[enum-wg] ENUM Adoption - Does a business case matter?

Hi,

Torsten Schlabach wrote:
> Hi Otmar!
> 
> First of all: Thanks for some new aspects your brought to the
> discussion. I didn't know that ZapMail thing before.
> 
> I think the VoIP as a product versus VoIP as a service thing pretty much
> hits it. Just I have to heavily disagree with what Clay Shirky wrote.
> 
> VoIP as a product (buy a device, connect, talk to people) has never ever
> worked. Not all all. Any this might be the problem after all.

Correct. This is in contrast to email, where you just

* get internet connectivity
* buy / get a mailserver software (commercial or open source)
* get a domain and set the MX records

and you're a full participant in the worldwide email ecosystem.

> If I could buy a VoIP appliance in the next electronics store which
> would allow me to start talking to other people who bought an appliance
> which uses the same protocol and if this appliance would be cheap and
> required no but a trivial configuration, that would be it. For sure.

An appliance alone would be not enough, as you still need the domain for
plain SIP and/or the enum delegation for E.164 based calling. But still,
why doesn't this happen the same way as it does for email?

In my opinion, the free SIP community never managed to get over the
Metcalfe's-law bump in the adoption curve as compared to the huge
Metcalfe-worth of the existing PSTN.

Any phone that can call the billions of existing PSTN phones is worth so
much more than any VoIP device that can only reach others using the same
technology. It isn't worth deploying an isolated, new telephony network.
(Even the IM-based phone networks are IM networks foremost and talking is
only a secondary usage. IMHO.)

What is another difference to email?

* Ad-hoc, settlement-free interconnection.

There is no need to sign contracts to transmit email and thus no money
passes hands. That's not the way the PSTN operates and as I explained
before, no sane PSTN operator will want to be the first to change it.

> But reality is that the VoIP operators behave exactly like any
> traditional telco. VoIP as a service. 

Because you need them as a gateway to the PSTN. You don't have the
contracts to terminate calls directly to BT, FT, ATT, DT, Sprint, ...
so you have to buy a this service from someone.

My private box can send email to all these big operators. My SIP proxy
cannot talk to their networks. That's the difference.

Free SIP service wasn't the problem. Just as there is GMX, gmail, Yahoo,
hotmail & other which give you a free email address, there were services
that gave you a free SIP address. (iptel.org, FreeWorldDialup, ...)

> You have to subscribe (ok, the
> subscription fee might be 0,00), there is a tariff plan, calls are only
> free on the same network or on a small number of interconnected networks.
> 
> What I always wondered is why people do accept that.

As a single person, they have no choice.

> Would you accept an ISP which delivers email only to accounts on it's
> own and on selected partner networks and who will print and snail mail
> or fax if the mail leaves their own network. This is excactly what
> happens if I am with VoIP operator A and I make a call to VoIP operator
> B. Unless two two have a peering agreement by chance, the call will go
> VoIP -> PSTN Gateway -> PSTN Gateway -> VoIP. How stupid is that?

It's not stupid. It's just a different economic stable state. The forces
who protect the status quo are stronger than the forces who want change.

Both interconnection schemes (pstn-type settlement and email-like ad-hoc
settlement-free) are stable states. Any single operator cannot change the game.

----

Side-remark: it's instructive to look a the satellite pay-tv markets all
over Europe: There are some countries where pay-TV emerged as the dominant
solution (France, UK) and then there is Germany, where free, ad-supported
TV basically won the game. Any new entrant into that market has to go along
with the dominant model. The typical German viewer will not sign a
contract, pay money, get the decryption equipment just for the privilege of
watching just one more channel. For a French, the infrastructure and habit
is there and people probably won't tolerate the ads.

Two stable states.

----
> 
> Now the question to me is: What the hell do I need a VoIP operator for?
> 
> I can make a device listen for incoming calls on an IP address on a
> given port. If we take IPv6, I can have a fixed IP address which will
> even be entirely independent of my ISP and it will even be mobile. 

Sorry, no.

ipv6 has nothing to do with getting everybody a static ip-address. Quite
the contrary, ipv6 was designed to make switching ip-addresses when
changing location / ISP / ... easier than it is for v4.

What you really want is a domain name which points to your current ip address.

ipv6 allows you to get rid of NAT.

----

> In case it would happen that ... say as many people as are using Skype
> today ... would use standard protocol IPv6 based voice over Internet I
> am sure this would make the telco's move about how to make that cloud
> reachable from their networks. And there ENUM would be back in the game,
> mapping a phone number to a protocol:IPv6address record.

Not at all. The current setup where the voip-users pay money to some
voip-operator which then hands termination-fees to the telcos suits the
telcos just fine. What advantage would they get if they would allow direct
calls towards them from individual, anonymous voip-users?

/ol
-- 
// Otmar Lendl <lendl at nic.at>, T: +43 1 5056416 - 33, F: - 933 //