WS 126 EURidUNESCO World Report on IDN Deployment 2012 - IGF ...

chulavistajuniorMobile - Wireless

Dec 10, 2013 (4 years and 23 days ago)

93 views

WS 126 EURid/UNESCO World Report on IDN Deployment 2012


opportunities and challenges
associated with IDNs and online multilingualism

>>

Giovanni
Seppia
,
: Okay. We have decided to start the workshop as we heard that the
--

was
told they were adjourned unt
il 3.15 but many of our obligations and other work shops we really
need to start and I would like to (very faint sound quality) really moderate if that is okay and
representing different stakeholders and this is second year that we have produced this
refer
ence and as of this year we have decided because the plan is to make it a regular annual
workshop and record which people can look at if you like to have a complete more information
about ID N and value for multilingualism and let us say for each division
of people. We are
registered and we have to support 33 languages not only at ID N level but also when we provide
support for our customers and essentially those languages that will become 27 languages in the
near future because there are changes I am tryin
g to discuss. So I am quite happy to open this
workshop and I would like to introduce, working together with UNESCO and she has
investigated ID N work at top level and at a second level and she has mate also comparison of
the results that we have obtained
over the last year against the results that we got. There are
some copies by the way of the work here, paper copies and you can find also this work that in a
PDF. So having said that I would like to leave the floor to Emily and she will introduce and speak

about ID N residence mapping which is a very (inaudible) thank you.

>>
Emily Taylor
: Thank you very much Giovanni and it is a real pleasure to join this panel and to
participate for the second year on the world report for ID N deployment with you rid and a
lso
UNESCO. This year we expanded the data sample that we were looking at and so it is now cover
in up to 90 per cent of the world S registered domains. Now of course we don't have every bit of
data for everything but we hope we can slowly fill in the blan
ks. I think the significance of the
time in of this study is something worth mentioning in that it is only in the last two or three
years that we have seen the registration of international liesed domain names throughout
history at the top level as well as

the second level and so our study is really marking and will
hopefully stay with this process as it really gets under way. There are copies, as Giovanni said, of
the report here in the room so please do take one but really just as a very brief overview we

have looked at a quantitative research on the takeup in different countries and regions and also
we showed that there is a very strong link with the languages spoken in a country or region and
those scripts that were the register chooses to support. It is

almost, it is over 95 per cent
correlation there between that, so this gives us a clue that we believe there should be a very
strong link with those internationalised domain names and the content found on the Web sites
associated with them. We look at the

time line. We also look at some general issues that we
believe are barriers to the take up of international liest domain names, such as the inconsistent
representations still in brow services. There is the inconsistent and the deployment of ID Ns in
e
-
mai
ls which is only just getting in its infancy, but also the supporting applications. For
example, Facebook with over a, billion users it is still not possible to create a user account with
an international liest domain name. All of this things throw barrier
s in the path of those who
would like to register and use in our view. We look at the experiences of different countries,
studies and then in the analysis, as Giovanni mentioned we start to build a matrix to help
understand why it is that say, for example
in the Republic of Korea we see very healthy uptake
almost immediately whereas in other countries or regions that is not the case so why is this?
And our hypothesis is that this depends on three broad factors: the first relates to the conditions
in the cou
ntry itself. Because is there, for example a big enough user base. How homogenous is
the country linguistically or culturally. Are there the conditions there that would foster a high
uptake but also what the broad band penration is like. I sop study has sh
own a very high
correlation between the infrastructure locally and the presence of local, and we believe that the
same factors apply also for uptake of international liesed domain names . Then having sort of
looked at the country or region we also believe
that the way that the registry locally is run,
registration the price of domains, how well known the brand is locally and also crucially
whether or not there is a network of local registrars. I was speaking with Lynn underjust before
the session started an
d she confirmed that in the Republic of Korea they have 31 registrars who
are all based in countries. Those are the classic channels to market that they are also serving the
local nationality with other services such as Web design, poa hosting and all of t
hese things that
create a virtual circle in our view towards the uptake of multi
-
lingual content supported by a
multi
-
lingual navigation system. So as you can probably tell I could go on for hours and hours
but I won't because there are many interesting pe
ople on the panel and I would

like to hand
over to vin cerf
who very kindly presented the full work to this report, thank you.

