30 Jul
2021
30 Jul
'21
8:08 a.m.
Dear Antonios,
Am 30. Jul 2021, um 03:18:53 schrieb Antonios Chariton (daknob) via 44Net:
Your two statements are again contradictory.
One point of view, taken by some amateurs back in the 90s was that
the communication between two Amateurs should ONLY occur via radio
waves. In other words, a "radio only" network should, as the name
implies, not be using any other mode of communication, for example
an IP tunnel THROUGH the Internet. That is a different, much more
extreme position than defining some sort of an "Intranet" for amateur
radio usage, which could include internet links, tunneled or not,
as long as both end users are licensed amateurs. This is another
example why TAC should be very careful with wordings, and maybe provide
stringent definitions of some key terms.
That is simply not true, since there is no homogenous radio network.
You (the TAC), made the claim and used it as a foundation
of your proposal. It is your onus to prove it or at least to plausibly
support it.
Sigh. This was not a yes/no question, but rather a question how far you
want to turn back the wheels of time for 44.128. Pre-2012, individual
networks did not have direct BGP announcements, but had connectivity
(or lets say the option to connect) through mirrorshades(.ucsd.edu)
which announced 44/8. I can bear witness that this worked that way at
least from 1995, but likely much longer.
Radio network access has at all times been set according to the
gateway's jurisdiction and ham radio regulations, namely
third party traffic. In most of Europe, any non-44 IP frame over an
amateur radio link was (and likely still is) illegal. Not so in
the US under third party traffic rules, albeit the situation has
become considerably murkier with the advent of encryption (HTTPS, SSH).
Again, please observe the careful distinction of "communicating
with the open internet" vs. "communicating through the internet"
You *never* can be certain. How many QSOs have you made where you
asked first for a copy of your partners license to be faxed ?
The problem is the problem of those allowing the traffic to cross
to a radio operated under amateur radio regulations, i.e. the operators
of that radio. Pre-2019, most of them would in the past have accepted
access to a sufficiently routed 44/8 IP as sufficient authentication,
although of course it is not perfect. Others required authentication
though a login on the gateway.
Hello Mario, please find my answers below:
Thank you very much for those insights. Although the image this paints
is very bleak indeed.
We do not want to limit the hardware, but we would
like to have an IP version of the frequency plan: 44.128/10, however you connect to it, is
the radio amateur band, and 44.0/10 is the commercial 5G band or the ISM band of WiFi. One
is for people that are licensed, and the other is simply the Internet that anyone can use.
Thats unbelivable, I can but hope that something is lost in translation.
Let me rephrase that:
The proposal is nothing short of ripping out 44.0/10 from the diverse
cloud of radio amateurs that AMPRnet stands for, and making it part of
the regular internet, available to anyone without a license and removing
traffic exchange with the remainder (44.128) of the AMPRnet. This
proposal is, in your spectrum analogy, taking away valuable ham radio
spectrum and dedicating it to general internet usage.
I call on the ARDC board to disapprove this plan.
Then again, this is the opposite of the language in the
official proposal (last page), making me wonder whether
this is truly what the TAC actually discussed and decided.
Are there any public TAC meeting minutes or the like by chance ?
The former could also be read as a
policy/guarantee that no
non-amateur-radio based means of communication are involved.
Is that intended ?
Yes, correct. One of the things that this proposal can bring is that a part of the
network is reserved for radio amateur to radio amateur communication.
I note that you also use the term
"radio-only network" on page 3.
Since 44.0/9 according to your proposal is not "radio-only", this
would mean that 44.128/12 should not be accessible from 44.0/9, which
is the opposite to your proposed resolution.
This is actually (part of) the proposal. That we guarantee that people in 44.128/10 can
only be reached by other people there, and people in 44.0/10 (technically /9) can be
reached from the entire Internet, except 44.128/10 (natively). This is similar to how only
radio amateurs can transmit in a ham band, but everyone can transmit to an ISM band
(including hams). b) Which route do I need to put into my router to
address the radio
network ? In particular, how can you answer this question without
considering the specifics of each individual case ? Why would there
be only one route?
You can address the “radio network” with a single route: 44.128/10. This proposal
guarantees that everyone there will be on the radio-only network, the same way
transmitting to 144-146 MHz in the EU is to reach hams only. Any traffic you receive is
(should be) from a ham, and you should only send anything if you are a ham, and you intend
to reach other hams. Transmitting to 2.4 GHz in the ISM band allows you to talk to more
people (anyone), but also anyone can talk to you. c) Can you back up the "originally
intended"
claim somehow ? I note that net-44 originated in the USA, which
historically has rather liberal third-party traffic rules compared
to other countries,
We probably have a lot of people in this mailing list that were even a part of this and
can speak up, but this happened before the Internet was (broadly) adopted and the 44/8 was
a way for this “Internet” project some people were working on to talk to this network of
these “radio amateurs” that they set up in the USA or Europe, etc. d) You propose a policy of not announcing the
prefix on the internet.
"the prefix" is presumably 44.128/10. Do I have to understand this as
going back to pre-2012 (no direct BGP) or pre, uh, 1990 (someone
remind me please when mirrorshades started providing encap tunnels and
announcing 44/8).
Yes, correct. This proposal wants 44.128/10 to not have any direct BGP allocations that
appear on the Internet. Connectivity of these networks should happen between themselves
(network to network VPN, radio links, …), the ARDC (or anyone else’s) PoPs, etc. and they
will not communicate through the open Internet. e) Is there a rationale why existing regional
networks cannot decide
themselves what level of internet connectivity they desire,
considering e.g. the local ham radio regulations
and keeping their numbering and infrastructure which have been
assigned to them long before ARDC existed as an entity. Is there
a particular reason for not grandfathering them ?
