[aprssig] Lotsa ears, big mouth concept.

[Heikki Hannikainen]

On Tue, 13 Oct 2009, Robert Bruninga wrote:

>>> It almost sounds like you are trying to
>>> replace the principal of having fill-in
>>> digis that respond to WIDE1-1 at low
>>> elevations and high WIDEn-N digis...
>>
>> This idea was born dead.
>
> Not so fast.  It was the only way in the USA to allow the
> construction of FILL-IN digis using the hundreds of thousands of
> existing available TNC's already in hand and not require the use
> of a $280 specialized new-N APRS TNC.
>
>> Our scheme, where everybody uses WIDE2-2,
>> and fill-in-digis respond only to packets
>> from a certain geographical region, works far better.
>
> Yes, it is slightly better because the length of each packet is
> shorter without having to carry the initial WIDE1-1.  And it is
> the best solution, if people are willing to buy all-new hardware
> for fill-in digis.  It is good to hear that your country is able
> to do this.

Not all-new. Just the $70-$80 TNC, or use a Linux solution on a little 
surplus PC with a CompactFlash card instead of a hard drive. Everything 
else can be surplus, too.

We don't really do much fill-in digis at all, which is rather fortunate 
for the network. We don't have any significant mountains in here, and we 
are happy to have very cost-effective access to some commercial antenna 
sites (see the second post on http://oh7lzb.blogspot.com/), thanks to a 
couple of individuals who have negotiated the deals with a couple of 
companies owning a lot of this kind of infrastructure. Public service, 
ARES, and hams employed inside those organisations help. Most digis 
service a pretty large area. And that's OK, since the user base isn't so 
huge.

A bunch of other digis are in other high sites, hilltop amateur towers, 
water towers, etc. Because there are no proper mountains, they cover large 
areas, too.

> We should be able to slowly migrate that direction now that many
> innexpensive new-N TNC digipeaters are available.  But for
> years, users will still have to use the path of WIDE1-1 until
> ALL digipeaters are converted if they want the best chance of
> getting digipeaterd..

Using WIDE1-1 does not really give any benefit over here. It does work, 
though (it's not deprecated), and we supported the WIDE1-1,WIDE2-1 new-N 
spec in the IARU-Region 1 meetings since it is compatible with what we use 
now, and having the same settings work worldwide, out of the box in 
shipped products, is a Very Good Thing.

The $70-$80 USD cost for a modern new-N capable TNC isn't really 
significant, when compared to the amount of WORK that goes in the digi 
set-up. And the cost the electricity, license, and potentially radio + 
antenna. We use mostly surplus antennas and radios, though. A major 
portion of the digipeaters here use the surplus HaMDR radio modem 
(http://wiki.ham.fi/Hamdr) which has a 5-watt VHF radio and a rather 
powerful processor and 128k of flash + 64k of RAM. And a very nice 
open-source ham firmware with a good smart digipeater. Too bad the supply 
of those modems has dried up.

We do also have a big bunch of digipeaters running TNC2 clones and UIDIGI. 
It would be very good to get those replaced with smarter ones with things 
like long path cutting (to drop those WIDE7-5,TRACE7-7 packets). The 
replacement needs to be open-source, so that it doesn't end up going the 
way of UIDIGI.

   - Hessu