packet today, after the internet bomb was dropped VHF gateways. on it, is a shadow of its former self. To most hams it looks like packet radio is useless and is now forgotten. Weather could be handled in exactly the same way. This needn’t be a forgone conclusion. It may be possible The trick is creating home software to handle the in- to continue in our networking hobby even now. It is coming information and to establish which network sites possible that packet radio is worth building onto. you want your home station connected to. What good is packet radio? Additionally it would be worth having data concen- There are three obvious features of Amateur Radio trators so you could gather all of the NY state weather that give reason to believe that packet radio will sur- and APRS sites with one connect to your home station. vive the overwhelming superiority of the internet over . Just a thought. the Amateur Radio network we created in the 1990s. There are SOOOO many things we could do with a 1. Some consider it fun to use hobby radio to perform a working reliable real time packet network, that won’t task that may be done more easily using other means. work unless you have full time internet to everybody’s 2. Amateur Radio is great for emergency preparedness home. 3. Amateur radio can be cheaper to use than commercial By the way, if the bottom drops out of something, what means for some applications. Sometimes the cost does that leave you with? differential is the difference between not being affordable, and being almost free. Fun to use Usefulness of Packet There are some conclusions you didn’t draw. The Packet then, has two choices - packetcluster thing is not packet-only. It works on the internet, possibly better than on packet. To evolve into a strictly techie whizz-bang domain with limited general ham appeal (a place where advanced The reason people use packet for PacketCluster is that techniques and hams live, not Joe Putz and his DRSI packet is 24 hours a day without tying up a phone line. card!) That conclusion is the fabric on which I base a hope that general hams could use packet (your point #2). I agree or with the rest of what you said. To attempt to become a medium with an appeal to the We’ve got a few niches to go for however. general ham population once again (as it was before the advent of the internet!) Emergency Traffic handling. If people applied packet correctly for this, it is unbeatable. Previously traffic If the first course is followed, we must accept that handling over packet has been based on autoforwarding packet is forever going to be a niche domain in our hobby. BBSs. Not the way to do it, in my opinion. Traffic han- Kind of like ATV is. A place where the newest techniques dling needs to be person to person, just like it is on HF are tested and used but it’s mainly done to satisfy the and on VHF voice nets. The only difference is that you tinkering Jones that most of us have. We’ll have to ac- can do the net over a packet network, coordinating the cept the continuing attrition of the packet population, system via a CROWD node or something like it. death of BBSes, etc. and understand that packet has become a strictly experimental mode with little use to You still need an acknowledgement that a PERSON the ham population in general. got your message to the next location. Autoforwarding on packet radio is for the birds in my opinion. I think

Continued on page XX?? N.E.D.A. Quarterly v5.2 page 9 Setting Your TNC's Audio Drive Level Why it's important, and how to do it... data; just under two cycles of 2200 Hz represents one by John Ackermann N8UR bit). Note that both tones are at the same amplitude. I'd lay reasonable odds that most 1200 baud packet The receiver at the other end of the link recovers these radio stations are getting less range than they should, tones, and sends them via its speaker jack to the TNC, and generating more retries than they need to. where a demodulator (the "dem" part of a modem) con- verts the tones back to binary form for further process- Why? ing. For best performance, the tones fed into the de- Because their TNCs are overdriving the microphone modulator should be at equal amplitudes, or biased a input on their radios. The audio output control on many bit toward the high tone. If there's a lot of imbalance TNCs is very touchy, and the difference between proper (or "twist") in the tone levels, demodulation performance audio and too much may be only a few degrees on the is likely to suffer. control. If you've just plugged your TNC into your radio More important, if the low tone coming out of the and gone on the air without adjusting the level, the odds speaker is stronger than the high tone, many TNCs based are very good that you are transmitting a distorted sig- on the TAPR TNC-2 design, including the very popular nal, and your packet radio performance is suffering as MFJ-1270 series, will have great difficulty demodulat- a result. ing packets. The demodulator chip in these TNCs, the The purpose of this page is to show why this is a prob- XR-2211, is particularly unhappy if it sees a low tone lem, and how you can fix it -- whether or not you have a that is louder than the high tone. Because XR-2211 de- bunch of fancy test equipment. modulators are so common, it's very important that we transmit signals that these units can decode, and that If you're anxious to learn how to adjust your TNC means signals that do not have a a twist favoring the without worrying about why you need to, you can go low tone. directly to Fixing the Problem, but I don't recommend it. The key message underlying this page is that packet tones should come out of the receiver speaker at equal Understanding the Problem -- amplitudes, or with the high tone a bit louder than the The Packet Radio Audio Path low one. Under no circumstances should the low frequency tone be louder than the high tone. A 1200 baud TNCs generate two tones, usually 1200 and properly adjusted packet station meets this requirement. 2200 Hz, to represent the one's and zero's of the binary data they are fed by their host computer. Those tones Here's an oscilloscope picture of packet tones at the are fed to the input of an FM transmitter, which trans- speaker jack of a receiver. The TNC and transmitter mits them just as if they were voice. This form of modu- generating these tones were adjusted to provide proper lation is called Audio Frequency Shift Keying, or deviation (more on that later), and the tones at the "AFSK". transmitter's input were of equal amplitude. Here's what an oscilloscope display of the audio com- ing out of a TNC looks like. The closely spaced waves are cycles of the 2200 Hz high tone, and the more widely spaced waves are the 1200 Hz low tone (by the way, each cycle of 1200 Hz represents one bit of 1200 baud

In a perfect world, the two received tones would also be at the same amplitude. Here, you can see that the high tone is somewhat louder than the low one; the transmitter I used for this test boosts the high frequen- cies a bit more than it should. Within reason, that's not a problem; a twist in the opposite direction, though, is a

N.E.D.A. Quarterly v5.2 page 10 different matter. By the way, you can click here for de- To get around this problem, someone had the bright tails of the test setup I used for these pictures. idea of boosting the high frequencies in the transmitted audio (like turning up the treble control on a stereo), so Here's an oscilloscope picture of the tones that come that they are louder than they should be, and then cut- out of the speaker when the TNC is over-driving the ting the highs back by the same amount at the receiver transmitter. -- in effect, turning the treble control down. That brings the voice signal back to its proper tonal range, but also knocks down any noise that was introduced along the way. The end result is that voices sound normal, but hiss is reduced -- the signal-to-noise ratio is improved. It turns out that this works really well for voice signals, and it's standard practice in all FM voice transmitters today. (And, by the way, this is how Dolby noise reduction re- moves hiss from cassette tapes.) The boost that's given at the transmitter is called preemphasis; it's a high-pass filter that emphasizes the high frequencies in the audio spectrum (hence the You can see that now, the lower tone is much louder name). Preemphasis increases the level of high frequen- than the high tone. That's the opposite of what we want, cies by 6dB per octave -- in other words, the voltage fed and a TNC will have a lot of trouble decoding this packet. to the modulator for a tone of 2000 Hz is twice as high Unfortunately, if you've never paid attention to the trans- as the voltage for a tone of 1000 Hz. Preemphasis should mit audio level setting on your TNC, there's a good cause the 2200Hz high packet tone to be about 5dB chance that this is what signals from your station look louder than the 1200 Hz tone. (Note: this discussion is like. about voltage and not power ratios, so 6dB represents a doubling, not 3dB.) Why does this happen? Here's what a properly deviated packet radio signal, The FM radios we use were designed to transmit the with preemphasis applied, looks like coming out of the human voice, not digital tones, and there are a couple of transmitter. circuits in these radios designed to improve voice trans- mission. Unfortunately, these circuits do very bad things to packet tones if the audio input level is too high. The first circuit (actually, two circuits -- one in the transmiter and a complementary one in the receiver) is made up of the preemphasis and deemphasis net- works, and the second is the transmitter's audio clip- per. Each serves an important purpose in voice com- munications, but when combined with excessive trans- mit audio levels from the TNC, they can distort packet data beyond recognition. Preemphasis and Deemphasis The high tone has nearly twice the amplitude of the Weak FM signals have an annoying hiss that reduces low tone; that makes sense since the preemphasis boost intelligibility. Preemphasis and deemphasis are a tool is 6dB/octave -- 2200Hz is nearly double 1200Hz. to reduce that noise. (There's a little bit of distortion visible in the low tones The problem is that most of the audio noise in a radio in this picture; I'm not sure where that is coming from circuit is at high frequencies, not low ones. Unfortu- -- it may be an artifact of the measurement setup.) nately, so is most of the information in your voice (low The complementary circuit in the receiver provides frequencies make voice sound good, but high frequen- deemphasis of 6dB per octave, but in the opposite di- cies make it intelligible). And to make things worse, rection -- the receiver circuit acts as a low-pass, instead most of the power in our voices is in the lower frequen- of a high-pass, filter. The combination of the cies. As a result, it's hard to understand someone whose transmitter's preemphasis and the receiver's deemphasis voice is competing with high-frequency noise or hiss. Continued on next page

N.E.D.A. Quarterly v5.2 page 11 Audio Drive from previous page Emphasis Plus Clipping Equals Trouble results in the received signal matching what was at the microphone input, with the benefit of reduced hiss. If the audio level the TNC feeds to the microphone jack is too high, the preemphasis circuit will still work This is what a properly deviated signal looks like at normally, and boost the level of the high frequency tone the receiver's speaker terminals; the deemphasis circuit by 6dB/octave -- making our 2200 Hz tone nearly twice pretty much (though not exactly) cancels out the boost as strong as the 1200 Hz one. However, the clipper will that preemphasis gives to the high tone. knock both tones down to the maximum allowed volt- age. The result: the level difference between the two tones will decrease, and if the overdrive is great enough, will disappear entirely. In that case, the 1200 and 2200 Hz tones will be transmitted with the same deviation level. In addition, because the clipping process flattens off the top and bottom of the input wave, it adds har- monic distortion to the transmitted signal. When the receiver picks up a signal that's clipped enough to equalize the deviation of the two tones, it will still deemphasize them, dropping the high tone by almost half compared to the low one. What was a good idea is now a bad one; due to Audio Clipping clipping, the deemphasis circuit is no longer acting on a preemphasized signal. Instead of seeing two tones By themselves, pre- and de- emphasis don't cause prob- at the same level, the TNC's demodulator will see a lems for our AFSK tones. However, when they interact 2200 Hz tone that's significantly weaker than the with another circuit that's included in communications 1200 Hz tone. And, of course, that's exactly what the transmitters, trouble lies ahead. TNC doesn't want to see. That circuit is the audio clipper. It's a handful of com- Here is the signal an over-driven transmitter puts out ponents designed to make sure that no matter how loud after the tones have been subjected to preemphasis and the audio input signal, the transmitter doesn't deviate clipping. more than a set amount (usually 5kHz for 2-way ra- dios). If the audio input is below the clipping point, the sig- nal passes through the clipper untouched. If it is above that point, the signal is chopped off (or clipped) at the magic level. This makes sure that voice peaks don't cause overdeviation. Because the clipper is a guardian that makes the trans- mitter stay within its legal deviation limits, it is usu- ally the last part of the audio chain, right before the modulator. Most important, it follows the preemphasis circuit. That's worth repeating -- the clipper works on preemphasized audio. The clipper has chopped both the high and low tones at the same level; there's no trace of preemphasis left. Why is it worth repeating? Because if the clipper op- In fact, the low tone is actually a little bit louder than erates on a preemphasized signal, the high tones are the high tone -- that's probably caused by the other dis- more likely to be clipped than the low ones. And tortion that "hard" clipping introduces. that, my friends, is why this discussion is relevant to packet radio.

N.E.D.A. Quarterly v5.2 page 12 You may think that this signal doesn't look too bad, The No-Test-Equipment Packet but remember that this picture is before the receiver's Adjustment System deemphasis acts on it. Here's what the signal looks like by the time it gets to the receiver's speaker. There's no excuse to transmit an over-driven packet signal. You can get your TNC's audio level set "close enough for government work" with nothing more than another receiver -- anything that will receive the trans- mitted frequency will do. Here's what you do: 1. Hook up the TNC to the radio, set the radio to low power and (ideally) connect the antenna jack to a dummy load. Tune the second receiver to the same frequency and remove the antenna; ideally, also connect its antenna jack to a dummy load (the idea is to keep the receiver This picture is guaranteed to give a TNC fits. Recall from being overloaded by the transmitter's signal). that the demodulator chip used in the most popular 3. Following the instructions for your TNC, put it into TNCs particularly hates this condition -- it wants the "Cal" mode and key the transmitter. Here are the high tone to be stronger than the low one. Even if the commands to do that on an MFJ-1270 TNC: signal is very strong, the TNC will have trouble decod- ing it. If the signal is weak, and there's additional noise a. At the command prompt, enter CAL. riding along with the tones, the chances of successfully b. Press K to key the transmitter. (The TNC's decoding a packet become very slim. timeout timer may limit the keyup time to ten seconds or so; if the transmitter cuts out before That's why setting your TNC's audio level properly is you're finished, unkey and then key again. important. You can more than double the effective range c. Press the space bar to toggle between tones of your station by sending packets that are easy for other until you hear the high tone in the receiver's stations to decode. speaker. d. When finished, press K again to unkey the Fixing the Problem transmitter, and Q to quit the calibration Now that we know what the problem is, how do we fix routine. it? I'll describe three different methods, starting with a 4. Make sure that the TNC is sending the high tone. quick-and-dirty approach that will get you into the ball park even if you have no test equipment available. 5. Increase the TNC's TX Audio potentiometer (the instruction manual will tell you where it is) until The main goal is to get the audio level low enough the tone you hear in the receiver doesn't get any that the preemphasized high tone gets through the trans- louder -- which indicates that the transmitter is mitter without being clipped. Then, we'd like to opti- clipping -- and then bring it back down again until mize the signal a bit further to get the best performance you hear the tone get noticeably softer. we can. 6. Turn the pot down a little bit more for good measure, Before we get started, here's one tip that may make and then toggle the tones to make sure that both are the adjustments easier. You might find that there's still coming through. It's better to have the deviation only a tiny bit of difference in the audio level setting a little too low, than too high. between no tone at all, and full clipping. That was the case with the TNC/radio combination I used for these That's all there is to it. You've now set your TNC so tests. You can improve that situation by putting a that both tones will pass through the transmitter with- resistor in series with the audio lead to the micro- out clipping, and with preemphasis intact. You are trans- phone. Finding the best value may take some experi- mitting a signal that's pretty close to what it should be. mentation, but something in the range of 4.7kohm to 47kohm will probably let you get the level setting control up into a range where adjustment is much easier.