>>Vint Cerf: Thank you very much
. I don't intend to take a great deal of time in this initial
intervention but one thing I want
to point out is that when the international liesed domain
names were first proposed around 2003 a great deal of work went into the design of the system
that would allow the use of non
-
(inaudible) characters. In such a way that none of us, the
servers and e
e (inaudible). What this forced on us is a new coding of new code sets in ASCII
codes through the processes of pi any code for no good reason an what basically we have done
is algorithm the Uni code and codes for specific symbols in a particular script and

put them into
a series of ASCII characters and that algorithm run in reverse to go from the ASCII I any code
version back to the Uni code version of the characters and then the we interpreted and
rendered as so that is one of the reasons that we have rath
er odd traditional step in mapping the
names. In fact if you look at the ASCII fringe for international liesed names they all start with the
characters X and N. This is chosen at random by the (inaudible) and done with great secrecy so
it is chosen so that

nobody when it is registered international lie
--

domain. So that is a reason
that we have to do it that way but there is a second thing that I cannot impress upon you
strongly enough and the domain names in order to be, it has to be compared with somethi
ng
that has equal standard ma newly or cut and paste (inaudible) for example browser O or
address file so successful matching is fundamental to the way, matching and it is very easily. We
were in sensitive to (inaudible) and the ease of designing the extre
me registered we are the
same trend simplified in this vur on of ASCII and maintain in the international liesed design in
code mapping as anything else. It was very important of representing these characters so that
up something high probability of finding

and correctly matching every one of these symbols.
Matching is fundamental to look up and people who were constantly using Uni code for the
purposes of simply representing symbols in the script and then turning had a lot of trouble
understanding why there

is more work involved in restraining what the coding was in order to
ensure the matching process. You can't use matching percentages to look up the entire
community addresses to go where we (inaudible) but there were a lot of constraints and got
subjet th
at they couldn't write. In an infinitely expressive way the characters even though under
Uni code you can write the same characters several times. The problem is that if you go one way
then the registration is done in a different way you can't match so the

ability to match is
fundamental. The same thing is going through with e
-
mail. Strings to the left of the hah sign
have historically been less constrained in the Internet world than string to the right of the sign
and the main names are very constrained. S
trings to the left of the sign represent mail box are
less constrained and therefore we actually have a bir challenge to do the matching. That is why
we need to resolve the process many people get very impatient. They say I can write anything I
want in you

any code on a Web page, what is the problem? The specific problem is getting the
computer to figure out the way you type something is matchable to something you are looking
for, for example somebody's mail box identifier. It is very hard sometimes to help

people
understand that this is a non
-
trivial exercise for a piece of software to recognise it is the same
string. Human beings look at the rendering and they say it looks the same it must be the same
string. The computer does not see it that way. It sees
it as a Kim bol. So there is (inaudible) in h
forward to this report I raised the issue of having to do the financial update of international
liesed (inaudible) there was much less than I had expected tr the model I had anticipated
wrongly is that as soon
as we have top level domains that could be expressed within these
international liesed scripts that lots of people would normally have had international liesed
name rz at lower levels, labels that were level two, level three or level four would block to
ie
wings the top level domains characters so that everything looked like it was come in from the
same script. So incredibly that has so this is a surprise that that is (inaudible) and.

So obviously something is different than the model that I anticipated an I

hope that in the course
of hiss session we can figure that out.

So that is where things stand right now. We are fully functional and top level domains fully
functional in the use of international liest string in the domain names fully functional codes in
the text of each e
-
mail Web pages and we are just getting going on the strings for e
-
mail. So I will
stop there.

>>: Thank you. I just want to catch up with just to say (inaudible).

>>: And they say that from that moment
--

IBM domain ds so I am very curio
us to see that as
well because that will.

>>: If I intervene one more time. I failed to mention one other problem area, it is called
bidirectional scripts. It is very, very, it is problematic to have the domain name that has labels
that are in different sc
ripts, some going right to left and some going left to right and there are
some conditions under which it is impossible to correctly render. Do I need this attached to my
nose. Is this okay.? All right. So the big problem (better sound) is that in bidirect
ional strings,
figure out how to render those strings correctly becomes quite difficult, especially if there are
doints the middle and there are some situations where the dot moves and you don't see what
was intended. So I am not I don't think we resolve a
ll of those cases and the only mechanism
that was available to solve the edge cases was to say don't register that particular label because
it won't be ren Drabble in a reliable way and this may also be a problem in the e
-
mail (much
better sound
--

tanks r
h.

