[aprssig] The final WIDE2-1,WIDE1-1 solution!
greg at clubnet.net
Fri Apr 1 19:28:43 CST 2005
If I understand this correctly, this configuration will create a
problem in areas such as metro LA where you have one-hop digis (unless
you have preemptive digipeating more on that in another message).
We instituted one-hop digis in metro LA about two months ago with a
standard path of WIDE2-2. These digis allow one-hop digipeating based
on receiving a WIDE1-1,WIDE2-2, WIDE2-1 or WIDE3-3 digipath. Two of
our 4 digis also respond to RELAY. These two run UIDIGI firmware, and
thus don't have the duplicate packet problem previously mentioned.
With this change, we have been seeing improved network performance.
Unfortunately there are a significant number of stations that still
unnecessarily use RELAY as their first hop. A path of RELAY,WIDE2-2
undermines the one-hop concept. A single packet that may have been
digipeated once without the RELAY now may be 4 digipeated times (and
that's just counting the high-level metro area digis).
As I see it, if these same stations use a path of WIDE1-1,WIDE2-1, the
same problem will exist (again, without preemptive digipeating). You
will automatically get two hops out of this path. Hopefully I've
overlooked something here.
My other concern relates to what Phil mentioned earlier. There is a
finite number of people that will be savvy enough to make such a
change, but as the configuration becomes more complex, fewer people
will understand and will probably default to the way they had it
before. Others may have their configs semi-hardcoded into their
trackers and don't want to bother making the changes. Before the
latest changes, we hadn't supported RELAY and WIDE for 5 years, yet
people still used them.
Overall, I think this latest change can help in many situations. I
just wanted to point out a couple issues that I had with it. I guess
there will never be a 100% solution. Hopefully this gets us a little
On Apr 1, 2005, at 12:39 PM, Robert Bruninga wrote:
> And now the New n-N Paradigm is complete!
> 1) WIDEn-N works everywhere for small values of N
> 2) The RELAY-DUPE and WIDE-DUPE problems are solved
> 3) We can still use fill-in digis
> and APRS paths are totally simplified to simply WIDEn-N
> and 2-2 for mobiles and the 1-1,2-1 options for those mobiles
> that think they need it.. Wow , what a day!
>>>> bruninga at usna.edu 4/1/05 3:09:49 PM >>>
> Wes, I think you got it!
> 1) We replace RELAY stations with WIDE1-1 stations.
> 2) The general mobile path is WIDE2-2 (or 3-3 if one insists)
> 3) for those indifviduals that insist they need a fill-in-digi can
> use either of these paths:
> WIDE1-1,WIDE2-2 (in areas that can handle 3 hops)
> WIDE1-1,WIDE2-1 (in WIDE2-2 areas)
> This completely eliminates the RELAY-DUPE-multiplication
> problem and is fully compatible with the New n-N Paradigm.
> But may not work in LA or areas with WIDE2-2-only digis,
> though I think we can fix that too...
> Is this workable?
> In those areas that can afford 3 hops AND need
> fill-in-digis, then mobiles in those areas can use
>>>> Wes Johnston <aprs at kd4rdb.com> 4/1/05 2:47:43 PM >>>
> I think I have it!
> A mobile station running W1-1,W2-2 gives 3 hops minimum becuase
> W1-1,W1-1 would cause every home station in town to TX.... But there
> a real case for running less than 3 hops... If a mobile wants to limit
> his path to fewer than 3 hops, he can "dummy up" a reduced Wn-n count
> using W2-1 instead of W2-2.
> I still think RELAY stations switching to W1-1 is a good idea.
> We could have three recommended paths for mobiles:
> W2-2 ->two hops off digis, if you can hit 1st digi direct
> W1-1,W2-1 ->one hop off house or digi, one more hop off digi
> W1-1,W2-2 ->one hop off house or digi, two more hops off digis
> In other words, if you want one hop, we have a solution.... WIDE1-1
> If you want two hops, we have a solution W1-1,W2-1
> If you want three hops, we have a solution W1-1,W2-2
> if you want four hops, we have a solution W1-1,W3-3
> And.... W2-2 still works everywhere because all digipeaters will
> to it.
> Robert Bruninga wrote:
>> ARGH! Ur right. Back to square one. Too bad the
>> WIDE0-1 idea didnt work... Bummer Bob
>>>>> "Curt, WE7U" <archer at eskimo.com> 4/1/05 1:15:29 PM >>>
>> On Fri, 1 Apr 2005, Robert Bruninga wrote:
>>> VIA WIDE2-1,WIDE1-1
>>> Will this work?
>> Close, but no cigar. If someone were using a path of
>> "WIDE2-1,WIDE2-2" and had "house" digi's configured to respond to
>> that first one, you'd end up with:
>> WIDE2-1*,WIDE2-2 (house digi relayed it)
>> WIDE2-1*,WIDE2-1 (hilltop digi relayed it)
>> WIDE2-1*,WIDE2-1* (all house digi's that could hear the relay)
>> Curt, WE7U. APRS Client Comparisons: http://www.eskimo.com/~archer
>> "Lotto: A tax on people who are bad at math." -- unknown
>> "Windows: Microsoft's tax on computer illiterates." -- WE7U
>> "The world DOES revolve around me: I picked the coordinate system!"
>> aprssig mailing list
>> aprssig at lists.tapr.org
> aprssig mailing list
> aprssig at lists.tapr.org
> aprssig mailing list
> aprssig at lists.tapr.org
More information about the aprssig