Continued on next page N.E.D.A. Quarterly v5.2 page 13 Audio Drive from previous page Everyone has their own pet value for deviation, but most of the recommendations I've seen are to set the The Homebrewer's Method high tone to around 3kHz deviation, and let the low The steps above will get you in the ballpark, but if tone fall where it may. The range of settings I've seen you're interested in knowing your signal's deviation with published goes from 2.8kHz to 3.5kHz. I usually aim a bit more precision, there's a simple way to do it that for 3.25kHz or so myself, but anything in this range is requires nothing more than some ingenuity, the will- close to optimum. ingness to open up a receiver and stick a wire in its guts, If the transmitter's preemphasis is working cor- and an oscilloscope. The receiver can be a cheap scan- rectly, 3kHz deviation of the high tone will yield low ner, and an old 'scope will work fine -- all it needs to do tone deviation of about 1.65kHz -- since the is display audio frequencies. preemphasis is 6dB (double the amplitude) per octave The end result will be a setup that can measure the (double the frequency), the low tone will be deviated deviation of an FM transmitter with surprising accu- 1200/2200, or 0.545, as much as the high tone. racy, and will let you adjust the audio level of your TNC Building the Deviation Scope to the deviation level you want. It can also be used to measure the relative frequency error of a transmitter, Since this is a homebrewer's project, I won't provide and to set deviation levels for voice radios. detailed instructions, but this should get you going. 1. Find the discriminator output in the receiver, and But What is Deviation, Anyway? attach one end of a 47kohm to 100kohm resistor to Before designing our deviation meter, we'd better know this point -- you may have to experiment to find a what we're planning to measure. value that will provide enough signal without loading down the discriminator. Make sure that you tap into Deviation is the term used to describe modulation of the discriminator before any coupling capacitor -- you an FM signal. The audio fed into the microphone input need to have DC coupling to make this work. Attach causes the transmitter's frequency to shift up and down the other end of the resistor to a piece of shielded from the center carrier frequency; the louder the input, wire, and route it outside the receiver. Attach the the greater the shift. That instantaneous frequency shift shield to ground in the receiver. is the transmitter's deviation. Deviation is usually measured as the peak positive or negative shift from 2. Set the scope for DC coupling, and attach the probe the carrier (in other words, it's "center to peak", not to the wire. Turn the receiver on, and adjust the "peak to peak"). scope's vertical sensitivity until the noise just about fills the screen. The oscilloscope pictures shown on this page with the words "FM Demod" in the lower left corner are cali- 3. With the transmitter set to low power and feeding a brated to show 2kHz of deviation per vertical division dummy load, and the receiver tuned to the of the display. Because we're looking at an audio signal transmitter's frequency and also hooked to a dummy after it's been demodulated, the frequency shift around load, key the radio with no modulation applied. the carrier has been converted to amplitude on the 'scope -- the larger the displayed signal, the greater the 4. The trace should now be a single horizontal line. devation. The goal of this project is to display a similar Adjust the scope's vertical position control to center calibrated signal on your oscilloscope's display. the trace on the screen. The radios that we normally use for ham radio FM, 5. Either adjust the transmitter up 5kHz in frequency, and which we are likely to use for 1200 baud packet, are or the receiver down 5kHz (do one but not both!) and designed for a maximum deviation of 5kHz, and their transmit again. The horizontal trace on the scope clipper circuits usually kick in at that point or a little should now be displaced either up or down from the lower. center point. If you can, adjust the vertical sensitivity control so that the trace is exactly 5 divisions from What should the deviation of a packet signal be? We the center line. If you can't do that, just note as know that the high tone needs to be less than the clip- accurately as you can the number of divisions of ping value, so you might think that "louder is better" displacement, and in what direction. From this point and that we should run the deviation right up to the on, do not touch the vertical sensitivity control. point where clipping begins. That's not necessary; it makes your transmitted bandwidth wider than it needs to be, and it can actually reduce performance in some cases.

N.E.D.A. Quarterly v5.2 page 14 6. Shift the frequency of the transmitter or receiver found in flea markets). A few years ago, TAPR sold a 5kHz to the opposite side of the desired frequency, deviation meter kit that interfaced to a scanner and key the transmitter, and once again note the amount provided a digital display. If you can find one, it and direction of displacement. The amount should works very well. AEA and MFJ have also sold devia- be very similar to the result of the previous step, but tion meters at ham prices. And, older communication in the opposite direction; if it is not, the receiver is service monitors can sometimes be found for surpris- probably out of alignment and you should take care ingly low prices, but beware -- some of these units can of that before attempting to measure deviation. be frightfully expensive to fix if they're broken. 7. Divide the displacement you measured in each of the An Idea for the Adventurous previous two steps by five. The result is the divisions per kiloHertz of your measurement system. For Here's an idea you may want to consider if you're not example, if the 5kHz frequency change caused the afraid to take the cover off your packet radio transceiver. trace to move exactly five divisions, each division Consider some of the facts we've been dealing with: equals 1kHz of deviation. • The TNC demodulator doesn't mind if the high You've calibrated your system to measure deviation. tone is somewhat louder than the low tone. You can transmit cal tones from the TNC as described • The TNC hates it when the low tone is louder in the previous section, and measure the amplitude of than the high tone. the recovered audio. Applying the kHz/division value • A clipped packet signal doesn't look too bad you calculated above, you can easily determine the de- before deemphasis is applied; the two tones will viation of your transmitter, and adjust the TNC's audio be at about the same level; it's only after level to get the proper value. deemphasis that the twist is really bad. • You're likely to be copying a lot of clipped Since this system is DC coupled, it will also show the signals. frequency offset of the transmitter as compared to the receiver. This is another valuable piece of test informa- Putting all this together, it might make sense to feed tion, but it can make measuring deviation more diffi- un-deemphasized audio to your TNC. Properly adjusted cult. If you want to look only at deviation and not fre- signals will have too much high tone, but the TNC can quency offset, you can switch the scope input to AC cou- deal with that. Clipped signals will actually be pretty pling when you're making measurements. well balanced, and will decode much more successfully than if they had been deemphasized. The Expensive Way I've been doing this for years -- transmitting Earlier, I said I'd describe three ways to adjust your preemphasized audio, but driving the TNC with audio packet station's transmitted audio. With two down, the directly from the discriminator, before the deemphasis third method is very simple. If you have access to a circuits. It seems to work very well, and I recommend deviation meter or service monitor, you can use that. that you think about this, particularly for nodes sites, The procedure is the same as in the previous steps, ex- BBSs, or other stations that will be communicated with cept you won't need to go through the calibration steps. lots of different stations. Although these instruments may provide an analog I'll be honest, though, and say that I haven't done or digital meter that displays deviation directly, I enough bench testing to determine how well this works. think it's still very valuable to display the deviation on These tests are on my list of things to do, and will be a scope if one is available (most service monitors can reported here when they're finished. do this). You can learn a lot about the overall quality of the signal from the scope display. Conclusion You may be surprised to find that a deviation meter I hope this page has provided you with an understand- isn't out of your, or your club's price range. There ing of why proper packet station setup is important, and have been a number of units sold for the ham market the ability to adjust your station for best performance. for a couple of hundred dollars or less (including a As simple as the "quick and dirty" method of setup is, Heathkit unit that was very popular and can still be it's very surprising how few people have used it. Now, you don't have any excuse... Copyright 1999 by John Ackermann Originally published at http://www.febo.com Used with permission.

N.E.D.A. Quarterly v5.2 page 15 Tech Committee Minutes from page 3 requirement that the commands or responses must be reasonable. This got many yucks and snide remarks. in network node implementation including network ar- Tadd went on to suggest that the three points of com- chitecture considerations. Later they reviewed the mis- pliancy we had discussed so far would not negate the sion of the committee and decided that what we really acceptance of a node that accepted 8 bit data but only needed was a list of things a network node’s software delivered 4 bits intact. must be able to do (or must not be able to do) in order to Don said that people wouldn’t use it. be compliant. Tadd said that it is in our interest to think of all of the The committee now reports that we have established things that we need in network software and that we a list of points of compliance. Tadd suggested that be- should add to the list of requirements as we go along, fore the list was read out or distributed that it might be both now and in the future, in order to optimize our worth while to poll the group at the meeting to get their network. We need to get as many of the good ideas and ideas in the hopes that we might get some new ideas. answers on paper as possible. This would help to forstall Don started out by saying that the NEDA compliant bad decisions by our membership by giving them a piece node must be a building block of a general purpose packet of paper with the good questions to ask. network. A general purpose packet network is one that Tadd said that Don and he had spent a few emails offers fair access to divergent data types and data pur- and a couple of hours coming up with mean things that poses and serves the interest of good amateur practice. software developers could do to us to slip under the wire. Burt brought up that we did have a definition at one There are more and more networking strategies coming time. Don said that he’d not been able to find it. Burt out all the time. It would do us good to try to prepare a said that there is a web document that describes it. Burt list of features that we desire, as well as those that we said that a NEDA compliant node shared the resources require. on an equitable basis so no individual person or type of The group tossed around what the legal requirements traffic could dominate it. It was open to all for inspec- of a node are. The FCC required that packet need to use tion so any parameters that affected network transport AX.25 to pass traffic. There is no FCC rule that governs and routing was visible to the common user. what protocols ride above AX.25. For example, if when Don suggested that he would read the list of notes message comes into the network it is converted to en- rather than having us spend time on redundancy. tirely upper case and the date and time is added to the beginning of every packet, this is fine with the FCC. It Tadd suggested that we brainstorm this and perhaps makes transferring binary information very difficult and the group might come up with something valuable. inefficient. NEDA must require that a packet node al- Don acquiesced and asked Burt to continue. low information packets to be sent without being modi- fied, within a given length restriction (TheNET limited Burt continued that node software that didn’t allow us to 200ish bytes or so). for inspection of the parameters should at least have a copy of the parameters available for download. For in- Tadd restated the question “What is required of a stance, in G8BPQ node software the parameters would network node software in order to be NEDA compliant?” have to be downloadable and the connect message should He asked of ROSE is NEDA compliant. He offered that indicate how to do so. the answer would be no because you can’t get routing information from a ROSE node (ed: actually you can in Tadd said he had a question to start things off. First later versions but only of the appropriate add-on appli- he asked if anybody else had a comment already. No- cation is uploaded to the node). Tadd asked what ques- body volunteered. He then asked what the club should tions must we ask about a software in order to not ac- do if somebody came up with a network software that cept it as NEDA compliant. Tadd asked if FlexNet is met the two criteria that Burt mentioned, i.e. it allowed NEDA compliant. BJG said yes. Tadd said the he didn’t shared access, and allowed inspection of routes and pa- have enough information to accept that yes on. BJG rameters. Tadd asked if this would be ok? He sug- said that FlexNet met all our requirements. Don asked gested that as devil’s advocate, what if somebody wanted what our requirements were. BJG said that it met the to write the worst software they could that met our defi- ones we just spelled out. Don and BJG laughed. Don nitions of compliant. What definitions would we want said that we were putting the cart before the horse and to counter this? that declaring that FlexNet was compliant was easy and Burt asked for an example of how someone could make wouldn’t create the list of requirements. Several attend- a bad network node that met those requirements. Tadd ees of the meeting reacted to BJG’s statement because suggested that a node where all commands had to be they weren’t aware that it met even the previously men- entered in Sanskrit would do it since we had made not tioned compliance requirements. Burt pointed out that N.E.D.A. Quarterly v5.2 page 16 we would be adding to our list of requirements for a 5. The software must be well documented, allow- long time. Don said this was true but we wanted to start ing most anyone to figure out how to use it effectively. with more than just a few. Burt pointed out that this is This documentation must include information on how the first time that TheNET has had competition that to map out a network built using this software. met even the few requirements we’ve just mentioned. 6. There is no requirement that frames other than G8BPQ had invisible parameters. AX.25 must be carried, but ability to carry other PIDs is Don said that part of being compliant was more than encouraged. just having the features in software. We can configure 7. The software must carry the data field portion TheNET to be non-compliant merely by turning off the of a frame transparently, within a reasonable length ROUTES and PARMS commands. However, he added, specification. In other words, it is OK for the software to the sysopping and architecture standards that must be refuse frames that are too long. followed are a discussion for another day and another document. Burt asked what he meant by that. Don 8. The software should perform automatic routing said that we wouldn’t declare a node compliant if it sent of connect requests. backbone traffic on a user channel to another node or 9. The software should meet all legal requirements server. We don’t believe that single ported node sites for use on Amateur Radio. are NEDA compliant general purpose nodes. We’ll talk about definitions and practices in another discussion. Don asked what the attendees would think about node software that does not digipeat. Could it be called com- At this point Dana WA2WNI entered. He was look- pliant? Tadd said that he thought node software could ing around at the cars in the room with our meeting. be NEDA compliant that couldn’t digipeat, and could (ed: I should point out that the meeting was being held also be compliant if digipeat was always available. Simi- at one end of a Mercedes Benz garage. There were brand larly a software could be rejected even though it did or new Mercedes parked about the place) The room is im- didn’t digipeat. Nobody argued or complained. Burt maculate. It is now 10:22AM. K1MEA checked in on said that we encourage people to turn of digipeating. the radio at about this point. Many cheers. Burt said that TheNET didn’t handle digipeated traffic BJG pointed out that we were talking about whether the same way it handled traffic where the user had con- a networking software was compliant, not whether a nected in and then connected away. particular node installation was compliant. He asked Don asked if point #8, that the node routes automati- to cut to the part about whether FlexNet was compliant cally was agreed with. He asked if we would reject node or not. software that required step by step navigation through Don wanted to make sure that the group understood the entire network. He also asked if we would reject that networking practices were important to NEDA but node software that denied the ability to step through not important to the approval or disapproval of a par- the network. ticular software package. Tadd said that he though either would be ok but that Don read the list of compliance items that the com- something that performed both would certainly be mittee had done. prefered. ROSE, for instance, did not allow you to go step by step. TCP/IP, doesn’t either, in general. You Working list of NEDA’s network software compliance have to have password access to each machine along the requirements: way in order to do this. 1. The software permits shared access to the net- Tadd said that if you couldn’t get routing information work for all users in a fair and equitable manner. from any and all nodes, that it would have to be rejected. 2. The actual node and network configuration set- It would be ok with him if you issued the “get routing tings must be viewable remotely by any user. This can- information at XXX node” at your local node and the not be only a text file whose content depends upon the network delivered. This means that, no, you don’t have sysop or others to update for this information to be avail- to be able to step through the network. able. K1CSB and K1MEA arrived 3. The node must be able to implement separate Dana pointed out that the node must be able to take backbone and user channels, and prevent user access advantage of hidden transmitter free backbones. Tadd on backbones and similarly prevent backbone access on suggested that it might be good to specify that it must user ports. be able to perform schemes aside from point to point 4. The software must carry data without regard to and retry-backoff, for instance, DAMA, and token pass- content or purpose, within legal and technical limits. ing. FlexNet can do DAMA. TheNET cannot.