>>: And I would like to give the floor to Janis of UNESCO and UNESCO has been producing this
record with us as they did last year and we are extremely happy an it is a great partnership and
we would like to continue in the next year and I would like to
leave the floor to Janis to speak
about UNESCO, multi
-
lingualism, what efforts comprise, Janis what you think is going to be in
the future of multi
-
lingualism and what can we do to support multi
-
lingualism.

>>
Janis Karklins
: So thank you Giovanni and likew
ise from the UNESCO side we are very pleased
with this co
-
operation and the outcome and the production of this role report. I will maybe
continue what vint was telling about a little bit of the history of development of IBM's from a
technical perspective.
There was also certain developments from the policy perspective and I
still remember that one evening it was in 2007 or eight in one of the Gulf countries
where
myself, and Chris Disspain

we were kind of not corps nrt but surrounded by representatives
fro
m the region and they almost told us we will not let you go from here if you will not do
something with international liesed domain names. We understood the request or the market is
there. There is a need and Teresa drafted the e
-
mail to Paul tu me Preside
nt and CE0 at that
time and that was the beginning of what we now know fast track IBM CCT O B fast track process.
We succeeded with the policy work I fairly quickly because there was this hype on let us do it
and then life will be completely different and
this was indeed very interesting policy process
with the big demand from member states from users side and this report currently was, or the
whole process actually, last year's report an then this report and the one that will come next
year, is intended re
ally to analyse and understand what favours the use of IBM CCT O Bs and
what are obstacles which need to be overcome because many hours of work have been put in
developing IBMs from a technical point of view and I think that we have reached the moment
that

we are, we have a house, we have everything which is needed for the roof but we do not
still have a roof and we need to put that roof and of course as we see from the report the
applications, either in the browsers or in social media or in particular e
-
ma
il is something which
will finalise this house if I may say and of course everybody is very eager to get this house ready
that we can use. There is also an aspect which we need to understand in answering your
comment, why the direct migration did not happe
n. Because if there is no, let us say the same
commercial value of the domain name as it is in ASCII with the ASCII extension, then people
think twice and no matter how attached they are to the language if there is no immediate
benefit or immediate conveni
ence in using that there is a certain period that the uptake takes
place. This report for UNESCO means the possibility and argumentation to make a policy
statement geared towards Government of the world where Latin is not used as script for the
language. S
uggesting that there are opportunities upcoming opportunities and what role
governments or what steps Government should play in facilitating this uptake and introduction
of international liesed domain names. We take as granted those who speak languages in
Latin
script or those who speak English or French or any language in Latin script. That Latin alphabet
is known everywhere and everybody knows it and believe me it is not true. As I do not
recognise
--

I can identify Arabic alphabet, but I have no idea wha
t sign means what. So the same
applies to ASCII characters. There might be people and believe not a one, billion on this planet
who do not recognise ASCII characters so for them that might be an obstacle. If we really can
offer the full package of services

we may hope that then the ID Ns will really take off and start
living their lives. It was also very interesting to see the different scenarios are introduction of ID
Ns. A number of ID Ns were introduced at the same time in Russia, in China, in other coun
tries
in some other countries. I think Egypt was also one of the first ones who got it and what are the
different ce na radios an again the analysis allows us to understand what influences those
different uptakes and different scenarios and countries and a
gain for UNESCO it is very useful to
understand and use this ourselves these argument in our policy work in advocacy work with
our governments.

We are willing to also to engage on the basis of findsings of this report with organisations of
technical nature

and I see already drafting the letter addressed to the Chairman of i.e T F, maybe
the CE0 of Facebook or Twitter, encouraging them to work on ID N capabilities of the services
they provide or encouraging technical community to work on necessary technical
solutions that
they feel this gentle and friendly pressure from the outside and that is always very stimulating.
So thank you you rid, thank you Emily for the work done and we will certainly be willing to
present this report to the member states of UNESCO
during the vi sis review conference in
February next year, again to just to show what are the advantages and what policies should be
put in place in order to make ID Ms live and used in their countries. So thank you, Giovanni.