Unfortunately this would be difficult to accommodate as the guarantees cannot be offered
then. If radio amateurs don’t have a dedicated band to talk to each other, and they have
to use the ISM bands, there’s no way to distinguish between normal people and licensed
hams. You can’t tell and there’s no guarantee that the person you’re speaking with is a
ham or anyone else.
Similarly to the RF world, in IP there’s this kind of problem as well. If you have IP
addresses on the Internet, you could receive traffic from anyone. Sure, you can use an ACL
or a firewall, but that’s not guaranteed. Packets could be spoofed for example. If you
have a special network where you know that all senders and recipients are hams, then you
can build things with different assumptions. You can build internal tools or apps,
websites, etc. It’s up to you. It’s a band where you will only find people of the same
hobby as you, that are licensed.
This used to be the description of 44/8 (now, sans the AMAZN part). To
my knowledge, this never changed (see ARDC's AUP). If it did, I would
support a reversion of that change.
The other part is like an ISM band. Sure, you can use
this to talk to other hams, and you can use it to talk to non-hams, and non-hams can use
it to talk to you, and you have to establish by your own means who is who, and ensure that
they can’t trick you.
What our proposal aims to do is to create a separate “Ham Band” / Intranet / 44.128/10
and a separate "ISM band” / Internet / 44.0/9. By using simple RF or IP you can’t
have them collocated into the same space.
This is the reason why we cannot have scattered space and we want to have it aggregated
and easy to address. Instead of our “band plan” being hundreds of lines and have it change
daily, and move band from “ISM” to “Amateur Radio” and vice versa, we want to create a
very simple band plan of 2 entries that don’t change. One is, and will remain to be “ISM”
(44.0/9) and one is, and will remain to be “Radio Amateur” (44.128/10).
Having a more stable and simple band plan is easier for everyone. They can make more
informed decisions for the future, they can choose who they want to talk to, and they can
even decide to use both bands: use a handheld radio (Radio Amateur) and a phone with WiFi
(ISM). This is what we try to do on a technical level. Clearly define the two bands, and
make sure that they are very few, and very stable.
People *DID* that, based on the current policy that each subnet can
decide if it wants internet connectivity. This was the policy that
governed AMPRNet for at least since the mid-nineties.
People built their network around that policy (and the legal
requirements). Later they *DID* opt for BGP or not.
It is the TAC with this proposal that is flipping over the apple-cart.
For such a proposal, in addition to the potential benefits that such
a plan may bring, TAC MUST also address the cost epecially the work you force
uppon the affected users.
In the IP world this translates to easier routing (each “band plan” entry is a route, and
if it’s just one, it could even be a static one), and less frequent changes. I don’t have
to consult today’s band plan to know why 44.5.5.5 does not respond from 44.128.128.128, if
the reason is that 44.5.5.5 decided to be Internet-only today or Intranet-only tomorrow.
We could have made use of complex routing protocols and policies that would dynamically
try to discover what each address or subnet is (because it’s not always clear and we can’t
always tell what each address wants to do, even if we forced everyone to connect to an
ARDC PoP) and then continually adjust this and maintain a complex state. This is something
that a lot of people would also have to do, or they would have to find someone to do it
for them (e.g. the ARDC PoPs). Going towards our value of being as inclusive as possible,
we did not want to force people that don’t want to to have to do this or to have to
connect via an entity that can do this. By having a 2-line band plan that doesn’t change
over time people can even hard-code it if they don’t want to deal with all of this
complexity or necessarily rely on someone to do it for them and then form a dependency to
them.
You are forcing your world-view uppon all existing users of 44.128/0
and require those that disagree to leave and expend effort to renumber.
I do not think that "inclusive" is quite the correct word for that.
There are people that are fine with having the world communicate with
their 44. IPs. Nothing in your ham radio license forbids you talking
to other people, as long as you dont do it over a (amateur-) radio.
You proudly wear baseball caps with your callsign on it. Why can't
you wear your 44 IP?
That said, removing the currently existing option to connect to the
internet by central gateway or direct BGP is a major, destructive
change of current policy. Sometimes, such destruction is requisite for
progress. It is the duty of the TAC to review such a change.
Such a review must be balanced, and necessarily must include a detailed
discussion of opposing views, a discussion why there is no
simple technical solution to the problem, and a discussion of those
negatively affected by this change. This all should be documented
for the record. Meaningful review would probably have also included a
documented poll of the affected network's coordinators.
At this time, I therefore believe the current TAC's proposal is
deficient and thus should be rejected (albeit without prejudice)
Furthering the analogy, a handheld VHF manufacturer
relies on a constant band plan to allow TX to 144-146 MHz and doesn’t have to build a
system for their product to download this hour’s or this day’s Amateur Radio allocation
and change the functionality based on that. You can also be sure that your local amateur
radio repeater won’t be today at 89.7 MHz and your favorite radio station won’t transmit
to 145.500 MHz this afternoon.
This analogy is not suitable. For starters, the 2m band extends up
to 148 MHz in IARU Regions 2 and 3, which should give you pause to
consider the difference of what you are used to, what is legal in your
country, and what may be so in the rest of the world.
Secondly, why would I not be allowed to listen to that amateur radio
repeater using my commercial all-band radio receiver?
Best regards,
Mario, DL5MLO
--
Mario Lorenz Internet: <ml(a)vdazone.org>