N.E.D.A. Quarterly v5.2 page 17 Burt asked if KA node is compliant network software. have no idea of what hardware a network link of the future may be running across. What if we had a link Tadd pointed out that TheNET can be, and is usually with a bit error rate of 1 bit per century but was rather used in, a non NEDA compliant fashion, with all nodes slow to switch from transmit to receive. Would you re- on one frequency etc.. A network made up of dual port ally want to insist upon an acknowledgement? KA nodes could certainly be constructed that used dedi- cated links and whatever, Tadd said. (ed: it would be Don said that he agreed that we shouldn’t knock down more expensive than TheNET though and not be nearly a node technology that might come into play in the fu- as easy to use or map) He added that he didn’t know if ture by restrictions required to support our current tech- there was a way to create backbone links that would nology. restrict user activity (ed: considering how much we Don said that this would be published in the Quar- hacked TheNET 1.0 I’m sure it could be done). Tadd terly in the minutes and as a separate article with a said that what we need to do was to declare this compli- request for comments. ance document in advance and that way we wouldn’t have to seem to make up answers on the fly the next (ed: this subject took 30 minutes). time somebody asks if such and so node software is com- pliant. TCP/IP via FlexNet Tadd pointed out that we already have a network of Bob (BJG) said that he and others did testing of Flex95 mixed node software and that this was ok. In the future IP package. He said that they ran into limitations with it would be real good if all of the nodes were running Microsoft’s TCP/IP stack. Don mentioned that Gunter software that met a consistent compliance document.. Jost (FlexNet’s author) had run into similar limitations. Bob said that the good news was that he could use mail Don asked about documentation. He said that we to connect to a remote server and get mail, but that’s want documentation on how to map the network when all. He said that the FlexNet software worked fine as a the network is made up of any compliant software. He transmission media. He said that only if you went and pointed out that neither TheNET or FlexNet had this bought the full Microsoft Outlook packet could you do documentation before NEDA wrote it. Tadd said that the rest. He said that what we’re doing about that in he though that any definition for compliance should state the NY area is that we’re taking the mountain to sites that the documentation on how to map the network must that were just routers before and converting them to be available and in English. He said that before any full blown mailboxes. He said that when this was com- software is used on our network that it is up to the pleted that they’d do further tests with windows IP pack- implementer of that software to make sure that the ages against the MFNOS machines on the mountains to mapping documentation (and any other usage instruc- see if they could send mail to them and get mail off of tions) must first be made available in English. Don asked them. As far as FlexNet itself and the Flex95 IP pack- if we should declare English to be the language of NEDA. ages, Bob and John (CIK) have done testing from one Tadd said that he thought so. Don asked Burt (VE2BMQ station to the other with the 95 package and that it works from Quebec) if he had any complaints. He said, with a very well. He said that they were on the phone with laugh, “no!” each other and that it works very nice. He said that the Don said that if we didn’t specify would could be stuck trouble they had was that they didn’t have the applica- with pig latin. Tadd added “isthey acketpay etworkney tion packages they needed for Flex95. isay appablemay.” Bob said that they had no trouble whatsoever carry- Don brought up some other topics: does there need to ing IP traffic on normal FlexNet PC since Barry modi- be hop by hop acknowledgements? How about end to fied MFNOS to use IPCAM. Don said that he talked to end acknowledgements. Should you be able to read the Barry K2MF recently and he was told that things were I-frame or address fields of the traffic between the nodes? flowing again. Is it ok to have the data compressed? Is it required to Bob said that he was operating a TheNET to FlexNet have the data compressed? Should the origin and desti- gateway and that the link over to K2MF’s system was nation station be available for each packet copied on a on TheNET. He said that there was a constant flow of backbone node? How about the ability to connect out of traffic through the gateway. Discussion ensued in which the network to a user station that is not connected to they talked about how traffic through the system and to the network? (note that we think DxCluster doesn’t let and from Barry via internet had resulted in a silly inci- you do this but FlexNet, ROSE, TheNET, Kanode do) dent. Tadd said that he didn’t think that hop by hop (ed: CIK sent us the web URLs for a few sources. See acknowledgements should be required. Don disagreed. NOS Info article) Burt pointed out that TheNET doesn’t do hop by hop of everything, nodes lists, for instance. Tadd said that we N.E.D.A. Quarterly v5.2 page 18 Hidden Transmitters and DAMA Jim asked for some specifics. Tadd said that he thought it was reasonable in the Don reviewed and pointed out that DAMA would keep near future to eliminate hidden transmitter syndrome one station from monopolizing the channel. Further, all as an effect in and around our networks. This would be of the delays implemented in standard packet to ease done through the use of the DAMA features of FlexNet. channel congestion, DWAIT, FRACK, PPERSIST, DAMA means Demand Allocated Multiple Access. This SLOTTIME, are no longer used. The channel runs as feature uses polling from a DAMA master to arbitrate fast as the active stations can handle. Somebody moni- when stations on the frequency would transmit. In gen- toring a busy channel would hear very few gaps and no eral this increase throughput by either a little, or a lot, collisions. As soon as a 3rd station is turned on line on if there are ever more than 2 stations on the frequency. the channel, the efficiency is increased. Burt pointed The elimination of hidden transmitter syndrome would out that one of the features of DAMA was that the be done in several steps. First we should test the DAMA amount of traffic allowed through a particular station is features on redundant user ports in areas where there based on the demand for that station. The DAMA mas- are volunteers. During this state we should convert all ter uses the recent history of the channel as part of it’s multiway backbones to DAMA supported backbones. calculations for channel arbitration. At no time is an active channel given none of the channel time. DAMA The second step would be to educate the local users request frames are the only frames that are sent out of through seminars and hand-outs (or via other means). sequence. SABM frames (normal packet connect request) We’d run demonstrations and whatever to help the are responded with an error message. process. John CIK mentioned that DAMA depends on line-of- Finally we’d restrict all access to the network to DAMA site like channel performance. He said that fading and ports. All users would have to be running TNCs that other properties of 6 meters, for instance, made DAMA supported DAMA (like many of the Kantronics models) on 6meter backbone ports less useful, especially if the or they’d have to have a personal computer that can run channel is not highly loaded. He has a 3-way backbone DAMA software to back up their TNC running in KISS on 6 meters where once in a while the path between any mode. FlexNet and Flex95 support DAMA and are free 2 of the 3 nodes fades and the signal can automatically to hams. be rerouted to the remaining 2 paths. This would not happen using DAMA. Tadd said that it might be worth while to insist in the next year or so that zoo channels (channels with mul- There was much talk about implementing test user tiple servers) or multiway backbone channels should be ports. converted to DAMA channels. Education is clearly required here. John CIK and Burt both pointed out that we are hav- Perhaps we should create and distribute a DAMA help ing a hard time increasing or keeping our numbers of sheet. users up. We shouldn’t come up with any restrictions on user access without a serious education and docu- CIK said that we made history recently by adding a nd mentation distribution plan in place up front. All agreed. 2 user on his 9600 baud port. Many claps. He said that the user was over in Plainview, Nassau county. Tadd said that the trick was to make this possible, then make it likely, then educate people, then, if it still Bob BJG pointed out that he’d had one on for a year a seems a good idea, attempt to impose it. while ago and that since nobody used it he’d pulled it. Burt said that he’d had trouble implementing one so CIK said that he thought the first thing we should do far. is to try it on the backbones. He said that he tried the TxDelay test on FlexNet that refuses access by users Bob BJG said that it should be a goal that DAMA be with TxDelay set too long. He said that the test was too implemented on the multiway channels. strict in that users could be dismissed for a TxDelay Burt said that it would definitely improve operation that worked at another time. Further the users would on a repeater. find their TxDelay too short under other band or load- ing conditions but would be refused access when the con- A discussion developed about how the DAMA master ditions improved. So.. that feature needs to be used was selected and what happens if it goes away. The with care on some ports or we’ll lose too many users. answer is that the DAMA master is selected by an op- DAMA would be bad too if turned on injudiciously. tion bit in the PORTS command. The other stations become DAMA slaves because of the presence of the CIK said that we should do nothing to discourage us- DAMA master. The DAMA master has to be the station ers. Dana added that “Education Must Precede Imple- with the best connectivity on the backbone. This is a mentation.” better idea than picking the one with the most data to N.E.D.A. Quarterly v5.2 page 19 send and receive. On a user channel the DAMA master Tadd went through a few examples of how adding a would be the one and only server. desktop MSDOS computer would cause trouble. He pointed out that some of the node sites are limited to 6 Good questions about DAMA inches or less high by 19 inches wide in an equipment If A, B and C are stations. If traffic is going mostly rack. A desktop computer could consume all of that from A to B and B to A, can C still be the DAMA master space. He pointed out that some node sites were lim- and will the traffic flow from A to C to B and B to C to A? ited to 12 volts from the battery in order to take advan- The answer is yes C may be the master. C will arbitrate tage of emergency power. A desktop computer needs when A and B transmit. C does not carry any of the 110 volts. traffic, it just gives marching orders to A and B. John CIK has brought a single board computer that What happens if the DAMA master goes off the air. If runs MSDOS out of ROM. He said that the one he had the DAMA master goes away, the other stations fall back was a discontinued product. His looked to be about 5” x to normal backoff and retry operation. 8” and he said it drew 660mA at 5volts which equates to around 3.5 watts.. What if you have a string of stations A, B, C, D in that order, such that only the immediate neighbor can be CIK said that he’s checked out some PCs and discov- reached. Which should be the DAMA master? The an- ered that they draw 45 to 55 watts. This included 20MB swer is none of them. This system won’t work with HD, floppy, RAM, 286CPU, with no keyboard. DAMA. All stations on frequency must be able to com- Tadd did some research and found a company named municate reliably with the DAMA master. JumpTec in Germany that makes a 33Mhz 80386SX If A, B, C are stations and B is the master, what kind single board PC called the DIMM PC/386. It is sold in of efficiency can be gather if only A and C usually have the US for $250. The device is a 1.5” x 2.5” board that traffic by making this a DAMA setup instead of a 3 way draws 300mA at 5v with a 2MB integral flash hard drive backoff-and-retry. The answer is that if this is a three- and 4MB of RAM. It looks like with an evaluation kit way backoff-and-retry backbone, then PPERSIST, SLOT, costing $500, the computer can be formatted with FRACK and DWAIT are implemented. This makes the MSDOS 6.2 and then FlexNet installed. It may then be link slow even if there are no collisions. DAMA would updated without the evaluation kit. The JumpTec board always be faster. If there WERE collisions, the channel is available as a 1JDMPI2 from EMJ http:// really goes down the tubes. www.emjembedded.com/ in the US. That is the $200 single board computer. You also need a connector and John, CIK made several good points about how DAMA RS232 driver board (the on board serial ports are TTL i/ is configured in FlexNet. Don pointed out that there o) It looks like the RS232 board is $28. We may need a was a DCC proceeding that covered DAMA. Perhaps connector to get power, keyboard, and reset signals to we should find this and reprint it. the board. It also has a MSDOS compatible parallel 11:00AM port. This computer could fit inside of a TNC chassis. This would give those with space, or power limitations a Ruggedized PCs FlexNet upgrade path that has little impact on the site. Tadd led into this subject with a review of PC- The unit draws less power than most TNCs. FlexNet’s requirement for an Intel personal computer The board that John CIK brought is compatible with as a control box. If we choose to implement FlexNet a passive PC backplane through which you could get using the MSDOS system (instead of the more expen- additional i/o and VGA output. This could be used for sive RMNC cards) to take advantage of existing TNCs development purposes. In actual use it could be oper- as modems, then we need a MSDOS computer at each ated with it’s on board RS232 ports. It has a keyboard site. There are several problems incumbent on a node connector as well. That board needed a software devel- owner trying to upgrade to FlexNet with regard to the opment package as well in order to set up the EPROMs. MSDOS computer. These are: The JumpTec board could be booted from a floppy drive The MSDOS computer is large by using the $500 development package. Then the on The MSDOS computer draws lots of power and that board flash ROM could be initialized by doing a format at 110 volts. /s. The MSDOS computer has moving parts (i.e.floppy Tadd said what we need to do is identify sources of and/or hard disk) and we think this might reduce reli- imbedded PCs and see what we can use to implement ability. solid state PC FlexNet nodes. The MSDOS computer costs money. John and Warren were talking about how passive PC backplanes were around and since we’d only need one

N.E.D.A. Quarterly v5.2 page 20 or two to use for setting up a series of passive backplane had concerns. This includes Montreal and etc. Burt based imbedded PCs this would save money over using brought up the fact that some sites can’t be visited for the JumpTec model. months at a time during the winter and unless we could have a machine on hot standby, he’d prefer a more reli- (ed: http://www.winsystems.com/embpc.htm is a able solution. Using a helicopter to do node servicing is URL for a system Dana sent in after the meeting) not reasonable. Ruggedized node hardware is impera- Burt warned about developing a methodology for us- tive. Burt said that he was a firm believer in Murphey’s ing a certain product when that product is either already Law. He said that it’ll work find so long as you can get outdated or likely to be so soon. We’d be better off find- to it. Once you are no longer able to go to the site you ing a source that is known for keeping a product on the can expect a failure almost immediately. market for a while, than using a cheap surplus solution. CIK mentioned that in any rate you should think about Burt suggested that we could adopt a technology that having a remote reset switch with touch tone access if could be used across a generic type of machine. possible. Tadd held up the TAPR flash card IDE hard drive It’s now 11:30am. interface board. He said this might be a solution for some users who don’t have space or power constraints Frequency Coordination. but still want a solid state solution. The flash card looks Alex, N3NMN said we should be performing some kind like an IDE hard drive if installed into a PC. of frequency registration to protect our backbone fre- A discussion was held about the types of Intel and quencies. Further, it is worth having a dialog with the other processors that don’t draw much power. CIK said frequency coordinators in order to not locate a backbone that ceramic package devices radiated more heat. on the same frequency as a repeater input or control link. Burt commented that the TAPR card was so simple that we should look at doing our own. CIK thought that The opinions of the packet network builders present it wouldn’t be worth it considering how few we’d need. about working with coordinators was wildly varying. Jim, K1MEA, thought we’d need one per node. Burt Jim MEA pointed out that working with coordinators thought that the flash card connector was probably the was much easier if you were in the middle of the area most expensive part. Don, CIK and Burt had a conver- that the coordinators protected/controlled. When you sation about how hard it is to make a board using 50mil are an area within 50 miles or so of a border between spacing connectors. He said that he could get some large coordinator regions. number of boards for $200. Perhaps he could make them for 20 for $200 and then get the connectors for $10 or so. Bob BJG and John CIK both suggested that the NY metro area was very difficult to coordinate because the An outstanding problem is how to make the system personalities of the coordinators (and the people request- boot from the flash ROM hard disk and then run out of ing coordination) were strong and often uninformed and/ RAM. A problem with FlexNet is that if it runs out of or not interested on consideration or objectivity. They Flash ROM it will also store temporary information, like reported that the backbone frequencies in use in the NY heard lists, and parameter values, onto that Flash ROM. metro area had been reported to TAPR for inclusion in This will prematurely wear out the Flash ROM drive. A the packet radio database that TAPR was operating. solution would be to have the unit copy the FlexNet sys- tem onto a RAM disk and then run FlexNet from there. Don suggested that he didn’t think that the coordina- The problem is that then if some sysop changes were tors necessarily knew what was the best frequencies. made the the FlexNet over the air, like routing informa- He thought that if the repeater coordinators did more tion changes or Text message changes, that those changes than suggest frequencies that they were not credible. would remain in the RAM. If the system was reset, Alex said that the potential frequency user should those changes would be lost. check out several frequencies that seem to work and to John CIK said that there was a FlexNet utility called be clear [through some amount of monitoring from both SERV that let the sysop have access to the PC by re- sites that the link uses] and then ask the coordinator if mote control. This could be used to command the sys- any of those is available. tem to copy the parameters from the RAM disk back to Don said that he didn’t think the packet network build- the Flash ROM. We’ll have to do an expose on SERV.EXE ers were obligated to follow the demands of the coordi- in the Quarterly. nators, unlike the repeater users. He said that the re- BJG had mentioned that he had no problem with hav- peater coordinators must be better informed that the ing a real hard disk on the PC. A discussion comparing link developer and that they would have good informa- the weather conditions in the various places that NEDA tion. The answer must be an open dialog between a link