>>: Thanks a lot Janis and I
just like to catch up with what you just said about the role of certain
providers and the role that can be played to support ID N, I would just like to ask a question to
vint as forward you just mentioned that one of the possible issues against that take u
p of ID Ns
is in decksation of the conference so how does this relate, what can be done, what few ideas to
share with us to improve also ID N uptake via the indexation of conference.

>>:
Vint Cerf :

So the issue here is two
-
fold. First of all in the curren
t Worldwide Web the way
you point things is use domain names and other worldwide Web strings like H T T P: slash slash.
If you use a domain name to build U R L and that domain name is not maintained, if you don't
pay your yearly fees then that U R L may no

longer resolve and, or alternatively it may resolve,
that is to say an IP address may be associated with it which is not the one that you intended so
you can imagine producing documents or other digital objects, putting them on a machine
which has a parti
cular domain name and leaving the them there for a while and then forgetting
to pay for your domain name registration and someone else gets that domain name and actually
puts content on whichever offence that domain name resolves to now, the new owner and
now
one of the old U R Les if it is simple enough and it does not have a lot of the additional right
-
hand side characters, parameters, is referenced, you may land on a Website and pull up a Web
page that you didn't create that somebody else did and this is

particularly awk mrt if it turns out
it lands on a Website that has not very nice material on it so your old pointer goes from being
resolvable to a place where your information is, to not resolvable at all, to resolvable to a place
where your information

is not but somebody else's is so you create ambiguity by a domain
name which is not maintained imperpetuity. My concern here is that it be international liest
domain names are registered and then they are found not to be particularly useful or attractive
and people abandon them. The objects that we are associated with them may disappear or
worse the domain name resolves to a thing which it did not originally point to. This all has to do
with the funny dynamics of domain names. Before we were charging for t
hem, it didn't matter.
When you got one it was yours, you kept it or you used it, no one else was reassigned that
domain name so the trap didn't occur. It may not resolve but it certainly does not resolve to the
wrong place. As soon as we money tiest demoa
in names for very good reasons, because the
infrastructure needed to be self supporting or self sustaining we created the possibility of a
domain name that ex parte for lack of payment could be accused
--

there is a solution to the
problem, one is a system

called the handle system developed by Bob conwho is CII where
strings are generated, pointers are generated that are not ever reused and if you never reuse
them then the worst that can happen is it points to nothing but no one else is assigned that
string

so the handle system is intended to be. With you of referencing information for arbitrarily
long periods of time. So those are the kind of things that have worried me about domain names
in general, not just ID Ns. But if it turns out that ID N uptake is s
omehow meak or it gets tried
and then is abandoned we may have these expiring and disappearing content folders and that of
course would not be a good outcome.

>>: Thank
s a lot vint. As Janis mentioned, some of the examples of ID N uptai especially at a top

level. I would like to leave the floor to min unParker and she will speak about the experience in
Korea and Korea together with the Russian federation, good example of ID N uptake at the top
level. So thanks a lot min un.

>>:

Minjung Park,

Think you Giov
anni and I would like to thank you ces go for organising this
work shop for ID Ns and for giving us the opportunity to torque about our status and the issues
related to ID Ns in Korea. My name is M I N unand I work for A key sa Governmental agency,
managin
g the domain names and ID N A reses in Korea. Let me first briefly introduce you of C
Korean cc L D, we call it a handbook to assist representation for our country. We have launched
the handbook in May last year, 2011 and it turned out to be quite successf
ul and we had about
220,000 registrations by the end of 2011. Since the launch of that handbook last year we have
been receiving some complaints from the end users about the limited usability of that handbook
domain names. Not so surprisingly most of the c
omplaints were from mobile phone users. Korea
has a very high pentation rate of Internet and, 60 per cent of the population are Smartphone
users so you can see that most complaints came from the mobile users in Korea. Some of the
issues, main issues that w
e have confronted were about the browser problem. Some of the brow
services were not supporting ID Ns and we also found out that the Android phone users they
had problems with ID N links, the automated link service were not provided for ID N users in
Korea
. (Laughter).

