[ax25-layer2] Still lost

[ax25 at pe1rdw.demon.nl]

Quoting James Ewen  <jewen at shaw.ca>:

> ax25 at pe1rdw.demon.nl wrote:
>
>> If you can install a single VHF antenna you can also install a dual
>> or triple band antenna thus creating the needed posibilety for a
>> backbone link.
>
> When you are given access to a commercial VHF antenna, it's pretty 
> hard to say thank you, but change that out to a different antenna 
> please!
>
I was under the impression you used your own antenna on a commercial tower.

>>> This new concept will still
>>> fail with digipeaters sitting where they can hear 300 users.
>>>
>> Not realy, for a 300 user lan you just need to upgrade the speed and
>> eliminate the hidden node problem, that can be done with dama or
>> digital repeater.
>
> How are you going to upgrade the speed of a 1200 baud TNC? Don't tell 
> me to change it to 9600 or 19.2K... it would no longer be a 1200 baud 
> TNC then.
>
> There is a finite amount of time available, and sooner or later the 
> channel will be full. The only way to create more time is to decrease 
> the user load by reducing the number of local users.
>
As long as you plan to be stuck in 1k2 land you will keep losing users and the
problem will solve itself, it might sound harsh but in todays world 1k2 packet
is like driving an donky, you might get there in the end but you'll get
frustrated doing it.

>>> For sparsely populated areas, which usually also means no access to
>>> high speed internet, this system will mean that packet radio will
>>> disappear.
>>
>> very doubtfull as RF linking has proven to work in europe, internet
>> linking has
>> for a very long time been iligal here so we had no easy way out.
>
> Great, send all the extra equipment and money you have laying around 
> so we can build this rf linking network. We'll also need enough money 
> to build the towers as well... Hey, who's paying for power?
>
What you need are sysops willing to spend their own money to build 
something to
be proud of, that is how it is done here and the results are very much worth
it.
The only thing donated is often a small place in a very hot boiler room in a
highrise building or sharing with a cellphone tower but usualy that is still
payed at a persentage of commercial room.

> Our local area network is currently 2 or 3 hops deep, and we would 
> like to increase coverage along the major highways in the area. The 
> network has about 20 local users, and we see <200 packets per hour.
>
> By restricting layer 2 to 2 hops, we will loose the areas that are 3 
> deep, and not be able to support the areas that we would like to 
> expand coverage into. It is tough enough to be able to get 
> digipeaters into the area, let alone try and set up a second link 
> frequency to create a LAN with a single digipeater with zero local 
> users.
>
a 3 digihop 1k2 link will drop down to about 75 baud effective but only on a
very clear frequentie, with more users it would be faster to use psk31, on 2
meter it will likely have the same range as 1200 baud packet given the low s/n
ratio it needs.

> I don't intend to speak for everyone, but in my local area, we build 
> and operate networks on zero dollar budgets. Everything is donated, 
> from equipment right down to mooching power from the commercial tower 
> owners. While the optimal system as described previously is great, it 
> most likely would not be feasible here. Not with the limitations we 
> have.
>
wel unless the sysops are willing to pay for their own systems I guess 
you'll be
stuck in the dark ages, you get what you pay for and with zero budget 
you get a
nearly zero functionalety, if your layer 1 is stuck in the 80's then 
nothing of
layer 2 can improve it.
Wel not completly trough, you can get a slight improvement by switching 
from end
to end acknowlidgment to hop by hop acknowlidgment but you will still need to
invest in pc's or tnc's that can function as a node rather then digipeater.
Then again everything that is discused here will need a pc or microprocessor
board to drive the tnc anyway.

> If money and locations were no object, I'd toss all this slow speed 
> stuff away, and look into building a HSMM network instead.
>
> James
> VE6SRV

Guess we are lucky to have sysops that care enough to invest something into
their hobby or we would never have the network we have now.

73 de Andre PE1RDW