N.E.D.A. Quarterly v5.2 page 21 developer and the coordinators. Since link developers Don pointed out that there are regions of the UHF don’t want to upset anybody they are likely to follow rea- band designated for packet radio that are rather large sonable suggestions from the coordinators. because they were defined for 100Khz channels. Don says that he doesn’t see any real issue with what If someone from NEDA wants to volunteer to be a Alex is suggesting but that he expects many link devel- coordinator laison or to manage a coordination group, opers to not cooperate. that would be fine. Don thought that this was beyond the scope of the technical committee. Alex warned that there was a P.R.N. circulating that could give the coordination board some teeth. He said Tadd pointed out that this was discussed and planned that there will be enforceable coordination. He said that in Oneonta in 1994 by Burt. Ray Hollingsworth from the FCC has said there will be. He said that if there is an on the air conflict between a Netrix & 6pack repeater and a packet link that packet will lose. Many Tadd suggested that the club might want to build a around the table disagreed. 6pack connection board that looks something like the Burt pointed out that coordination of packet radio is Netrix. It would be a strip of board that had four con- different from coordination of voice repeaters in that a nectors to plug into 6pack TNCs. It would have a fifth packet link has two coordinated ends where a repeater connector to go to the Intel PC. It would have no active is only one site. Burt said that many packet link devel- components. It would want jumper pins that you could opers don’t expect the coordinators to understand the cover up to bypass a missing TNC for the 6pack loop. difference. In general you don’t know where the user John CIK, said that he thought it would be too ex- sites are going to be on a repeater so you coordinate the pensive. Tadd said that he thought it would cost $10 repeater site to be quite a distance from any other re- each. John thought that since most builders would use peater site on the same frequency. On packet you are surplus PCs with slow serial ports that they’d be better trying to coordinate two sites to make them talk to each off dedicated one serial port per TNC. You’d have to other in order to form a network. On packet the links have a higher speed serial port. He said that the sur- are directional and the coverage area could be much less plus computers won’t handle a 19,200 baud data stream than a repeater. Indeed, backbones could be on the same without an upgraded serial port. He said that he frequency in the same region because of the use of direc- thought it was more cost effective to go to a 4 port se- tional antennas and adequate power. Repeaters use as rial card instead of building a netrix-ish card. much power as they can get away with and afford. Links use only what they need to get some amount of fade Tadd said that most of the nodes in the network are margin. running four or five 1200 baud ports. Thus a pair of 9600 baud 6pack ports would do just fine. Tadd is think- Burt said that the most important think was to create ing that an imbedded single board computer with two a database of all of the links and then cooperate with the ports could run two six pack loops and handle almost coordinators to register the backbone links. every node in the network, with one loop running 19,200 Bob LWR commented that he’d run up against a re- baud and the other running at 38,400 baud. You could peater owner that was certain that a coordination gave put four 1200 baud TNCs on one and four 4800 baud the repeater owner a clear channel throughout the state TNCs on the other. of PA. Bob BJG pointed out that 9600 baud technology has Don said that he figured that NEDA participants are been around since 1988. Tadd pointed out that what probably very cooperative compared to many hams imple- technology exists has nothing to do with the goal of menting fixed long term transmitters. Burt said that he making as comfortable a conversion to FlexNet as pos- thought that packet radio, including full duplex repeat- sible. ers, should not be co-channel with voice repeaters, even Don asked for a conclusion. Tadd said that he thought at great distance because it is the nature of automated we could make 100 four port 6pack loop packages. Don packet stations to not notice lift/band opening conditions. asked if we should sell just the board and leave the Voice operators stand off and listen when the openings connectors up to the user. John CIK pointed out that occur. A digital repeater operator would jam all of the the connectors would be expensive if bought individu- repeaters that are reached during the lift. This would be ally, compared to buying pre-assembled serial port annoying. It would be good for repeater coordinators to cables. Don said that he gets gold plated DB9s for 23 leave room for digital full duplex repeater pairs that are cents each. Don said that he thinks the kit cost would available across coordination areas. be $5. Don said that if you want an easy 6pack solu- tion, this would be it. It doesn’t make sense to use this solution if you are running several 9600 baud TNCs.

N.E.D.A. Quarterly v5.2 page 22 Tadd said that he had been charged by people who said that he is very worried about internet forwarding live in an area that has lots of FlexNet nodes to promote trashing the BIDs. FlexNet into areas that do not. He said that he lives in an area that does not have many FlexNet nodes and Show & Tell that this 6pack loop board would help him sell FlexNet Don showed off his laser link. It used an unmodified to the local node ops. He says that that is perfectly rea- laser pointer whose power supply is switched by the TTL sonable to come up with a solution, even a costly solu- output of the serial port on a TNC. He said that it turns tion, that trivially upgrades the nodes from TheNET to on and off at about 11Khz. He said 4800baud is pretty FlexNet. Just the existence of that solution will be fast. Since it’s a laser it could be full duplex with two enough to get some of the TheNET node ops thinking in pair of transmitters and receivers. The receiver uses a terms of doing it. Some of those may choose cheaper lens and photodiode from radio shack. He uses a com- computers, multiport serial card add-ons, or whatever. parator to reform the pulses back to TTL. He said the He said that some of the existing TheNET node ops con- lens is used to allow the laser to be pointed more inaccu- sider the idea of putting a full size MSDOS computer at rately. The laser beam is actually still fairly tight at the node site a horrible idea. Tadd said that he knows of what he thinks will be the maximum range of the link, 9 node sites in New Hampshire. He said that power hundreds of feet. He’s been sending square waves budget and enclosure size are real concerns of at least 5 through it to test it and measuring the receive signal on of those sites. Moving parts issues are also concerns. If a scope to compare it to the transmitted signal. It looks we can develop and deploy a single board computer so- pretty good up to 11Khz. lution and a simple 6pack board, for less than $350 we may see FlexNet take off very nicely. Dana, WA2WNI, John CIK related that he needed to make a high speed said that he’d go for that. He is resisting so far. Tadd wireline link through 60 floors of an office building. He said that once we make this an off the shelf solution said that he changed the resistors on the modem in an some will look at FlexNet as viable, even though they MFJ1270 to make it higher bandwidth. Instead of run- may take another course to implement it. Tadd said ning it at 1200 and 2200 hertz he ran it at 10khz and that we spent several years convincing people that 18khz. He was about to run 9600 baud data across a G8BPQ was not to be used at node sites for reasons of hundred feet of twisted pair. John was suggesting that PC reliability. Now we’re trying to convince them that it might be worth running AFSK instead of straight PCs are reliable? baseband. Don said that his system cost about $50 bucks for two Networking Standards & BBSs ends. Bob talked FlexNet TheNET gateways. He said that The meeting broke down into a discussion of using you need to assign an SSID to each port that is going to 2KW cutting lasers to run long distance links. face a non-FlexNet station, and that includes users, and TheNET nodes. Furthermore, you must have digipeat Recommendations to the board enabled on the TheNET node that FlexNet faces. This means that TheNET 2.10 will not work with FlexNet as Don said that he was going to recommend to the board a gateway link. We have to add a new column to the that they persue the 6pack loop device. He was going to TheNET parameters list in the Quarterly to show Flex- pass the coordination info from UNYREPCO to the Net gateway as a port type. This gateway should have board, and finally a request to be able to publish the digi enabled. Jim and Tadd pointed out that most of the NEDA node compliance draft in the next Quarterly. backbones in New England were TheNET 2.10. John brought up a reminder about adding a new line Bob BJG said that there were three-ways and zoo to the X1J spec for the FlexNet gateway support. Pa- channels in the network still although they knew that rameter 23 has to go to 1 and Mode 17 has to go to 0. dedicated links were good and Bob and the other net- TheNet 2.10 won’t allow a FlexNet connect. work builders were pushing and implementing 9600 Burt and Tadd decided they should make FlexNet a baud dedicated links. new kind of port with it’s own column. Another com- He said that the BBS operation in his area were get- ment from CIK was that the node name should be ting worse and worse. He said that the distribution changed to FXGxxx so you could tell it was a FlexNet groups, like NEBBS and TRIBBS were abused and for- Gateway node. warded to obscure locations. He said they were getting Jim asked if FlexNet had a conference capability. John section manager reports from Nebraska in the NYC area. CIK answered that there were two different conference Bob said that there were serious problems with BBS capabilities. One was that there is a round table fea- sysops not paying attention to good practices with re- ture entered into with a C command. The other is the gard to where bulletin messages get flooded too. He

N.E.D.A. Quarterly v5.2 page 23 ability to send a one line message to a connected user with the T command. CIK brought up FlexTalk which he was wondering about. Don said that FlexTalk is a FlexNet-compatable voice communications system. Don thinks that it needs 3.2Kbaud of throughput per voice which means it should be running on a fairly fast network.

Adjourned Don closed the meeting at 12:16pm. –NEDA

N.E.D.A. Quarterly v5.2 page 24 N.E.D.A. Quarterly v5.2 page 25 Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps Maps John Kushneir, N2UBH John is from Syracuse NY. He has been instrumental Can you help with Maps? in making the central New York packet radio network There are two processes that we need to take place. the best part of the north east TheNET packet network. One is a process we need done, the other is development John is the man responsible for the maps. He’s updated to simplify the process. many of them so far and is working on the rest. See article below. The Process We create our maps from a network survey. The map created from log files of actual network commands.

Log File • For a TheNET, that log file should contain: • the infotext, • node name from each user port, • callsign from each user port, • nodes list, from each user port, and • routes from each of the backbone ports. • For a G8BPQ node the log should contain: • the node name, • nodes list, and • routes from the node. Map Report • For a FlexNet the log should contain: • the node location, • callsign, Thanks for any support you can give! • PORTS command results –N2UBH - the map guy It helps the map making process to start with all of the log files already accumulated. A volunteer should gather the log files from local nodes and forward them to somebody who wields the graphics program. (NEDA bought all of the map makers their own copy of a graph- ics program several years ago.)

Draw Map From the log file a map is created or updated. The map is drawn to show the backbones between the nodes, as shown on each map. See the inset box on that map for a list of link types that are shown. We were using Aldus Freehand (now owned by Macromedia) until 1993 or so when we switched to CorelDraw. Development What we really need, is a program that runs on a per- sonal computer, and that connects from a user’s station out into the network, and that gathers the log file speci- fied above. This should be pretty easy in C++ or possi- bly even BASIC. Anybody got time for a nifty project? If you wonder how much this project would be appreciated, think of the names of the BBSs out there. W0RLI? F6FBB? G8BPQ? So... create your node logger software and call it YOURCALL. I promise you we’ll forever refer to it by your callsign! —NEDA Editor, Tadd, KA2DEW

N.E.D.A. Quarterly v5.2 page 51 N.E.D.A. Quarterly v5.2 page 52 North East Flea Market Dates From: [email protected] http://mit.edu/w1gsl/Public/ne-fleas Last Update 4-1-1999 de W1GSL 2 May Yonkers NY Metro70 @LincolnHS $6@9 $25/T@7 ...... Otto WB2SLQ 914 969 1053 F 7,8 May Rochester NH Hoss Traders @FG x13 rt16 ...... Joe K1RQG 207 469 3492 List is normally updated twice a month - look for 10 May Greenfield MA FCARA Mon 6PM @WESCO $3/tg ...... Richard KD1XP 413 665 2211 + the latest version 14,15 May Dayton OH Hamvention ...... http://www.hamvention.com Source: 15 May Forestdale RI RIAFMRS @VFW rt146 8A flea+auct ...... Rick K1KYI 401 725 7507 F= Flyer T= tentative early info 16 May Flea at MIT ...... Nick 617 253 3776 F + = new info this month 22 May Londonderry NH IRS @Lions In+Out $10@6 $2@8 ...... Paul K1LL 603 432 1538 A= ARRL 29 May Vernon CT NARC@TollandAgC I84x67 $3@9 $15/T@6:30 ...... Wayne 860 487 1921 V= VE list D= W1DL 30 May Sorel-Tracy PQ HamfestduQuebec $15/T@6 $5@9 ...... Jean VE2JCE 514 587 2986 W= web 4-6 June Rochester NY ARRL Atlantic Conv ...... Harold K2HC 716 424 7184 A M= W1JTH WR NV 73 CQ QST = 5 June Herman ME PineStARC...... Rodger KA1TKS 207 848 3846 A Mags 5 June E Hartford VintageR Museum @1231 Main $12/Tg@7 $1@8 ...... John 860 675 9916 F 6 June Corona NYC @ HallofScience 47-01 111st ...... Steve WB2KDG 718 898 5599 A This list has been compiled from many 6 June Newington CT NARL @HS $5@9 $20/T@8 $10/tg ...... John KA1HQK 860 666 8569 F+ sources. While we 12 June Goshen CT SoBARC @FG rt 63 nr4 ...... Lee K1LEE 860 435 0051 A believe the info to be accurate the author can 12 June Falmouth MA FARA ...... Ralph N1YHS 508 548 6405 A+ not be responsible for 12,13 June Lancaster NH MooseSwappers @FG $10/S$5B@7 ...... Russ N1YZE 603 922 5514 F changes or errors. 13 June Bethpage NY LIMARC ...... Rich N2WJL 516 520 9311 A Check with the sponsoring organizations 20 June Flea at MIT ...... Nick 617 253 3776 F for more details. This list 10 July Union ME Pen-BayARC @FG $4/T@6 $5@8 $2tg ...... Will KD1ZS 207 785 2739 + will be posted monthly to USENET. Mailed copies 11 July Patchoque NY MidIARC ...... Mike N2OX 516 736 9126 A are sent when additions 18 July Flea at MIT ...... Nick 617 253 3776 F are made. 24 July Nashua NH NE Antique RC $5@8 $1@9 @ Res Ctr Church...... 617 923 2665 Additions/ Corrections via Internet [email protected] 7 Aug Plattsburg NY CVARC ...... Bernhard KC2ALG 518 643 9657 A US Mail 7 Aug St Albans ME @ Snow Mobile Club ...... Howard WA1SBI 207 876 3702 M+ W1GSL 15 Aug Flea at MIT ...... Nick 617 253 3776 F POB 397082 MIT Br 22 Aug Yonkers NY YARC ...... John WB2AUL 914 969 6548 A Cambridge MA 02139 29 Aug Enfield CT East V-UHF Conf ...... [email protected] Mark K1MAP 413 566 2445 (c)1998 W1GSL 11 Sept BallstonSpa NY SCRACES fri6P $5+5/T+15cmp ...... Darlene N2XQG 518 587 2385 http://flealist.senie.com/ 11 Sept Windsor ME AARA @Fairground $5@8 s@6:30 ...... Frank N1ITR 207 623 9217 + SASE for updated copy 12 Sept S Dartmouth MA SE MassARA @club $2@9 $10/S@7 ...... Bill K1IBR 508 996 2969 F as issued. unlimited reproduction permitted in 18 Sept Forestdale RI RIFMRS @VFW rt146 8A flea+auct ...... Rick K1KYI 401 725 7507 entirety 18 Sept E Hartford VintageR Museum@1231 Main $12/Tg@7 $1@8 ...... John 860 675 9916 F 18 Sep Marshfield MA GenesisARS @FG Rt3A $20FriPM$5@9 ...... Lou N1WNT 781 837 6651 + 18 Sept Lincoln ME Bagley ARC @BurrSch $3/S@7:30 ...... Sylvia N1JNR 207 732 5185 + 19 Sept Flea at MIT ...... Nick 617 253 3776 F 19 Sept Bethpage NY LIMARC ...... Rich N2WJL 516 520 9311 A 19 Sept Newtown CT CandlewoodARA ...... Jeff WB3DLG 203 798 6860 A 26 Sept Yonkers NY Metro70 @LincolnHS $6@9 $25/T@7 ...... Otto WB2SLQ 914 969 1053 F 25,26 Sept Lancaster NH MooseSwappers @FG...... Russ N1YZE 603 922 5514 F 3 Oct Queens NYC @ HallofScience 47-01 111st...... Steve WB2KDG 718 898 5599 A 8,9 Oct Rochester NH Hoss Traders @FG x13 rt16 ...... Joe K1RQG 207 469 3492 T 10 Oct Wallingford CT NUTMEG CT Conv $6@9 $15/tg@6 ...... Gordon K1BIY 860 342 3258 F