>>: Mkr Chairman, may I intervene briefly.

>>: P
leas do so.

>>: First of all I am not going to try an defend anything I think sometimes software doesn't
work, what would be extremely helpful I would ask as an action item if it has not already

happened that the documentation about the problem be made available and I will see to it it
gets to the people responsible for the Android development.

>>: Thank you. Because Korea has been having some problems in contacting the right person
about this pr
oblem and yes, thank you for the assistance.

And yes, these were some of the problems that we were confronted with and the browser
problem was solved as the browser is, the Internet explorer, MS has provided the upgrai dt
version providing the support for
the ID Ns so it has been solved but the ID N link problem has
still remained so we have to work with the relevant stakeholders and
--

so we have been
working with the application developers and mobile phone manufacturers as well, Korean side
and et cetera
to improve the usability of that handbook since last year and we have had some
success in some part but we are still tangling with M some other issues and we also plan to
promote some more promotions and campaigns to promote the use of that handbook to dom
ain
names, such as, for example, co
-
operating with the public service agencies to put the Korean
names as their main U R L to make it more, to raise the end user awareness. We had about
220,000 registrations last year but it had dropped to
--

I just found
out that it dropped to
100,000 so it is a sharp decrease. We have actually
--

actually we found out that there are about
30 per cent of actual use of domain names last year so internally we had a discussion about
whether to have a promotion or campaign tar
geted to the annual renewal date which will be
May where we decided to have the campaign at the end of the year because we didn't think the
actual usage rate was healthy so we decided to let it drop. We didn't think that it would drop
this much but we ende
d up with 100,000 say healthy registrations about more than half of them
in actual you rate so we planned to have a very aggressive campaign with the co
-
operation of
the Government agencies and public service sectors as well as our registrar rz. I am start
ing at
the end of this year which will go on to next year so we hope to see a more healthier growth in
the registrations of ID Ns in Korea and with the growth of ID N e
-
mails, probably next year, we
also are expecting to have a more uptake in the ID Ns in
the near future and, yes, thank you. This
is the status of Korea for the time being.

>>: Thanks a lot.

Can I ask you something because you just mentioned something I spent 10 years in the registry
business now and you spoke about healthy registrations and
do. Think it is valuable to have
another campaign to promote ID Ns to push volumes or shouldn't we, because we have
implemented ID Ns as well, shouldn't we push the use rather than volumes, so push more what
more on healthier registration, the main names t
hat are effectively used?

>>: Yes, we also had a very hot debate on that specific subject and we
--

our conclusion was to
focus on the healthy usage first and to promote the volume as a next step, but it is just like in
chicken and egg game, it goes hand i
n hand. You need to have more volume to make people
aware of the ID Ns and to make, get as a pressure to the application developers also as the
famous sites to actually implement the ID N supporting programmes so the volume is
important, yes, but we decide
d to go with the usage, the actual usage first and then to target the
volume next.

>>: Thank you. I just wanted to intervene.

>>: I just wanted to ask one question in relation to a situation in Korea. What is the participation
or enthusiasm of the Governme
nt agencies in applying and running ID N domains of their own?

>>:
Minjung Park ,
For the time being most of the Governmental agencies, I guess, more than 90
per cent, have the domains with Korean CCT L D but the problem is not many of them are using
Korea
n CCT L D domain name as their main domain name so they are still using ASCII domain
names to promote their home pages so we are trying to persuade them to use more Korean
version of U R L to promote their service.

>>: Please
--

thank you.

>>: It is really

an observation listening to to, mi N underbecause we were we were looking at the
diverse uptake of ID N, the republic of Korea stood out as one of the countries which had the
most favourable environment for uptake for strong uptake. Particularly the cultu
ral linguistic ho
mo je spnei ti, excellent Broadband penetration and presence of I X Ps and so I think that the
experiences that you are describing as this being not just a one off thing that you can just click a
button and then hfam everything is suddenl
y multi
-
lingual, the small steps that you have to take
and the involvement of multiple take holders both as superusers and very visible users, trying
to motivate the application creation industry to see this as something that they must serve
because there
is a benefit and also to harness the marketing power and the channel to market of
your registrars. I think just my observation is that when you look at the ID N readiness if you
like that we think your country should have and the incredibly positive steps
and positive
outlook that you obviously have towards ID N uptake it just I think highlights the scale of the
challenge in moving towards a multi
-
lingual addressing system in my view.

