TID: FMail-W32 2.3.0.1-B20240319
RFC-X-No-Archive: Yes
TZUTC: 0200
CHRS: CP850 2
MSGID: 2:280/5555 6885f8a4
REPLY: 2617.ipv6@1:154/700 2cea28b6
Hello Nick,

On Saturday July 26 2025 09:38, you wrote to me:

 NB> And you always will run into those walls. This is somewhat comparable
 NB> to sysops that still wish to run BBS software from the 80s/90s, that
 NB> don't and will never support things like a TZUTC kludge. It's
 NB> ultimately their choice.

The effect of not supporting TZUTC is less than the effect of not supporting
IPv6 will be. And, yes it is ultimately the choice of the sysop.

 >> For those confronted with te reality of IPv4 exhaustion and the
 >> shattered brick wall of denial, there is brick wall #2. Hang on to
 >> IPv4 no matter what tricks it needs.

 NB> Some people don't like change, until they're ultimately forced to do
 NB> so. If IPv4 is ever fully phased out, my guess is Fidonet will lose
 NB> another big chunk of nodelisted sysops. Much like when darktech.org
 NB> and no-ip.com were abandoned, but on a bigger scale.

Possibly. Read on...

IPv4 will stay around for quit some time. It won't go out with a bang, it will
just fade away. Like IPX/SPX. For all intents and purposes IPX/SPX is gone, or
at least it does not play a major role any more. It is gone from my house, but
surely it will still be in use somewhere. Something similar will probably
happen to IPv4. In the long run there is no business case for keeping both
protocols. So when "everyone" has IPv6 the benefits of maintaining support for
IPv4 along with IPv6 will eventually no longer outweigh the cost and affort so
manufacturers and providers will eventually stop support for IPv4. Like
support for IPX/SPX has stopped.

That will not happen tomorow. I may or may not live to see it. What will
happen in the foreseaable future and is already happening in sone parts of the
Internet is that providers no longer offer a globally routable IPv4 address to
their customers, they will have to make do with CGNAT.

 >> IPv4 exhaustion may not be a serious problem for the incumbents in
 >> parts of the world where IPv4 was historically issued as if it
 >> would last forever.

 NB> I imagine some of these bigger ISPs, mine included (Spectrum), bought
 NB> up a ton of IPv4 blocks because they saw all of this coming and could
 NB> afford to take it all away from smaller ISPs for themselves. Heck,
 NB> there may even be enough IPv4 addresses at these large ISPs to go
 NB> around still, but they are hoarding them for their own (current and
 NB> future) customers.

Ever since the well of IPv4 fell dry, there has been a market for IPv4
adresses. And when demand exceeds supply, the price goes up. And with dynamic
prices speculation and hoarding comes into play. But in the early days there
was no level playing field. Some early major players were issued very large
blocks of IPv4. And some have sold it what they did not need, some are still
sitting on it. In the meantime the peak is history. Price for an IPv4 address
has peaked at about $60 a couple of years ago. Now it is around $40. So some
of the hoarders have fished behind the net alraedy.

 NB> While I don't follow the situation very much, nor do I have any
 NB> experience with it (I have native "dynamic that never changes unless I
 NB> change hardware" IPv4 that I have always had, as well as native
 NB> "dynamic that never changes unless I change hardware" IPv6), but do
 NB> you think these bigger ISPs rent out some IPv4 addresses to smaller
 NB> ISPs just so they can do said DSLite kind of setups?

 NB> Is this DSLite setup a bunch of private addresses not open to the
 NB> outside, that run off of one or a few public addresses so they
 NB> basically disable any incoming traffic so they can make more use of a
 NB> single IPv4 address?

Full DUal Stack is a technology where IPv4 and IPv6 are on the same carrier as
"equals" next to each other. DsLite is a technology where the main connection
is IPv6 only and IPv4 is offered as a service via a 4in6 tunnel. DsLite is
always combined with CGNAT. But CGNAT is also used in other settings.

CGNAT stands for Carrier Grade NAT. A means to share IP addresses among many
users. It is basically the same as the use of NAT in your home network, but
the process takes place at the site of the provider. The range 100.64.0.0/10
has been set aside for the "private" side of the CGNAT network. But some
providers also use 10.x.x.x or some other RFC 1918 range. The use of
100.64.0.0/10 avoids conflicts with local adresses of the user.

Anyway, what is relevant to Fidonet is that with CGNAT the user no longer has
a globally routable IPv4 adress for himself. He can no longer run a server. No
server, no incoming calls.

Accepting incoming calls is an essential part of running a Fidonet node. It is
possible to run a Pvt node, but the network can not function without nodes
capable of accepting incomng calls.

Presently this affects only a handfull of nodes. But if this practise becomes
more common, it may have a large impact on Fidonet and we may indeed see the
numbers of nodes drop when IPv6 adoption does not rise in respons.


Cheers, Michiel

--- GoldED+/W32-MSVC 1.1.5-b20170303
 * Origin: he.net certified sage (2:280/5555)
SEEN-BY: 19/10 103/705 104/117 105/81 106/201 124/5016 128/187 129/14
SEEN-BY: 153/757 7715 154/10 30 110 203/0 218/700 221/0 226/30 227/114
SEEN-BY: 229/110 206 317 400 426 428 550 616 664 700 705 240/1120
SEEN-BY: 240/5832 250/1 263/1 266/512 280/464 5003 5006 5555 291/111
SEEN-BY: 292/854 8125 301/1 310/31 320/219 322/757 341/66 234 342/200
SEEN-BY: 396/45 423/120 460/58 467/888 633/267 280 281 410 418 420
SEEN-BY: 633/509 2744 712/848 770/1 902/26 5019/40 5020/400 545 1042
SEEN-BY: 5053/58 5075/35
PATH: 280/5555 464 633/280 229/426