N.E.D.A. Quarterly v5.2 page 53 Board Meeting Minutes continued from page 3 Secretary’s Report extra copies during the 3rd and 4th quarters as that isn’t Don asked if the secretary (dana) had anything to re- flea market season. port. Dana said that the recorder seemed to be record- ing. Don asked if the secretary needed any supplies or Don said that he still hadn’t done his e-mail/internet anything he needed other than his dictaphone back. promotion campaign that is also likely to bring in new Dana said that his appointee (ed: I was sitting next to members. Dana) seemed to have it well in hand and regarding his Tadd said that he proposed doing an overprint for the dictaphone, he didn’t think we wanted to go there. July issue to support the fall flea market season as well. Jim moved that we accepted the minutes Jim asked how the printing costs were affected by quan- tity. The answer is that as the quantity increases we Ray seconded get to increase quality. The price doesn’t go down much. Passed For instance, we can use a copy plate and photo process for $20/page. Having made the copper plate, the print Editor’s Report run can be made on a printing machine, instead of on a Tadd said that he had several articles and no new photocopier. This drops our price by a little, and in- maps yet for the next Quarterly. He said he’d like more creases quality by a lot. That’s not economical with a articles and new maps and that was all he had to report small quantity print run. Tadd guessed that we should since we’d talked about the newsletter during the be at about 1000 quantity, or $1600 or so for the print treasurer’s report. run, before we can do that. Jim said that we should make another note that UBH Dana brought up that once we got our membership and YHR were instrumental in getting the maps done numbers up higher that we could send out 3rd class/bulk and should be congratulated in the minutes. All clapped. rate. Don reported that the number seems to be about Jim said that the only complaints the map makers had 200 before we can do the mailing at all. Don reported was that they couldn’t interpret Bob BJG’s extremely that the post office charges something on the order of 25 detailed text maps to translate them. They’d rather have cents per issue for 3rd class, compared to $1.65 per issue log files for all of the nodes than having other people’s for 1st class. Burt said that we should just send extra interpretations of the current state of the network. copies of the newsletter for promotion to increase our quantity. Tadd said that the disadvantages to going bulk Membership Report rate are that it takes longer and you don’t get Tadd said that we had nothing newer than what was undeliverables returned (nice for finding out who moved in the Quarterly or whatever). John MKH brought up that there was some confusion The problem is that maintaining a 3rd class permit about membership expiration dates. It seemed that and doing the sorting ourselves is expensive (around N2UBH had his expiration incorrectly handled. Dana $160 application and $80/year) and complex. The bet- said that he thought his was wrong too. Tadd said that ter way to do this was to hire a mailing house to do it. he understood from the membership director – WB2DWD This will cost $40 to $60 upfront plus the mailing costs. – that there were somethings that he was correcting in Plus it has to be prepared in a certain fashion. Tadd the database when he heard about the problem. Appar- said that Bob QBQ had decided not to use bulk rate a ently DWD had some difficulty starting up as the new quarter or so ago and that now that our membership is membership director at the beginning of the year. Mem- back up we might consider it again. Don said that for bers who believe that there is a problem with their en- 110 it is definitely not worth it but for 170 it’s coming try in the database should definitely send a message to close. Jim said that at the last stamp increase there Membership Director at the PO Box or can e-mail Bob were also major cost increases (at the post office) in 3rd at [email protected] class. We weren’t doing it doing that period. We have to look at it again. (ed: I called the PO. They said that Tadd pointed out that N2UBH’s membership is listed presorted 3rd class non-letter is on a cost per pound and as running out in January 2000 in Report 5.2 but in the cost per item basis. It costs 68 cents per pound plus database that DWD gave to Tadd just recently it shows 10cents per item. That means an 8 ounce item will cost as running out in January 1999. Clearly an error hap- 34 cents plus 10 cents each to mail) pened in the handling of the database. Burt said “Y2K” bug haha. Tadd said that Bob may have been using the Dana moved that the report should be accepted as wrong copy of the database from Cal or something like reported that. Jim second All approved. N.E.D.A. Quarterly v5.2 page 54 Don asked for a notice in the next issue of the Quar- posal was made for a liaison to perform this duty. Alex terly that asks people to check their own membership warned that coordination might be mandated in the fu- roster listing and request an adjustment if one is needed. ture. Don thinks that the FCC will mandate coordina- tion of control links and repeater frequencies but thinks Maps Report that packet is not going to be coordinated. Don would Tadd says the map makers want log reports. This is like NEDA to consider voluntarily working with them. the prefered method of delivery of map update informa- Tadd asked how NEDA could do this. Dana doesn’t think tion. The log report format is on the page after the maps that NEDA needs to do this at all. Don says that since in the Quarterly. Tadd briefly went over the form at the NEDA doesn’t have any packet links that it should be meeting. Anybody noting an error on the maps should easy for the club to cooperate. Haha. Don thinks that go to the node and log the node information that is dif- we should have an article in the Quarterly saying that ferent than the maps. UNYREPCO has asked for this and we should consider working with them, as node ops. Jim and Dana pointed John MKH showed an error in the maps that wasn’t out that the board is on record with this about 2 years picked up and a link that was deleted that only went ago, and 5 years ago and nobody has every carried out away for a short time during the map making process. the task. Tadd asked who the contact person for The problem was probably caused by the fact that the UNYREPCO is that should be contacted by link link that was away made it difficult for the map makers implementers. John MKH said that John N2UBH is to look at the site about which the error was printed. the contact person. John is One thing that we should fix is that there should be a [email protected] and is N2UBH@N2UBH. way to indicate that a link went away, but was there He’s the packet chairman at UNYREPCO. Tadd offers during the last map update cycle. (ed: I spoke with John that we could have a static piece in the Quarterly to put UBH about this. He’ll work out a way of showing links information about who to contact at the coordinating that broke) councils. Don points out that UNYREPCO should look Tadd pointed out that UBH was updating the maps at the TAPR information. John CIK said that the TAPR from several year old information. His first mapping system is updated for NY metro as of recently. attempt showed many missing links. The maps would have looked pretty ridiculous with links that the locals Old Business knew to have been missing for years shown as “unreli- able.” NEDA compliant node definition This was covered above Technical Committee Report Don the chairman of the meeting hands the meeting FlexNet over to Don the NTECH chairman. Tadd moves that we endorse the use of FlexNet. The technical committee requests to be able to pub- Dana seconds lish a draft of it’s NEDA node compliance document in The only worry is that we might find a problem with the Quarterly with a request for comments. Tadd moved it. Burt points out that we found things wrong with that we publish it. Jim seconds it. All approved. TheNET along the way and we just worked around them. The technical committee asks the board to authorize Tadd offers that we might change the amount of TheNET the pursuit of a Flextrix 4 board with the intent to even- stuff published regularly. We could refer to the web page tually have the club fund a run of 100 or so of them. or previous documents for the unchanging TheNET in- formation. John MKH seconds it. Burt asks if this can be done on a single sided board. Don says yes and a conversation breaks out around this. Jim asks if we are excluding TCP/IP users from use Apparently it will be easy to produce inexpensively. Dana of the network. Tadd said that we are certainly break- makes a motion that we bless the pursuit of this project ing the network in terms of the use of IP. We are also at no cost to the club. Tadd seconds it. All approve. making the network a hazard to local BBS sysops and DxCluster sysops. During the transition, use will be The technical committee comments to the board re- more difficult. IP use is particularly difficult because it garding Frequency Coordination. Alex, N3NMN, pro- is not possible to make the transition between the Flex- moted that the club should cooperate with frequency Net portion of the network and the TheNET portion of coordinators by obtaining and publishing lists of back- the network unless there is an appropriately located bone and user frequencies. The list would be securely passed to local frequency coordination groups. This co- operation would be voluntary on the part of the group and on the part of the node operators. No offer or pro- Continued on next page N.E.D.A. Quarterly v5.2 page 55 Minutes contined from previous page Dana asked how we should support TCP/IP at the points where there is a gateway between FlexNet and router. The BBSs can forward across the gateways as TheNET. Would a router located at that point fix the keyboard users do, but this requires more work than problem? Tadd said yes, but it would be cheaper to con- usual to keep up with where the gateways are. Areas vert the site to FlexNet and move the problem up the that are contiguous TheNET and areas that are con- line to the next guy! Lots of laughs. Dana asked if we tiguous FlexNet are easy to traverse by all software. could unplug the gateway support router and hand it to Don related how TCP/IP didn’t work well with TheNET the next guy to use. 10 years ago and how eventually a TheNET module was added to the NOS software to work across TheNET. Don Jim asked if FlexNet and NOS could be run in the said that there was a software drivers available for NOS same machine. John, CIK, said the answer was yes, for that support FlexNet in the same way. It is called some kinds of NOS. The NOS would have to support WAMPES. It is reasonable to move on to FlexNet and DPMI. But then you have a problem of reliability be- accept that some adjustments will need to be made to cause if either software crashes, it takes down both ser- some of the applications that use our network. vices. JNOS runs DPMI. The answer to this problem is to not run the NOS router on a mountaintop node. Jim Dana asked about the people who were using the vari- could do a dedicated link from his node site to a location ous software packages with the existing TheNET net- that can house the NOS router computer. work. These people will (and have) think that the net- work is broken. The answer is that those people that Dana reviews that the answer is that Jim could have want to make the most of packet radio will pay atten- the BBS and router at the same location. Jim says that tion to the news sources on the subject. These include he already has that at his house. the NEDA Quarterly. We’ll also have to do some dog Don says that we should vote the motion. Don says and pony shows at various clubs to promote these that if anything new comes along with TheNET that we changes. should still report it and cover it and we should com- Several people commented that there will be a media pletely document FlexNet as necessary to support NEDA circus regardless of what we do. Tadd asked what the members who want to implement it. various states do when they are about to totally disrupt The vote was called. The vote was unanimous. existing highways in order to make improvements. “Your Tax Dollars At Work”, “We’re working harder to make Web Page Stuff your commute easier” etc… “Temporary Inconvenience, Since last time that we met the domain Permanent Improvement”. packetnetwork.org has been registered. http:// Bob, BJG, told us that WAMPES is built into WNOS www.packetnetwork.org/neda points at the NEDA web and HRL miniNOS. It puts a router into the NOS that page. Tadd says that northeastdigital.org is available if is exactly like FlexNet. The other driver is IPCAM that the club wants to get it. Tadd said that we have an offer makes IP look like text for passage across FlexNet. from an ISP in Washington State that has NEDA life MFNOS has IPCAM in it now. It also has some new member N0NDO as a partner. John, N0NDO has of- timing changes. It now works. MFNOS does not have a fered that if we are interested in having a computer at FlexNet router in it yet. his ISP (www.conflux.net), that he will put it to his part- ners to host it for free. He doesn’t see that this would be Jim summarizes that as he understands it now, Flex- any problem. Tadd is thinking of doing this and is look- Net will build a general purpose packet network and all ing for help. Jim asked if there is a problem with where of the known servers that use TheNET can be adapted the web page is currently located. Burt said no. Tadd to use FlexNet, which will do a better job for all of us. said that the only problem that he has is that once in a The group discussed Iprouters and what it would take while he gets into the mood to do some web page devel- to create the analog of a X1J IP router for FlexNet. Don opment and he’d like to help once in a while. He said pointed out that FlexNet does not care about the data he’s not willing to commit to the job of web master and carries. Some sites would need to have routers located Burt likes the job anyway. Burt says that sometimes he at them. Some already have them. doesn’t feel like working on the page but is interested in staying with the job. Dana asks if we are providing pointers to the people Tadd asks if it is ok to leave his URL who are supplying the technology needed by server us- http://www.packetnetwork.org/neda ers in order that they convert to using FlexNet. The pointing at the NEDA web page in the mean time and answer is probably no but we’re working on it as fast as if we can list that URL as NEDA’s official web address. we get questions. All agreed.

N.E.D.A. Quarterly v5.2 page 56 Tadd asks that the pointers to his web page be moved Jim pointed out that Bob QBQ has a bulk rate mail- further up the NEDA page. Burt says he’ll do that. He ing house. Also WB1DSW had one in Nashua. says that he likes web pages that load fast. Don calls the vote. All approved Dunning Letter Meeting Location Don says that the Dunning letter seems to have got- Tadd suggested that we should have the next meet- ten us at least 50 members by spending $600. We don’t ing in Niagara Falls on Sunday the 8th of August or Sun- have the numbers. (ed: get the numbers) day the 15th of August. John MKH pointed out that there Tadd says that we have to double our membership was an Amateur Radio event in his area that would in- from what we have now. This would be easy if just the terfere with that date. Most decided that Niagara Falls sysops in the region joined the organization. was too far away regardless of it’s alternative entertain- ment opportunities. The group decided that Utica would Tadd says we could target those sysops with free maga- be the ideal place for the next meeting. John MKH vol- zines and pleas. unteered to host it in Syracuse if a host for the Utica Don suggests that we expand the coverage of the club location didn’t pan out. We decided that we’d figure out by getting names from outside the region. a place over the next couple of weeks.

Tadd suggests that we allocate some money to send NEDA promotions/hamfests free copies of the newsletter to sysops. Tadd asks that everybody that can do some hamfest Don said he’d rather specify a number of quarterlies. going. He said that those who wanted to lead such a Tadd said we could pull callsigns off of our maps and move could contact our membership director and get lists send to those people. John CIK suggested that it would of local members’ packet mail and e-mail addresses to be worth doing that just to get out of the first class cat- contact in order to help run a booth. egory. Constitutional Change Requests Don suggests that we just do this quarter by quarter. Tadd submitted two constitutional change requests. Burt suggests that we just send enough to make it bulk Both are to be published in the Quarterly to be voted on rate. Don suggests we just try it for one quarter. Dana at the next meeting. One is to remove the Grounds for agrees. Dissolution of the club, the other is to generalize the Another discussion went on about bulk rate pricing. periodical requirements of the club. Burt said that the way we handled bulk rate in the Board Member Alternates past was by supplying a diskette of labels to a mailing Tadd asked that he be allowed to appoint WB2CIK, house and they charge some up front price. Jim asked John Papson, to be his alternate. No objections? John where the newsletter was printed. Tadd said that it is Tadd’s alternate. Jim MEA asked for VE2BMQ to be was at Budget Printing in Amherst New Hampshire his alternate (across the street from Tadd’s office). Dana said that he’d like to see this done a particular Oswego County Hazmat way in order to escape being accused of playing favor- A letter was read to the group regarding the Oswego ites. He says we should pick a specific area and take all County drill held May 1, 1999. The drill went well. The of the node ops and sysops who are not members and letter is to be reproduced in the Quarterly. send them all a invitation to the club. This time that invitation should include a Quarterly #5.2 when it is Meeting Adjourned mailed. (ed: KB2TNR suggests that we preceed the de- livery of the free magazine by a post card that announces the free magazine) Dana made a motion that the first area we do it to is the NY metro area. We could take it off the list that Bob BJG mails out. We should send Quarterly #5.2 as part of the bulk rate. Tadd seconds it. John CIK said that we have a list server that we should gather some callsigns from as well. Don says we should do that in the future. Dana’s idea is a good one for this time. Tadd suggests that John has a pile of Quarterly #5.1s that John could send to any people he think might join.