>>: I recall having a discussion with a colleague whose use of ID Ns was
primarily Chinese and I
was speculating thekr that once the ID Ns were available there would be a very rabbit uptake
and I was told it might not happen that way because many of the people who would normally be
expected to introduce international liest doma
in names found it harder to enter those than they
did to enter the ASCII characters in the first place and in fact I don't know if this is true for the
Korean case but in China someone typing on a keyboard will often type an ASCII string whose
pronunciatio
n sounds like the letter that they want or the symbol that they want M and then
drop town menu pops up and they pick the one that they wanted which although I am amazed at
how quickly people can do that it looks fi clumsy and so the result is that sometime
s the ASCII
characters are preferred. I would be happy to hear some better information I have a feeling Miss
Prk is going to tell me what happens in Korea.

>>: Actually Korea's case is different from China case because we have a Korean alphabet, works
the
same way as that alphabet so you can just type it as it is, yes. One problem that we are facing
with the users again is that
--

(Laughter) we have been trying to persuade the mobile
manufacturers as a Samsung and apple to have a key pad to put the
--

just
like they have
a.comkey but the problem is that you need to type in much more letters if you type in
(inaudible) it is actually composed of six letters, it is two characters but you need to type six
more letters.com, three letters, you have to type in tith

muilfth this small mobile key pat, so we
are trying to persuade them the volume is that not that much so they are not willing to put that
much money in, so it is one of the problems that we are still facing.

>>: Excuse me but don't you make a lot of these

in Korea?

>>: Yes, so we are still talking with Samsung as well that they won't listen for the moment,
persuading
--

(Laughter).

>>: The people that make them in Korea don't want to make them for you I don't know what to
do about that!

>>: Well, we will t
ry harder.

>>: Okay, but you committed for the Android.

>>: What I have committed in the Android manner is to convey the problem to the people who
have to decide later and how they will solve it.

>>: Yes at least that. Okay. So thanks a lot mi N underagain

and I would like to move from that
handbook to Africa with the experience of (inaudible) who recently joined and she will share a
bit of the what are the issues with the African communities from place to place as with ID Ns
and if there is a demand of ID
Ns in Africa or other priorities are there so I like to leave the floor
to (inaudible), thank you.

>>: Thank you Giovanni and thanks to U rid and you ces go for inviting me to talk on this panel,
so the perspective, yes, I am going to talk about mostly Afr
ica south of the sa hara because as
you know from the experience of I can, a lot of the area in the north hat speaks Arabic is
practically covered. Africa is I think you all know one of the places where we have the most
languages on earth. About 15 or 1600

out of the 35,000 that exist. So, yes. They are very diverse.
Most of them are not written but the ones that can be written can also be written with the Latin,
Arabic or in core alphabets and cois one of hose that is invented by some of the African lingui
sts
from cam roon so that is already an issue, the fact that they are not written and the fact that they
can be written in so many different alphabets. My dad was one of those first UNESCO experts
back in the 60s who actually coded the alphabet for some of

the after can languages an I grew up
with six different ones so this has been fi dear to me from a very early age. The African
ministerial conference on education actually agreed that in 2010 that dual track education
should be a standard in Africa and th
is is also a way of basically making sure that are can
languages uptake comes up but having said that, the issue is when you do not have
--

so you do
have alphabets but when in the same country you have experts that would write the same
language in Arabic
in coor in Latin then there is not the right standard to follow so which one do
you chiewz? Which one do you actually take to do the education in schools is something that is
not done and a lot of times we have the issue of also finding, I mean, having gon
e to the
ministerial meeting and having set yes, dual track education is a must, you know, making or a
that that happens on the grown is not easy because you need the Lynn gfyists to agree on the
alphabet. You need after that the material to be produced an
d you need the kids to really start
learning it and the curriculum, that would be of course featured there. But there are a few
advances. You have places in Africa where for example sfa hilly being one of the languages is
taught, hal sa is also done and ha
l sa is done either with, again, Latin or Arabic alphabet and
haRick is one of the languages we have also in the eastern part of Africa, ee Theo pea, E ri trear
where they Ali trat in those languages, so we have quite a few people working on the encoding.
It has been done but that has not been translated into content, for example on the intranet, or in
African languages and this is just because, again, first you need people who ask for that content.
Given that we have a lot of literacy issues and we have qu
ite a few countries where literacy rates
are low we have issues with people who would simply read and then coupled with that now is
the uptake of intranet on mobile devices whereby basically people tend to speak and go look up
things that are, let us say m
ore on the I coNick and graphical stuff more than anything because
they don't know how to read and also yesterday I was on a panel and this was one of the things
that we were debating, how to actually make sure that the uptake of also languages and content

