Sunday, May 9, 2010

WinLink Failure


At the April 8th meeting of the Green Bay Mike & Key Club, Dennis, KC9OIS had an ARES report on a demonstration of a mobile Winlink station and it's emergency email capability.

The station was setup by Outagamie County ARES at the Paper Valley Hotel in conjunction with the WI ARES/RACES at the WI Governor’s Conference in March on Emergency Management....

While the demonstration was successful, John, NA9J and my (Dennis, KC9OIS's) messages never reached their destination successfully.


Paul, N5XMV writes:
I don't see the purpose of continuing to find new ways to use Ham Radio, whether for experimental, emergency, or commercial use, if the average Ham is not able to understand the full concept, build it, and implement it... What happens when it fails, reach over, and get a new piece out of a stockpile of new stuff??


Paul, N5XMV, makes a good point. Emergency communications weakest link is in the operator behind the equipment.

At the same time, exercises and demonstrations like this let you experience the mistakes and correct them before the time comes that you actually need to rely on it. My question is what is being done to prevent a repeat performance?

I don't know much about how the Winlink network works. But it seems to me that if "emergency" messages can be lost, this is Not a well thought out design.

Are outdated networking protocols are the basis of Winlink? It sound like a kludge of much how the old KA-nodes worked, and the hierarchical BBS message forwarding using a system of mail rewrites and mail exchangers.

Ouch!

RFC's for SMTP mail delivery ensure that the sender should get a return message in the event that a message is undeliverable to to a broken path route or server being down. If Winlink cannot at the very least support this, perhaps it's time to re-think the network topology.

Think "real-time network" .... it can be done. In the 90's we here in Wisconsin could test the deliverability of a message with the ping command.

traceroute is another good network debugging tool.

Here is an old reference on Emergency Operations and Packet Radio
http://www.qsl.net/kb9mwr/wapr/wiscnet/part31.html

Here is some interesting history on Winlink.

July, 2003: In cooperation with its partnership with Homeland Security & at their recommendation, the ARRL Board sought to provide a Nationwide digital system to enhance the communications capability of the Amateur Radio Emergency Service (ARES) There are situations, the Board said, when ARES "must have the capability to pass digital traffic across the Nation quickly and accurately.

ARRL Resources Volunteer Committee determined that the new network should: provide rapid transfer of emergency traffic between sections; use available and future digital modes, interface with commercial communications systems such as conventional telephone, cellular telephone, and the Internet, etc., have speed, performance and accuracy.

The digital network will provide a value-added service for ARES and will continue to be viewed very positively by our served agencies," the committee said in its report. "This allows ARES to be viewed as modern and necessary instead of antiquated and invasive."

A quote from the former FCC Director of Engineering and Technology:

In the past, hams have adopted more spectrally efficient technologies - for example, by migrating from double-sideband amplitude modulation to single-sideband modulation and, more recently, by shifting to more efficient modulation for digital modes. I would urge you to continue shifting towards more spectrally efficient communications techniques - especially digital techniques. Such a shift has a number of benefits:

First of all, it demonstrates to policymakers and regulators that you are good stewards of the public's airwaves even without direct economic incentives.

Second, by using what you have efficiently, it strengthens your case when you need to ask for additional spectrum.

Third, by allowing more users to access the available allocations simultaneously, it improves the amateur experience and ultimately increases the attractiveness of the service to new and old users alike.

Fourth, it provides the opportunity or "headroom" for increases in data rates to more closely match those available on wire line networks and, in the future, on commercial wireless networks as well.

Fifth, as the rest of the telecommunications world makes the transition to digital techniques - and there are very few exceptions to that trend - the amateur service will look antiquated if it is not making progress in that direction as well. So looking to the future of the amateur radio service in the new century, I would urge you to continue your traditional role in public service by being prepared for and providing communications in times of emergencies, conducting experiments, providing training in radio communications, and encouraging international comity. But I would also urge you to focus particular attention -- for the reasons I just mentioned -- on experimentation with digital techniques."


So far the basis of Winlink on a county/state level seems to rely on 30 year old packet radio technology.

For example it will take over and hour to send a 540 KB .xls file of names... or in just a half an hour you might be able to send a 270 KB attachment at 1200 baud.

So even if the demonstration messages would have made it, Winlink still appears as a failure to me in that regard.

7 comments:

goody said...

In my opinion Winlink has problems on several levels. You mention SMTP. It seems the designers of Winlink ignored the enormous suite of time-proven RFC standards and chose to reinvent the wheel. It's like it was developed in a complete vacuum. Second, the network depends on a centralized system. The system is redundant, but it all runs on the same code so it could all come down from the same bug or exploit. Third, the choice of PACTOR-III as a primary mode was just a bad idea. It's closed, proprietary, and expensive to implement when compared with other choices. Even the name "Winlink 2000" itself speaks of two inherent design shortcomings.

Another big issue is that development and administration is controlled by a small group of people. Amateurs can't just create an ad hoc Winlink network or join the network quickly in the event of a disaster.