N.E.D.A. Quarterly v5.2 page 57 N.E.D.A. Quarterly v5.2 page 58 N.E.D.A. Quarterly v5.2 page 59 N.E.D.A. Quarterly v5.2 page 60 N.E.D.A. Quarterly v5.2 page 61 The second course is much harder to achieve. Packet to publish. This will be a short article on how to use a simply *cannot* compete with internet where things like FlexNet node to access local and distant services. I’ll email are concerned. Trying to position packet as an also include pointers to more info and phone numbers. internet competitor will simply result in packet as the What say? guest star in a painting titled “Still life with roadkill”! There’s one big clue out there in the packet world as to how it might be able to become a mass medium once more - packet clusters! These things (admittedly, some- times a pain in the ass!) continue to thrive and present an active face to the packet world! Why? Simple - They Don’t Compete With A Service That Exists On The Internet! Instead, these systems provide a fast, reliable notifi- cation service that cannot be easily found on the ‘net. They’re single purpose and are tailored to do one thing exceedingly well. There are a few would-be clusters on the ‘net, there’s IRC of course, but they can’t compete with packet clusters on the radio. The trick then is to offer services that the internet cannot provide. Right now, packet offers a roadway but few unique destinations (read: applications) - the ‘net offers a faster roadway and many, many, many destina- tions. (This is why clusters have survived but packet BBSes are getting to be an endangered species!) If NEDA (and hams in general) ever want option #2 to be a possibility, by all means upgrade the road, BUT GIVE PEOPLE SOMETHING TO USE IT FOR! If we don’t, then all we’ll wind up with is a better road that’s still unused because there’s nowhere to go on it! We have packet clusters. What about data clusters? What about search engines for packet networks and data clusters? What about graphical interfaces for nodes? What about deviations from packet clusters into nodes running software for real time scheduling of contacts? Where are the county-hunter-specific node applications? This is our weakness - we have many road builders but few who are interested in building the malls with which to line them...... Whew! You got me started there, Tadd! This is my stance, anyway. Hey, clean it up, put in the next news- letter - see what people say - see if it stirs anyone up! I’ve found the same thing. This gives us some oppor- tunities too though. We can make things the way we like and then pro- mote it, rather than try to migrate an active audience which always has its disappointments. I’m thinking that we swap node softwares to FlexNet and then turn on DAMA on the user ports. This gives us a user port with no hidden transmitter syndrome. I’m going to write a how-to article for the local ham radio club newsletters

N.E.D.A. Quarterly v5.2 page 62 Continued from page 1 We who are promoters of Amateur Radio Packet Net- working need to figure out how to promote our hobby in order to build the network and make it more usefull. This is one man’s idea of how to promote packet net- working. The key, to me, is that packet radio networking is poorly understood. We node builders know what it can do. We also know how much fun it is to watch it work. The trick is to identify what it does and try to present it to the other hams. One of the things that overshaddows any such effort is that those hams who are interested in things digital are often on the internet. They see the internet as capable of everything packet radio is capable of, but better. I’m trying to build up a list of things that packet radio does that internet doesn’t do. Things packet radio does that internet doesn’t do or doesn’t do affordably • free 24 hour/day connect to hamshack • selective calling to any participating ham • 24 hour a day data transfer for weather and other projects • allows for exploration of the network • portable operation • add own services as desired • blinky light on TNC when you’ve got mail • operate with extremely low tech/inexpensive computer gear • it’s cheaper. Internet users think $240/year to $600/ year is reasonable to pay for a single digital channel. For that money you can do a lot with packet! Things a dedicated link based general purpose packet network does that some other packet schemes don’t do • high amount of data throughput per amateur voice channel used - multiple conversations on each channel • long distance communications • variety of data content (keyboard, automated data gathering (wx), bulletin traffic, dx spots) • multi-person chats • reliable selective calling

N.E.D.A. Quarterly v5.2 page 63 NEDA Constitution 1. Purpose of this Article e. The alter nate has full voting rights at boar d a. This article lays down the rules for operation of meetings in the absence of the board member the North East Digital Association. No other which he or she represents. N.E.D.A. document may change or replace the f. It is the r esponsibility of the boar d member and rules set down in the Constitution. The his or her alter nate to maintain reasonable Constitution may only be modified by the communication so that the alternate my act on procedures described herein. behalf of the boar d member in an inform e d 2. Of ficers manner. a. There are six Boar d of Dir ectors positions plus g. Any alter nate may act on behalf of any absent appointments and alternates. The board of board member, who’s alter nate is also absent, if dir ectors ar e elected for two year ter ms. Three necessary to provide a quorum. The member he of the dir ectors ar e elected annually. or she is originally designated to r epr esent must also be pr esent. The alter nate would have the 3. Appointments same voting rights as in (e.) above for the member a. Appointed positions include T reasurer, he or she is r epr esenting at the meeting. Membership Director, Board Member Alternates, Chairman of the Technical Committee and 5. Removal of a Person From Office or Revocation Network Regional Sysops. The Network Regional of Membership Privileges Sysops report to the Chair man of the Technical a. A petition for r emoval of a person from of fice or Committee and are considered members of the membership must be submitted in writing to the Technical Committee. board of directors with a minimum of four b. Other appointments may be made at the signatures of voting members. The petition must direction of the board of directors. These be presented at least two weeks before a appointments are made by the board of dir ectors. quarterly board meeting in which it is to be acted Only voting members may be appointed to a upon. The board of directors must vote on the committee chairmanship, board member petition at a quarterly boar d meeting. The alter nate or of fice position. Boar d members may document will be kept in the club archives also serve other appointed positions and unless removed and expunged at a later board appointees may serve multiple appointments. meeting. b. This person being removed is held as a removal- 4. Board Member Alternates pending member for one quarter and then is a. Each board member may appoint an alternate reviewed at the following quarterly board to represent him or her at board meetings in the meeting. This issue is then presented in the event that the boar d member is unable to attend. minutes in the Quarterly so that it may be b. The alter nate must be approved in advance by reviewed by all the membership and commented the board during a board meeting in which the on before the following quarterly boar d meeting. board member presenting the candidate for c. A person removed from membership is not alter nate is pr esent. The candidate must also eligible for voting membership unless the be present and agree, or fur nish written consent privilege is r estor ed by an act of the board of to serve. dir ectors at a later date. c. Appointment of an Alternate may be terminated 6. Membership at any board meeting under any one of the following conditions: a. Membership is open to all. Dues are at least 2 levels for individuals. One of these levels is • At the r equest of the board member the called Voting Membership. Voting membership alter nate r epr esents. is open to all except as defined under ‘Removal’ • At the r equest of the alter nate. above. • Using the same procedures as removal of a board member, (Article 5). d. The alter nate appointment is automatically cancelled when: • The alter nate is elected to a boar d position; • The alter nate is no longer a voting member; • The member the alternate r epr esents is no longer on the board.

N.E.D.A. Quarterly v5.2 page 64 7. Dues e. Board meetings must be held 3 times a year. a. Dues are paid to the Membership Director or The 3 board meetings should be held as close his designee who then forwards the funds to the as possible to any of the months of January, Treasurer. Dollar values of dues is set in the April, August or October. Additional boar d NEDA bylaws but the dues level for a Voting meetings may be called by the board of dir ectors member is $25 or greater. Dues are used to fund: with a vote of 4 board members. A board meeting • operating expenses for the club; is required in order to: • development costs for club products that • Estsblish or make changes to the annual facilitate network gr owth. budget. • documentation in the form of an Annual and • Spend club funds. Quarterly • Discipline a member; • documentation in the form of free technical • Change the appointment of chairman of any documentation distributed for the benefit of committee. packet networking. • Assignment or re-assignment or of any board • documentation in the form of free promotional member alternate. literatur e on NEDA and on packet • Appointment of any member to any of the networking. positions as detailed by Article 3. 8. Membership Privileges • Change the constitution or bylaws a. Voting Members receive the 4 copies of the NEDA • Appoint a chairman for the Board of Directors. Quarterly per year and a copy of the Annual each • Form or disband any committee. year. The Annual is delivered to the member at f. Actions which must occur at the board meeting renewal time (after r enewal) or at the anniversary include the r eading of a curr ent NEDA treasury of the member’s membership. report. This will be r ecorded in the minutes and b. Voting members are invited to attend the Boar d printed in the subsequent NEDA Quarterly. of Dir ectors meetings, run for of fice annually and 10 vote for of ficers by mailed ballot. 11. Elections c. Additional privileges ar e defined in the bylaws. a. Elections ar e held by mailed ballot after the 9. Board Meetings October Board of Directors Meeting. a. A Board of Directors Meeting is a physical Immediately after the October Board of Directors gathering of the board members. Meeting attendance of each member, over the b. A minimum of 4 directors or appr oved alter nates previous year’s board meetings, are tallied. Any must be present to open a board meeting. The voting member who has attended half of the board meetings must be announced via the year’s boar d meetings, and who are not alr eady NEDA Quarterly or via packet Email to every in the middle of a two year ter m ar e automatically voting member at least two weeks before the nominated and are listed on the ballot. The or der meeting. Each voting member is responsible for of appearance of the names of those listed on keeping the membership manager notified of the ballot shall be deter mined randomly. his/her current packet email address. If a b. This ballot is sent to all NEDA voting members quorum of board members or alternates is not complete with a self addressed stampled available to start the meeting, a new meeting envelope. The envelope also has a retur n must be scheduled and new announcements address label with a note stating that the r etur n must be sent out. address must be filled in for the ballot to be c. Board meeting should be held in dif fer ent cities counted. The ballot includes instructions that each time to allow voting members from different the voting member should order all of the listed NEDA regions to be able to attend board people in ascending order, 1 for first choice, 2 meetings and have access to administrative for second choice. This way the results will still proceedings. be meaningful if one or more nominated d. Board meetings may be attended by voting members are unavailable to fill the positions. members or those given special dispensation by The ballots ar e mailed to the club POBox and the board of directors or any approved by the then counted by one of the board members bylaws. whose term is not expiring this year. The balloting pr ocess, and the counting process must be operated with a process which maintains confidentiality of the ballots.

N.E.D.A. Quarterly v5.2 page 65 13. Filling Spots on the Board Due to Board Member Resignation c. The ballots must be mailed out to all NEDA a. If a boar d member resigns or is otherwise no voting members within two weeks of the board longer available to fulfill the r emainder of his or meeting. They must be retur ned to the club her ter m a new board member is selected to serve POBox within five weeks of the board meeting. until the next annual meeting. The new board Results ar e included in the Quarterly or are member is selected from those voting members mailed out separately to all members to arrive who were previously board members and who at least a week befor e the winter board meeting. ended their ter m as board member in good d. The results include the following statistics: standing. • total number of ballots sent; 14. Network Maps • total number of ballots r etur ned.; a. Network maps must be maintained and are • list of all nominees; presented in the Quarterly. The maps must consist of at least the callsign, nodename, • list of the thr ee new board members; location (at least r elative), user access • and a list of nominees who abstained but who frequencies for AX.25 (if any) and backbone had a higher vote than the selected board connectivity for all NEDA network nodes. members. e. If three new board members are not chosen by 15. NEDA Quarterly this pr ocess then a board member may be a. The NEDA Quarterly is published within 60 days chosen by consensus of the founders and the after the quarterly boar d meeting. The Quarterly existing boar d from those voting members who is fully described in the bylaws but as a were previously board members and who ended minimum must include the minutes of the board their ter m as board member in good standing. meeting (including the tr easurer’s r eport), the If ther e still ar e not thr ee eligible new board network maps, and membership roster. members then the club must be dissolved. b. The board may delegate the task of production and mailing of the Quarterly but maintain the 12. Board Member Responsibilities responsibility. a. Board members or their alter nates must attend c. In the Fall edition of the Quarterly whatever the quarterly board meetings or obtain an results that ar e available fr om the annual alter nate to handle meetings the board member elections ar e printed. This may include the cannot attend. Failing to do so twice in a single nominees or the final results. year is gr ounds for removal from of fice. Boar d members or their alter nates are also obligated 16. NEDA Annual to attend additional board meetings called by a. The NEDA Annual is the current statement of verbal agreement by any four of the board NEDA packet network involvement. This members. includes user infor mation for usage of the NEDA b. Board members represent NEDA and are network as well as lessons in the technology obligated to carry out the NEDA Charter in needed to fulfill the goals of NEDA as stated in regards to dealings with other members and non- the charter. members. b. This document is delivered annually to each and c. The board of directors as a body are responsible every paid member of the club. This document for seeing that the NEDA Quarterly and the should be updated at least once annually to NEDA Annual are published on time. As these reflect the curr ent state of networking technology are the instruments of the club and as the NEDA in use by NEDA. Quarterly is the means by which the financial c. The Annual is the responsibility of all of the operations of the club are published to the board members. The board may delegate the membership, the paying membership has the task of pr oduction and mailing of the Annual right to expect these documents. but maintain the responsibility.

N.E.D.A. Quarterly v5.2 page 66 17. Changes to the Constitution 19. Grounds for Dissolution a. Changes to the Constitution may only be made a. If the boar d of dir ectors doesn’t hold 3 board by the following pr ocess: meetings during the year or if the club is unable b. At a regularly scheduled quarterly Board of to hold elections or ther e wer e not thr ee eligible Directors meeting a proposal for a change is and willing candidates or if the Quarterly in at submitted in printed or typed for m (8 copies) to least it’s minimum for m) isn’t deliver ed on time each of the Dir ectors, to the editor and to the then the club must be dissolved. secr etary. The item must be presented in person 20. Dissolution of the Club by a NEDA voting member. a. After paying out any pending bills the tr easurer c. The for mat of the submission is in bulleted is directed to write a check for the r emainder of sections. The following sections must be the club tr easury to the American Cancer Society included: TITLE, PRESENTED, BY, BRIEF, and to close the all club bank accounts. The SPECIFICS, PURPOSE. The page is headed with name of the club (i.e. North East Digital “Constitutional Change Request”. TOPIC is Association) and it’s logo NEDA become the followed by one line which identifies the change property of the founders of the club, WA2WNI, request. PRESENTED is followed by the date of W A1TPP, KA2DEW, K1MEA, NQ1C, WA2VAM, the board meeting. BY is followed by the name KC3BQ, to do with as they wish. All paperwork and callsign of the author. BRIEF is followed by pertaining to software management of individual a single paragraph description of the change. nodes is deliver ed to the node/site managers. SPECIFICS is followed by a paragraph by © North East Digital Association 1989, 90, 91, 92, paragraph description of the changes including 93, 94, 95, 95, 96, 97 refer ence section and paragraph numbers. PURPOSE is followed by a justification for the change. A sample change is available fr om the club. d. The proposed change is entered into the minutes of the Boar d of Dir ectors meeting at which it is presented. Discussion may follow. No vote is taken at this time. e. At the following boar d meeting the change is brought up as old business and after discussion is either ratified or not. No change is made if a tie occurs. f. If a change is ratified then the new copy of the Constitution is printed in the following Quarterly in its entir ety. 18. Changes to Bylaws a. Changes to the bylaws may be made at a single board meeting with the vote of a majority of the board members present. If a tie occurs then no action is taken.