on the intranet in African languages comes when we have a place where we have the next,
billion that will come or hundreds of millions that will come from our vision will most probably
be on mobile devices and might be working with those I cons that they
recognise that are
becoming the Lynn via for an ca of basically the intranet and especially on tablets and things like
that, instead of the regular language so we are not too sure what that will give but this is one of
the things that we are working on and

I have been working with the association for the
development of education in Africa and some of the things that we are trying to experiment is
actually a few things with distance splabs that that are being provided by I tri reply and a few
others that we
are kind of pushing in, how do you say, pilot projects to see how that will work so
this is where we stand and of course one of the issues that we are having is also that in terms of
CCT, or these for example in the African region taking it up is just simp
ly the architecture and
the infrastructure is not strong enough and the people who are doing that are not savvy enough
in terms of how do you say, the what to do, basically to get ID Ns out there available but, you
know markets exist and knee shes, like, f
or example, I won't name names but a phone operator
has been able to have phones that speak sfa hilly and haRick in those neekh markets and it is
absolutely fabulous, you can't go to those countries and find a phone that speaks the Latin in
whatever alphab
et. I wouldn't know how to dial a number if I was using that phone because it
speaks in haRick or it speaks sfa hilly if it was, if the device is sfa hilly in Latin characters I can
still do it but if it is in Arabic it is a bit difficult so this is where
we stand in the region. A lot of
challenges. Thank you.

>>: Thanks, and just I would like to say that we have a moderator for the remote participation
and if there is any question coming from the remote participants or if there is any question romt
audienc
e, it does not
--

from the audience does not mean the people who are on this sitely want
to speak, please feel free to ask questions and interact with us. So thanks again A ra Shell and I
like to pass the floor to by har ass met of I conan I would like by
har to touch base on what I
conis currently doing to support ID Ns, what we are all aware of the process launched by I
conseveral years ago I think 4 years ago now for having the ID N facility, we have seen the
application with a limited number of ID Ns so

thanks by Rick for sharing your information,
yournology about these matters, thank you.

>>: Thank you Giovanni. Good afternoon everyone. Let me first congratulate UNESCO and U rid
on this excellent piss of work, the report. I have had the chance to go thr
ough the soft version of
the report a while back and I found it very informative and providing very interesting analysis
that is not only useful to us as experts in the field and the field but I guess it also establishes a
foundation for further work to be

done, especially on the usability aspect of ID Ns, given the
experiences of registries, registrars and user community application rovders and so on so that is
really a very good step forward. On I can and what is going on at I can so we introduced ID Ns a
t
the top level and we thought we were done but apparently we are not and what is coming is
probably more difficult and what is going on right now is much more complex than what has
been done in terms of fast track or having ID N used by people in Web brow
ser and so forth. So
a couple of years ago I can community started an effort to look into the issue of variance and just
for those who are not familiar with the term, variance is a term that has been used in very
different a ways and after almost a year of

the study on what variance are, the conclusion is
there is noot single definition of variance so it basically indicates that there is some sort of
relationship between lai peles, whether the full label or some characters of the label and this
relationship

varies according to the script itself so in some cases relationship implies that those
variant labels should be dealt with as if they were one label. In some other cases it could be that
if we are allowing one label to go on the D N S system then the rest

of the variance should be
band or deserved or blocked or whatever so the variant issue as an issue has been looked at
through a series of case studies in six different scripts, ASCII, Arabic, Chinese, agree, se ri lick
and Greek and this was through commu
nity experts who led the work through working groups
and all this and they came up with a set of reports identifying spulfts fiern the issues that either
relate to specific script or general issues that has to be addressed at a more of an overarching
level