I'm disappointed that Winlink is the best we could come up with and an alternative like PSKmail which is open source and de-centralized can't seem to gain ground. In my opinion our wireless email network "standard" is the biggest amateur radio failure and missed opportunity, ever.

Anonymous said...

Winlink represents a lot of work by volunteers, but it's still a kludge. ARES uses it because it works (sorta), and because, well, there isn't an alternative that works better.

Building a better mousetrap for digital and/or emergency communications requires people who know math and computer science.

Amateur radio doesn't do much, if anything, to reach those types of people. Want something better? Take some of the money that goes into ARES repeaters that no one talks on, and put it into a summer research project for undergrads at a university.

The result can't be worse than a complicated email system that doesn't quite work even when there's no emergency.

kb9mwr said...

Precisely what I have been trying to point out here. "Ham radio doesn't do much to reach the types of people that can in-turn give back to the hobby."

I agree 100% that there needs to be grants or something available for the people who do and can build a better mouse trap, as you put it.

TAPR sort of fits that bill, but not really.

Again here is an example of something more logical in my mind than Winlink:

http://kb9mwr.blogspot.com/2010/01/emcomm-amsats-geostationary-satellite.html

Anonymous said...

Funny, I received the messages at the Governors conference. So WINLINK made it thru. We even digi-peated the messages to my home QTH and one other local station.

Now for the REST of the story. The Governors conference was attend by several hundred EM's, police, Fire and various other Emergency personnel and many volunteer agencies.

We set up our ARES trailer and received 17 messages for 10 of the Emergency Managers. All messages were delivered, to EM's at the conference we could find. We had a table in the Vendor's area and as EM's came by, we delivered the messages. That particular EM did not come by after messages came thru.

Paging the EM in such a conference would be distracting to the EM and other attendees.

There are numerous breakout sessions and side bar meetings at such a conference, and looking for someone in that scenario is distracting for all.

Under a EOC type situation, the message would of course make it to the appropriate individual(s) or the position station.

So Winlink did not fail. The message, fully readable, made it to the location.

Winlink is not perfect, and no system is. What makes it useful is:

1. Most people with minimum training can have a station set up in a short period of time with minimum expense. We cannot always expect our most experienced person(s) to be available. A "newbie" might be the first to respond. No use having her/him stand around waiting for expertise if he has what it takes for our very usable system.

2. Just about any emergency agency's personnel we would assist understands MS outlook/express email. Familiar procedures is important in a stressful situation, instead of trying to figure out various agencies forms to say the same thing we or they can put in email format. Remember, our objective is to assist served agencies, not make them an expert in our procedures. We can even turn the keyboard over to them and let them type it out and stand by to review and send. This way they can say want they need to and we can keep it legal.

I checked out your link to the satellite system. Read about it elsewhere a while back. Should we wait another 10 years for this to get launched?

We (ARES) continues to promote what works now and we do look to future technology that is useful for both the fun side and EMCOMM use in amateur radio. "Layered" systems to cover and compliment gaps is always the best. But for now, I cannot just shrug my shoulders and ask our EM "can you hold off emergencies for 10 years?"

Just the facts.

Stan Piekarczyk
outagamie Co ARES EC
SMSgt USAF Retired

kb9mwr said...

I'm just reposting what Dennis reported at the meeting. Maybe the message made it later or something.
I'll be honest, I know very little about Winlink. I guess my main concern was that it sounded like you don't even get a bounce message if it's undeliverable, so you just have to assume it made it? Perhaps you can clarify that. Sound like operator error since Winlink is new to us Brown County guys.

ws4e said...

>well, there isn't an alternative that
>works better.

Not true.

NBEMS(Narrow Band Emergency Messaging System) designed by W1HKJ (author of fldigi), is a much better and 100% open tool for handling digital traffic.


http://www.w1hkj.com/NBEMS/


ARES and RACES groups need to get off the Winlink proprietary PactorIII bandwagon.

Anonymous said...

Winlink does not PING per say. However it will store the message on the email address for about 30 days. Not sure of exact number of days. Best bet, receiving net should have (yes me) responded with could not find the EM)

Plus you have the option of addressing to numerous addressee's email addresses. So with multiple "targets" you can send to hams, direct to EM's and others you have addresses for (showing the importance of good working relations-not just popping in when the stuff hits the fan. Thus the blackberry's and other devices can get info from Winlink.

I checked out WS4E's link on the NBEMS system. I think it will be worth trying, but sounds like winlink lite on 2 meter sideband. But I'll give anything that does not get me arrested or divorced a try.

Steve, I have 2 meter SSB, we are about 40 miles apart, want to try NBEMS in early July? ( I have full sked with work for next 6 weeks).
If you want to work EMCOMM either fully ARES dedicated or a walk in- think Layers- Commercial means (served agencies fax/phone and if asked radio. Then Amateur Digital, Voice, C/W, Boy Scout flags.

or as the Russian Tankers would say: If you can't be a tank, be a cannon, if you can't be a cannon, be a pill box, if you can't be a pill box, be an infantry man. When you can't be, well you get the picture.

You can get my email address from the Brown Co ARES folks.

Stan K9STN
etc etc