N.E.D.A. Quarterly v5.2 page 67 This is the NEDA membership roster as of 4/05/99 for all members expiring after 03/99. Please check your Membership Roster listing and renew if necessary. Thanks. callsign first name last name city st zip exp home BBS internet e-mail address NU1A Frank Finger Sutton Mills NH 199905 A call@bbs [email protected] AA2AC Vernon Siegel Clarence NY 200002 V AA2AC.#WNY.NY.USA.NA [email protected] VE1AIC Ron MacKay Cornwall PE 200002 A VE1AIC.PE.CAN.NOAM [email protected] KA2AON Brien Mathews Hannibal NY 199905 A N2UBH.#CNY.NY.USA.NA [email protected] KA1APA Joe Lelievre Sanford ME 200002 V WA1WOK.ME.USA.NA [email protected] AA9AW Walter Altus Onalaska WI 199911 A AA9AW.EN43KM.WI.USA.NA [email protected] K2BEH James Wenskus Rochester NY 200002 A WB2WXQ.#WNY.NY.USA.NA [email protected] K2BJG Robert Anderson Oakland NJ 200002 A wa2sna.nj.usa.noam [email protected] VE2BMQ Burton Lang Howick PQ 999999 V VE2FKB.#MTL.PQ.CAN.NA KC1BT Allan R. Machell Barre VT 200002 A KA2TCQ [email protected] K1BXG Joe Devlin Chester VT 199904 V WA1ZYX.NH.USA.NA [email protected] W2CG Marty Grozinski Flemington NJ 199910 A N2QAE.#NNJ.NJ.NOAM [email protected] WB2CIK John Papson Melville NY 200001 V WA2PNU.#NLI.NY.USA.NA [email protected] K3CKO Robert Swarm Corry PA 199912 A KB2OBB.#WNY.NY.USA.NOAM WB2COP Edward Kracum Middletown NJ 199912 V WB2COP.NJ.USA.NA [email protected] KB2CS Jack Abel Albany NY 999999 C KB2CS K1CSB Ray Feeley Southhampton MA 199912 V K1MEA.#WMA.MA.USA.NA [email protected] WB3CUF Wells Farr Cntrl Bridge NY 200002 V N2NQH-4 [email protected] N2CWY John Taylor Port Jervis NY 200001 V [email protected] [email protected] WA8DCE James Turrin Stow OH 200002 A WA8DCE.#NEOH.OH.USA.NOAM [email protected] N1DCO Donald Clark W Newbury VT 200006 V W1ET.NH.USA.NA [email protected] W9DDD John Koster Richardson TX 999999 C KA2DEW Tadd Torborg Amherst NH 999999 V ka2dew@wb1dsw [email protected] AA0DH Donald Waring Imperial MO 200001 V VE2DJE Richard Aubin Laval PQ 199906 A VE2TOY.PQ.CAN.NA [email protected] K4DMU Joseph Leitsch Lousiville KY 200102 A W4CN.KY.USA.NA [email protected] K2DN Fred Skinner Cortland NY 200002 A KB2FAF.#WNY.NY.USA.NA [email protected] VA2DOSLaval Laurentide Radio Club Fabreville QC 199911 A VE2CRL.#MTL.QC.NOAM [email protected] WB1DSW Herb Salls E Kingston NH 999999 V WB1DSW.NH.USA.NA WB3DTG Bob Unger Nazareth PA 200001 V WB3DTG@WB3DTG [email protected] WB2DWD Robert Seastream Manchester NH 199905 A WA1WOK.NH.USA.NA [email protected] N7ENT Peter Kallio Port Orchard WA 200002 A N7ENT.#WWA.WA.USA.NOAM [email protected] KC3ET Charles Gessner West Mifflin PA 200002 A KC3ET.#SWPA.PA.USA.NA [email protected] N3ETP Robert McGrath Salisbury MD 199902 A N3KNT.DE.USA.NA KY2F Fred Swiatlowski Oswego NY 199903 A WB2FAW Nicholas Dudish Herkimer NY 200002 A WA2TVE.NY [email protected] KA2FIQ Jim Morgan Ossining NY 200001 A WA2AWG.#ENY.NY.USA.NA WA2FNQ Jerry Mehrab Northport NY 199903 V WA2FNQ.#NLI.NY.USA.NOAM [email protected] N9FQF Gerald T. Kelley Jeffersonville IN 199911 A [email protected] N7FSP Scott Cronk Seattle WA 199905 A N7FSP.#SEA.#WWA.WA.USA.NA [email protected] W1FVQ Aime A. Beaudry Manchester NH 200001 A WA1WOK.NH.USA.NOAM [email protected] W1FYR Alan Merrill Gilsum NH 999999 V W1FYR.NH.USA.NA K1GQH Roger Guillemette Manchester NH 199904 A WA1WOK.NH.USA.NA [email protected] N1GUJ Peter Ferguson Granby MA 200003 A K1MEA.#WMA.MA.USA.NA [email protected] WA4GWG Bob Ewing Lincroft NJ 200002 V WB2COP.NJ.USA.NA [email protected] KC1HH Robert Merril Goffstown NH 200006 V WA1WOK.NH.USA.NA [email protected] KD1HL Maurice Richesson Northboro MA 200002 A WA1PHY.#EMA.MA.USA.NA [email protected] N1HO Bayard Coolidge Epsom NH 200102 A [email protected] N1HSM Stan Graziano Atkinson NH 200001 A WB1DSW.NH.USA.NA AD1I William Kenefick Chelsea VT 200002 A W1ET.NH.USA.NOAM [email protected] W2IH Isaac Hathaway Jr. Ithaca NY 199903 V WA2TVE.NY.USA.NA [email protected] N1IMO Bernard Peabody Hollis NH 200002 A N1FT.NH.USA.NA [email protected] N1IN William Sexton Richmond MA 200001 V [email protected] N2IRZ Don Rotolo River Vale NJ 200101 V WA2SNA.#NNJ.NJ.USA.NOAM [email protected] N1IUP Harold Read Berlin MA 200102 V [email protected] WD5IVD Greg Jones Denton TX 999999 C WB3IWY Gerald Engman Warren PA 200001 A KB2OBB.#WNY.NY.USA.NA [email protected] K2IZA Jack Aber Canisteo NY 200002 A WB2WXQ.#WNY.NY.USA.NA [email protected] NV2J Anthony Volino Elmira NY 200001 A N2JBA Edward Rubin Amenia NY 200002 A WA2PVV.NY.USA.NA [email protected] N3JBG John Filiatrault Russell PA 200002 A KB2OBB.#WNY.NY.USA.NA [email protected] N1JEO Joel Curneal Meriden CT 200004 V N1JEO.CT.USA.NA [email protected] W1JFP Calvin Stiles Hanover NH 999999 V W1JFP.NH.USA.NA [email protected] KA2JFU Richard Billings Oriskany Falls NY 199905 A WA2TVE.NY.USA.NOAM [email protected] N3JNS Dan Sekera Kingston PA 199904 A K3RLI.#EPA.PA.USA.NA N2JPQ Rick DuBrava Marathon NY 199903 A N2JPQ.#WNY.NY.USA.NOAM [email protected] KA2JXI Roger Ousterhout Ogdensburg NY 200002 V KA2JXI.#NNY.NY.USA.NA [email protected] G8KBB Dave Roberts Ipswich 999999 C W2KGY Cadet Amateur Radio Club West Point NY 999999 A W6KK Charles Spetnagel Alta Loma CA 200002 A W6TJ [email protected] W8KVK Ted Jacobson Athens OH 199912 A [email protected] N7KXI Jim Raehl Orem UT 199912 A N7KXI.UT.USA.NA [email protected] KB2KZB Floyd Harding Liverpool NY 200002 V KB2DIO.NY [email protected] N6KZB Mike Burton Temecula CA 200001 V KJ6VC.#SOCA.CA.USA.NA [email protected] WZ1L William G Poulin Amesbury MA 199910 V WZ1L.FN42MU.MA.USA.NA [email protected] N2LEX John Wright Albany NY 200002 A WA2UMX.#ENY.NY.USA.NA [email protected] W2LGF Reuben Merchant Henderson NV 200101 A WI7D.#SONEV.NV.USA.NA [email protected] KB2LML Alvah Haggett Champlain NY 200102 A KD2AJ.#NNY.NY.USA.NA [email protected] WA3LWR Robert Chimel Clarks Summit PA 199903 A KB3BHH.#EPA.PA.USA.NOAM [email protected] W1MBG John Keslo Pelham NH 199905 A [email protected] K1MEA James Wzorek Jr Easthampton MA 199912 V K1MEA.#FN32PG.MA.USA.NA [email protected] N2MKH John Driscoll Syracuse NY 200001 V N2UBH.#CNY.NY.USA.NA [email protected]

N.E.D.A. Quarterly v5.2 page 68 callsign first name last name city st zip exp home BBS internet e-mail address N1MR Mark Robinson MILTONBORO VT 199912 A KD2AJ.#ENY.NY.USA.NOAM WY2N James Brewer West Winfield NY 200002 A WA2TVE.NY.USA.NA [email protected] N0NDO John Painter Everett WA 999999 V N0NDO.#SEA.WA.USA.NA W1NMQ Joseph Boudreau Jr Fiskdale MA 199904 V KB1H.CT.USA.NA N2NSA John Romano Bronx NY 200003 A KB2VLX.#BRONX.NY.USA.NA [email protected] W7NTF Gary Kohtala Spanaway WA 200102 A w7ntf.#wwa.wa.usa.noam [email protected] WJ2O Dave Farnsworth McConnellsville NY 200002 V WB2BIN.NY.USA.NA [email protected] KD3OA John Tobias Mt Pleasant Mill PA 200002 V kd3oa.#sny.pa.usa.noam [email protected] N2OCW Lawrence Ashton Berkley Springs WV 199904 V N2OCW.WV.USA.NOAM [email protected] W1OQ Hartley Gardner Phoenix AZ 200002 A N7MRP.AZ.USA.NA [email protected] N5PBC Hans Zorn Spring TX 199910 A WA4IMZ.#SETX.TX.USA.NA [email protected] K0PFX Mel Whitten Bridgeton MO 200002 A k0pfx.#stl.mo.usa.noam [email protected] K7PO Donald Sturtevant Billings MT 200002 A N2POR Karl Hemker Jr. Albany NY 199907 V WA2UMX.#ENY.NY.USA.NA [email protected] WA1PTC Michael Staines Rochester NH 199912 V WA1WOK.NH.USA.NOAM [email protected] N1PVJ Harry J Coon Plymouth CT 199906 A KE2PW Rusty Seastrum Bemus Point NY 200002 A KA3SFC.#NWPA.PA.USA.NA KA9Q Phil Karn San Diego CA 999999 C N2QAE Arthur Martin Long Valley NJ 199905 V N2QAE.#NNJ.NJ.USA.NA WA3QAG Sanford Reedy Canton PA 200002 A KB3QW.#EPA.PA.USA.NA [email protected] WB2QBQ Robert Seger Altamont NY 199904 V WB2QBQ.#ENY.NY.USA.NA [email protected] WB2QJA Richard L. Benda White Plains NY 200002 V WB2QJA.#ENY.NY.USA.NA [email protected] W1QJH Gerard L LeBoeuf Hudson NH 199903 A [email protected] W2QLI James Dates Corning NY 199908 A WB2QBQ.#ENY.NY.USA.NA KA1QP Paul Chauvin Manchester NH 200001 A WA1WOK.NH.USA.NA [email protected] K2QQY John Darling Oswego NY 199905 A [email protected] KB3QV William Watts Hudson FL 199905 A N3KNT.DE.USA.NA W1RFP Blanchard Pratt Norwich VT 200002 A W1ET.NH.USA.NA [email protected] W2RH Richard Black Brockport NY 199904 A WB2WXQ.#WNY.NY.USA.NA [email protected] W1RK Ralph Karcher Gloucester MA 200002 A BBSUGM [email protected] VE2RM PtClairDorvl PQ 200003 A VE2FKB.PQ.CAN.NA [email protected] KB2RQB David R. Allen Dexter NY 200102 A KA1JXI.NY.USA.NOAM K1RSC John Johnston Rye NH 200001 A WB1DSW.NH.USA.NA WB2RUM Richard Schisler Rutland VT 200002 A WB2RUM.VT.USA.NA [email protected] K2RW Richard Wujciak Rockaway NJ 200002 A K2RW.#NNJ.NJ.USA.NA [email protected] K2SBQ Dexter Berwald Owego NY 199905 A WF2A.NY.USA.NA W2SN Edward P. Madison Brooklyn NY 200002 A WA2ZWP.#SI.NY.USA.NOAM [email protected] N2SNL Jack Tripp Pittsford NY 199902 A WB2WXQ.#WNY.NY.USA.NA [email protected] WA2SOK Irv Walter Palmyra NY 200002 A WA2WXQ.NY.USA.NA [email protected] KD1SQ Lee Reynolds Lempster NH 200002 A WA1ZYX [email protected] KD6STL Hugh Jamison Herlong CA 200002 A W3SYY Richard Bender Edensburg PA 200002 A W3SYY.#WPA.PA.USA.NA [email protected] N2TKX Stephen Auyer Liverpool NY 200002 A KB2DIO.#CNY.NY.USA.NOAM [email protected] KB2TNR Nancy Torborg Amherst NH 200002 V wb1dsw.nh [email protected] W2TTM Edward J O’Connor Morgan NJ 200002 A wb2cop [email protected] KA1TUZ Richard Doherty Newton Center MA 200002 V KA1TUZ.FN42JH.MA.USA.NA [email protected] WA2UKX William Reiter Pen Yan NY 200002 A kb2txp [email protected] KB2UQZ Edward A. Gutowski Corfu NY 199905 A KE2VW#WNY.NY.USA.NOAM [email protected] W1UU Peter Butler N Andover MA 200002 V W1UU.#EMA.MA.USA.NA [email protected] K2UZV Merrill E. Ryder Quoque NY 200002 A N4GAA [email protected] NY2V Fredrick V Adsit Syracuse NY 200002 A n2ubh.#cny.ny.us.na [email protected] N1VE Mike Stone Gilford NH 200002 A n1ve@wa1wok [email protected] N6VR Ray Benny Oak View CA 199906 A [email protected] WB2VUN R George Newton Skaneateles NY 200002 A WA2TVE.NY.USA.NA [email protected] N1WCU Marc Chauvin North Conway NH 199903 A [email protected] K2WG Wayne Gearing East Chatham NY 200002 A WA2UMX [email protected] WA2WNI Dana Jonas Valatie NY 200001 V WA2WNI.#ENY.NY.USA.NA [email protected] N8WQG LeRoy Anderson Crystal Falls MI 200002 A KC8VC.#UP.MI.USA.NOAM [email protected] NH6XO Robert Hlivak Kaneohe HI 200102 A [email protected] KB1YL James Robertson Rumford ME 200002 V W1TLC.ME.USA.NA [email protected] K1YPP Dennis Blanchard Hampstead NH 200002 A WB1DSW.NH.USA.NA [email protected] N2ZHS Al McChesney Scotia NY 200002 A WA2UMX.#ENY.NY.USA.NA [email protected]

