[governance] Re: The Internet (as we know it) can never be "private"

[Daniel Kalchev]

Please note what follows may be too old for many and sometimes with more 
detail. I also try to not draw much conclusions at this time.

On 19.07.11 23:42, Miguel Alcaine wrote:
> In this case, it is convenient to follow the levels in the suggested 
> order: First actions at the local, then national, etc. We need to 
> overcome the barriers to cheaper interconnection costs. One way of 
> doing so, is promoting IXP at the local, national and regional levels. 
> The order is also important to have more meaningful discussions at the 
> next level.
>
> Here, I am also very interested in the experience in Bulgaria. How 
> ISPs finally understood the convenience of IXPs ? Could Daniel help us 
> in finding out more information about it?
>
> Another usual missing part of the equation is the promotion of 
> content. Infrastructure and content need to evolve hand in hand. Any 
> good experiences from anybody to share on this?
>
In the Bulgarian case, the development was driven primarily by "private" 
demands, that is, costs and user demand.
Of course, local content played significant role in this development.

In the beginning, Internet connectivity to Bulgaria was very expensive, 
due to relying on the monopoly telecom pricing. Connectivity within the 
country was not less expensive, for the same reasons, although the 
telecom possessed and operated modern infrastructure. But they were not 
letting anyone "buy" it -- you had to wait several months, sometimes 
years in order to get a line from A to B, while it was strikingly 
apparent that infrastructure and capacity is available.

I remember two such cases, that are indicative:

For a while, we have been using analog circuit for international 
connectivity. So far so good -- TCP/IP Internet was available! But, it 
was slow and unreliable (in today's perception even more so), so we 
wanted to go digital. Digital circuits were available, although very 
expensive and ... the telecom refused to provide the circuit for 
whatever reason. We eventually succeeded and it was only possible, 
because it turned out the PTT wanted to provide digital capacity first 
to their own X.25 network, then to "the competition" --- was it our 
fault that their network could not make use of this connectivity faster? 
I havent' looked in archives, but from memory 64kbit line to anywhere 
would cost like 6000+ SDR a month.

Another case was for national connectivity, same telecom. We wanted to 
connect a major city and were looking for a location to put our 
equipment (not much at that time). After discussing with the local PTT 
branch we were offered to collocate our equipment in their "new all 
digital switch room", a place where the optical rings across the country 
interconnect. To my surprise, everything was connected and passing dummy 
traffic there -- but no real users! We were excited and immediately 
ordered digital circuits from that city to few others. At a price like 
4000-5000 SDR for 2 Mbit for 100+ km distance.. Outrageous price, even 
if international connectivity was more expensive. We got no response... 
Two years later, the manager at the local PTT responsible for the 
digital infrastructure eventually got Green Card from the US lottery and 
left the country --- ironically, we got "permission" for the first 
circuit few days latter... During that two years, we were forced to use 
multiple analog circuits and to whatever tricks to have some 
connectivity to that large city..

Since then however, things developed in a different way. It was obvious 
that the state owned and regulated telecom would do whatever it take to 
stop this "Internet thing" from developing. However more difficult it 
seemed, we were forced to go forward and build our own networks, all 
over the country. The in-city networks were easier/cheaper to build and 
the initial inter-city links were all wireless, not so fast and reliable 
but way, way cheaper and instantly available --- going even to the 
smallest village, that didn't have phone coverage. Then, as traffic 
needs became more significant, few optical networks were laid across the 
country. Some even went that far to lay optical fiber cross-border, thus 
providing massive capacity opportunities internationally. The telecom, 
which was later privatized was practically forgotten and now they are 
insignificant player, although they still continue to have the most 
extensive infrastructure available (form the times the state invested in 
laying cables all over the country).

For some time, the phenomenon of "Local city networks" emerged in 
Bulgaria. These were networks with relatively poor Internet 
connectivity, but offering LAN speeds (10-100 Mbit) to local content. 
This built local Internet communities and demand and also local 
interconnects. Eventually with inter-city and international connectivity 
becoming cheaper, these networks got better Internet connectivity to 
non-local content and regional/national interconnects. Some of these 
networks also became larger ISPs offering wider selection of services.

As to IXPs development. In the beginning there were some attempts to 
organize the IXPs by few ISP organizations. But this never worked. At 
that time, private interconnects were the rule and indeed, few of the 
ISPs were offering "free lunch" to their peers, that is, transit to 
other national networks. There were few "official" IXPs but because they 
were hosted by one or more ISPs other didn't find this neutral enough 
and were reluctant to join. We tried to also organize IXPs based on 
Academia grounds, as an neutral place to interconnect, but this too, 
newer worked.

There was the telecom initiative to provide city-wide and nation-wide 
MAN networks, but they never got the price/attitude right and although 
these services were used for interconnects by many, and are still used 
-- this initiative was never competitive enough.

The first successful "real" IXP was purely private and focused effort -- 
to fill the need. A privately owned enterprise, especially founded for 
the purpose and focused exclusively on providing IXP services. So far 
has exceeded expectations. Of course, private interconnects between ISPs 
still do exist and grow. But the IXPs have permitted smaller players to 
have access to the same Internet connectivity/capacity as did larger.

Besides IXPs, what helped development and driving costs down was the 
creation of few data centers. The first data centers were rather 
primitive by today's standards but provided cheap means for data 
providers to host their infrastructure without inuring enormous costs 
for connectivity.

What is common in all these success stories: 100% private initiatives 
and 100% lack of regulation.

If, say 20 years ago, the government has chosen to other the monopoly 
state owned telecom to at least provide connectivity without delays, we 
could have achieved this much, much faster. But apparently, at that, 
already democratic times those in the Government who were "governing" IT 
and telecom development didn't see any value to developing Internet and 
so, their "governance" was in effect trying to prevent it from happening.
Not so, with the private people. It is important to understand, that by 
"private" I do not mean corporate, by any way! Most of the ISPs in 
Bulgaria started very, very small. Sometimes I guess even as family 
business. Those people all invested their efforts for the Internet to 
happen. Even if some of them dreamed of "control" and were greedy (I 
have many examples, observed over the years), they soon realized, that 
they can achieve this on/via Internet as much, as they can achieve it in 
their normal human society.. and if they were unsuccessful in the human 
society, they would fail the same way in Internet.

Years ago, there was a great debate in Bulgaria, whether Internet should 
be subject to regulation. There were intentions to require ISPs to 
obtain licenses etc. Ironically, to my knowledge, that regulation was 
invented by some of the existing ISPs, that didn't really understand how 
Internet work and dreamed of having absolute monopoly (or oligopoly, 
rather). Many claim to have direct influence on the outcome, but my 
position is that the outcome was community effort -- at the end the 
Government backed out of regulation. I believe they are now happy with 
that decision and the further Internet development has happened without 
any effort on Government's part.

Current raw connectivity costs in Bulgaria are well below 1 euro per 
megabit for both inter-city and international connectivity. Connectivity 
costs have reached the level at which users are not really paying for 
connectivity, but more for service and support.

This has become longer than expected, but let's home some find it useful.

Daniel
____________________________________________________________
You received this message as a subscriber on the list:
     governance at lists.cpsr.org
To be removed from the list, visit:
     http://www.igcaucus.org/unsubscribing

For all other list information and functions, see:
     http://lists.cpsr.org/lists/info/governance
To edit your profile and to find the IGC's charter, see:
     http://www.igcaucus.org/

Translate this email: http://translate.google.com/translate_t