I plotted a graph of the membership numbers as of April 2, 1999. This is the number of members we have in the database for each given expiration date. In other words, if we count “real members” as being people who expired after April 1, 1999, we have 131 “real members”, as of April 2, 1999. 9809=159 9903=138 9909=115 0003=25 0009=22 9810=155 9904=131 9910=115 0004=24 0010=22 The twelve 9811=155 9905=120 9911=115 0005=24 0011=22 members listed for 9812=152 9906=117 9912=115 0006=22 0012=22 0102 (Feb, 2001) 9901=147 9907=116 0001=84 0007=22 0101=19 and forward were 9902=145 9908=115 0002=28 0008=22 0102=12 all life Bob, WB2DWD, is NEDA’s membership director. members with expiration dates of Contact Bob via the PO Box, or e-mail at 999999. [email protected] or ICQ# 5769071 or via packet at [email protected] N.E.D.A. Quarterly v5.2 page 69 Membership Application

Welcome to NEDA Packet Radio This is the official or four times a year in various locations within the club's membership form for NEDA. area of interest. Meetings are open to the voting mem- bership. Some general information about NEDA: NEDA members sponsor general interest and special NEDA is a club formed in 1989 to promote packet ra- interest packet meetings throughout the region. NEDA's dio and to lead the development of a general purpose, focus is to publish information on packet radio and packet user-accessible wide area packet network. radio networking. NEDA does NOT fund the building, NEDA's primary area of interest includes the north operation nor maintenance of any packet networking fa- eastern United States, Quebec, Ontario and the Mari- cility. time Provinces of Canada. The membership/dues structure of NEDA is: NEDA publishes documents each year including offi- Associate with US address ...... $15 cial meeting minutes and articles of interest to packet Voting with US address ...... $25 networking. Voting and Associate members receive all Canadian Address surcharge ...... $5 NEDA publications. Non-US/Non-Canadian address surcharge . $15 NEDA's administration is based upon six directors, Upgrade to Voting, all countries ...... $10 alternates, and several appointees. The six directors of All membership rates are US funds only. Foreign the Board are elected for two year terms by Voting mem- members should send funds in a Postal or Bank bers, three are elected each year. The Board meets three Money Order in US funds. Thank you. This form is dated 19970707

Full service BBS at which you get your packet mail. Name: Example: WB2XYZ@WA2WNI.#ENY.NY.NOAM Call BBS Address: @

State or Home City: Province: Phone: ( )

Office Postal Code: Phone: ( )

Country: If a NEDA member gave you If not USA this form, what is their callsign?

e-mail County: address: Membership Dues are payable in US Membership desired - Check one Funds Voting Associate Make Checks payable to NEDA USA - $25/yr USA - $15/yr Address this form and all correspondence to: Canada - $30/yr Canada - $20/yr NEDA Foreign - $40/yr Foreign - $30/yr PO Box 563 Manchester NH 03105 Renewal Upgrade to Voting Information Membership, $10/yr Update only NEDA Representative Office Manager Membership Chairman Date: Date: Database: Amount: Amount: Check #: Check #: Intro package Delivered: Intro Package Mailed:

N.E.D.A. Quarterly v5.2 page 70 NEDA TheNET X1J Node Spec Parameter and Mode Specifications for a NEDA Compliant Node Revised: May 99

X-1J PARMS user #DPPL #HTSF Gate non FlexNet # Function port m’way way ideal gateway 1 Size of destination node table (A) 100 100 100 100 100 100 2 Minimum auto update quality 50 63 (B) 63 (B) 50 50 50 3 HDLC (radio port) default quality 0 0 (C) 0 (C) (D) 0 0 4 RS-232 (Crosslink) default quality 203 203 203 203 203 203 5 Initial obsolescence count 55555 5 6 Min Obs to broadcast 63263 6 7 Nodes broadcast interval 900 900 900 900 900 900 8 Initial time-to-live 13 13 13 30 13 13 9 Transport FRACK timeout (seconds) 200 200 200 600 200 200 10 Transport RETRY counter 11111 1 11 Transport acknowledgement delay 11111 1 12 Transport busy delay 180 180 180 180 180 180 *13 Transport window size 22222 2 *14 Transport overfill limit (frames) 44444 4 15 No-Activity time-out (seconds) 7200 7200 7200 7200 7200 7200 16 Persistence (n/256) 64 255 (E) (E) (E) 255 17 Slottime 20 1 20 20 20 1 18 FRACK (T1) time 4 1 1 (F) 9 1 *19 AX.25 windowsize (MAXFRAME) 1 (G) 1 1 1 1 (G) 1 20 AX.25 retries 10 10 10 10 10 10 *21 ACK (T2) time (L2 RESPTIME) 50 20 20 50 50 20 22 Active check (T3) 32000 32000 32000 32000 32000 32000 23 Digipeat 00000 1 24 Callsign validation 11101 1 25 Beacon mode control 20011 2 26 CQ broadcasts 10001 0

MODES # Function 1 RS-232 host mode 00000 0 2 CWID repeat period (seconds) 00000 0 3 CWID keyer speed (J) 66666 6 4 Selective nodes broadcast on ports 23332 3 5 RS-232 crosslink protocol 00000 0 6 Transmit keyup delay (TXD) 35 (H) (H) (H) 35 (H) 7 Full duplex (set as req’d) 0/1 0/1 0/1 0/1 0/1 0 8 Crosslink node b’cast interval (I) 400 410 425 440 450 410 9 Node b’cast algorithm port ctrl 33333 3 10 Beacon period (seconds) 600 600 600 600 600 300 11 Connect redirector (J) 00000 0 12 User message control flags (J) 27 19 19 19 19 19 13 Hash (#) node b’cast port control 33333 3 14 Extra alias 10000 0 15 Auto reconnect to node 10000 1 16 Control of slime trails. 22202 2 17 Digipeat up/downlink control 33330 0 FlexNet gateway * These parameters are good candidates for further experimental research. node nmemonic (A), (B), (C)... etc indicate notes following page. should be FXGxxx This drawing represents the node quality value for a single node as where xxx is local it propagates through several node hops. site abbrieviation Local node 1st neighbor 2nd neighbor 3rd neighbor broadcasts 256 = 128 = 81 = 51 <—User ports—> <–bkbn ports–>

N.E.D.A. Quarterly v5.2 page 71 Notes Regarding TheNET Parameter Spec

(see previous page) A/ It has been observed that this number is the H/ Transmit Keyup delay (TXD) for Dedicated User port: total number of ROUTES (up to three routes per point-to-point link, HTS free backbone and gate- Visible node port where live keyboard stations node) for all the nodes entered in the nodes way. This should be experimentally determined access the network and access resources over the table instead of just the number of nodes en- for each system and set to 1.25 times the lowest network. No station is heard on-channel by the user tered. If your node contains a large number of value that works. “real” alternate routes for many nodes, it may be port that transmits >3K/hour or 300K/month. No I/ Crosslink node broadcast interval. To prevent necessary to increase Parm #1 to 150 or higher node to node communications on HDLC Pside. No all the nodes on a node stack from trying to to prevent nodes being lost from the nodes node broadcasts on HDLC side. All node broad- broadcast simultaneously, Mode #8 should be table. The use of the alternate broadcast algo- casts on HDLC side ignored. set to a randomly selected number between rithm should prevent the proliferation of most of Non-ideal User port: 400 and 450 seconds. the “trivial” routes. Visible node port where we don’t want to do node to node communications but where there are other B/ The minimum auto update quality on hidden nodes or servers that this node can hear, on fre- nodes should be set to 63 since anything lower J/ These parameters do not affect the network. quency, that are causing HTS effects. This might be will not show at the nearest visible node anyway Set them according to your needs. a 2m user port that has gone to seed and sprouted and fills valuable buffer space in the node. automated stations that transmit more than 3K/ C/ The HDLC default quality on hidden nodes K/ Slime Trail Control Mode #16 hour: such as BBSs, NOS, KAnodes, Digipeaters, should be set to 0 to prevent unwanted or Slime Trail Control is made up of two bits. Bit#0=(1) HF gateways, APRS stations, DXClusters, weather intermittently received nodes from corrupting controls the display of slime trails in the node nodes, TheNET nodes, switches, etc. This port the network. However Parm #3=0 makes it list. Bit#1=(2) forces the node to ignore slime doesn’t broadcast nodes on HDLC or accept broad- difficult to get to the node following initial instal- trails. It has been observed that if both of these casts. lation or cold reset. If you burn parm #3=63 into bits are set (ie Mode#16=3) then Bit 0 overrides Gateway: the EPROM, you will be able to connect and Bit 1 and the effect is equivalent to setting sysop the node. You then set Parm #3=0 and Visible non-2 meter node port that acts as a Mode#16=1. If your intention is for the node to lock in the desired neighbor nodes at qual- hopping point between networks. Accepts node ignore slime trails, then you must enter ity=203 using the locked ROUTES command R broadcasts from both HDLC and RS-232 ports but Mode#16=2 only broadcasts itself. The purpose of this is to allow 0 + 203. two networks that have conflicting parameters to Note: Your initial parm #3 must be at least equal meet, allowing users and services to cross over, but to or greater than the minimum auto update quality If operating a BPQ switch with an application (BBS without having the node lists intermingle. We’re (Parm #2) for the route to be automatically added to or DXCluster) on a network with the slime trail never going to get a gateway defined exactly be- the Route list. control enabled, it is important that the applica- cause they will be customized for the various cir- tion first connect to the BPQ switch and then D/ HDLC default quality for gateway ports. This cumstances every time but this is the general idea. connect from there to a nearby visible node in value must be determined by the sysop such the network. The application usually will be set #HTSF HTS Free multiway: that the gateway port can fit into the neighboring with reduced quality to limit its node propaga- Hidden non-2 meter node (#node) for a protected network as well as being accessible to the tion to the immediate region. The lack of slime HTS free backbone. All stations on frequency follow NEDA-compliant network. 82 is a good value to trails will prevent direct connections from the NEDA parameters and all agree to not have more burn into the EPROM. No matter what you have application to more distant parts of the network. participants on frequency without further agree- the HDLC quality set to, the nodes that the Certain applications (such as FBB BBS’s) will ment. All HDLC routes are locked in. The default gateway hears over the radio will not be broad- force the BPQ to use the application callsign HDLC quality is set to 0. Persistence is set to 256/ cast over the RS-232 port because the mini- instead of the BPQ callsign in its L4 connec- (N-1) where N is # of transmitters. ACL is used to mum obsolescence to broadcast (parm #6) is tions out into the network. Connecting first to selectively enable participating neighbor nodes. set higher that the initial obsolescence. Do not the BPQ switch and then to a nearby visible Default ACL=7. lock in any nodes at >50 quality. node will avoid this problem. #DPPL Dedicated Point to Point Link: E/ Persistence for Hidden Transmitter Free L/ Alternate Nodes Broadcast Algorithm Port Con- Hidden non 2 meter node (#node) for a protected Multiway, Gateway and Non-Ideal user port. trol HTS free backbone with only two radios on channel This value must be calculated based on the within range of either site. Both stations follow intended aggressiveness of the port and the In the X1J documentation, the author noted that “it NEDA parameters. The HDLC neighbor station is number of transmitters on the frequency. Gen- made little sense to use it on the HDLC port but what locked in with ROUTE command at both sites. erally it is calculated as = 256/(N-1) where N is the heck, it is included for completeness. The only Default HDLC=0 ACL is used to selectively enable the maximum number of transmitters on the settings that made sense are 0 and 2.....Setting it to participating neighbors. Default ACL=7. channel. 1 or 3 will result in some pretty weird effects.” We theorized that the alternate algorithm would be F/ FRACK time for a gateway node. This must useful on any radio port that was SHARED. At first NOTES follow whatever standards are being used by we tried it on a digital repeater node radio port in a the rest of the stations on the radio channel. */ These NEDA recommended TheNET X1J pa- backbone and it worked fine. There were no ill rameters are designed to suit and standardize G/ MAXFRAME on user ports. To allow users an effects observed and it solved the problem of mul- MOST of our network node needs. However equal share of network access, the maximum tiple trivial routes generated in the repeater environ- they are just recommendations (not cast in node transmission to a user should not exceed ment. Later we also tried the alternate algorithm on stone) and can be adjusted to meet the needs two seconds. This corresponds to a regular Node radio port and did not observe any ill of SPECIAL circumstances. Don’t make the MAXFRAME=1 at 1200 bps, MAXFRAME=3 at effects. So while it would not seem to enhance a mistake of someone recently who thought that 4800 bps and MAXFRAME=4 or 6 at 9600 bps LAN or dedicated port, it does no harm and it does we were against full-duplex links because we data rates on the user ports. MAXFRAME on help a shared port. Therefore we recommend previously specified Mode 7=0 (full duplex off). dedicated network links is a controversial issue setting Mode#9=3 for all ports. If you choose to adjust these recommended and until this can be resolved through experi- parameters however, PLEASE make sure you mental research, NEDA continues to recom- know ALL the effects of the change. If you are mend network MAXFRAME=1. not fully aware of the node operation, ASK BEFORE MODIFYING.

N.E.D.A. Quarterly v5.2 page 72 Board Meeting Notice The May 2, 1999 NEDA Board Meeting will be held in northern NJ just off of the Garden State Parkway. Con- tact Bob Anderson, [email protected] for more in- formation and to RSVP. Alternatively, contact Don NEDA World Wide Web NEDA Rotolo, [email protected]. Home Page address

If you did not get a meeting notice via packet or e- URL = http://www.packetnetwork.org/neda mail a few weeks ago, make sure your information is correct in the roster. The invite was sent to all voting members whose expiration date was after February 1999. The invitations were sent out based on a database list- ing generated on April 1, 1999. –NEDA

N.E.D.A. Quarterly v5.2 page 73 The NEDA Quarterly is an official publication of the North East Digi- tal Association. This document is published after each Board meeting, generally three times per year. The Quarterly contains the Minutes of Board meetings, a Membership Roster, and other business matters of the Association as well as articles and information pertinent to the construction of Amateur Radio packet networks. NEDA is a nonprofit association formed for the purpose of promoting free-access general-purpose amateur radio packet networking. Paid membership at the time of this publication was approximately 150, and over 250 copies were distributed. Copyright © 1989->1999 North East Digital Association. This document may be reproduced in full for nonprofit purposes. This document may be used as a source for material, but NEDA must be credited and the NEDA Quarterly Volume 5 Num- ber 2 (July 1999) must be listed as the source document. NEDA's Mailing Address is: NEDA PO Box 563 Manchester, NH 03105-0563

N.E.D.A. Quarterly v5.2 page 74