at the D N S level or the root level and the same reports also identified a number of areas
that require further work to be done. Among those areas three that again I can community have
already started working on. One is related to the user experience, th
e usability aspect and this
seemed from the reports to be the most important, or one of the most important aspects
because it tackles or it addresses a number of areas from end user being the reji strand who
registered the domain name or the user who just
uses the browser or e
-
mail or any online
service or it could be the application developer who developed software or it could be the
system administrator who runs networks and the D N S service et cetera it could be the
registrar who maintains the zone file

and so on. And so there is one working group dedicated for
this particular areas and they issued an interim report and public comments have been received
and from now until probably the next I can meeting in the next six months there will be more
work on
the final report that will be presented then. The other two areas relate to what we call
the ID N table and this is simply the table that identifies the characters that could be used for the
specific script or language and the ID N table aspect, there are
two issue: one is the rules needed
to generate the labels an to identify the labels that could or could not go on the root and again
there is a dedicated working group looking at this issue and the third and last working group is
looking at the format of t
he ID N table because ever since we started having ID Ns even at the
second level, I can and through the repository has been maintaining the ID N tables for the
scripts that have been used but there has not been any specific format or even sort of policy o
r
rules or anything that defines how the table should look like what it should and it should not
include and all this so there is an Internet draft that sort of provided the initial view on using
something like XML as a standard technology for that and aga
in, people could actually have
provided comments on this draft and again in the next couple of months the process will
continue through i.e T F on developing this draft to the next steps so this is what is going on
with regard to variance on a different tr
ack the cc N S O and this is the constituency within I can
that look after CCT L, the policy issues, the cc N S O has been working on the long term policy of
ID N C T L thes because what happen when the fast track was introduced 3 years ago, the
thought wa
s that the fast track is a fast track it does not tackle each and every aspect of the ID
Ns and it was a short
-
term solution so the long term solution is having a long term policy on ID
NCC T L Des and this is something that will then doing some progress w
ith and just to mention a
part of it which relates to the use of geographical names in ID NCC T L Des, how the country
names would look like in different languages and scripts and this is something that UNESCO is ^
hekping I can with and helping cc N S O w
ith through the participation in the working oop that is
looking into this aspect and UNESCO has issued a survey to some of the member states, 30
something member states asking a set of questions on how do country names would look like
and different langua
ges and when you talk about country names we are talking about official
names, talking about short names, talking about famous or popular names and so on so it is
another track of effort. Something else that is also taking place right now is what we call t
he
universal acceptance and previous speakers have touched upon this issue when they mentioned
applications for services like Facebook and others when you can't have or see addressees or
names and ID Ns and the issue of universal acceptance in fact is more

broader than the ID Ns,
even with any new top level domain, whether in ASCII or in any other language, the fact that
many of the online applications do not recognise new top level domains for the simple fact that
the reason they I mean those applications
verify that this is a T L D is somewhat based on very,
like heart coded sort of tools that are not updated regularly so there have been discussions
within the I can community, within the technical community with business, from companies and
so on and this
is actually an effort that our role there is merely to raise awareness. It is not
really, I mean it is more of community pefrt, we need bises, we need Facebooks and the likes to
do more effort there to get those names recognised in a very efficient way. So

I will stop here
and would be happy to answer any questions if any. Thank you.

>>: Thanks a lot, by here I was about to stop you in any case. And I just wanted to ask you
something because, I don't know if you have any perception what could be the, after
the new T L
D, what could be the real demand for new T L D, new ID N G T L D. Did you hear anything in the
community. I heard a lot of passion in your voice for all ID N related matters and I think that
comes also from the fact that I can has developed a s
et of policies in different areas that you
mentioned the variance, you mentioned the ID N infrastructure S S T L Des and all this policies
have been refined over the year as they were subject to some sort of criticism from the
community but I must say that

I can has taken on board a lot of the comments from the
community so those policies have greatly ain jt but what would be next for the new ID N G T L
thes, what do you think is going be the demand of those ITN G T L thes.

>>: Well I think so we have recei
ved 10016 ID N G T L D applications with a range of scripts. I
think answering your question I think it is very difficult to predict. I think one of the le phons
learnt from introducing ID N at the top level and fast track and all this, it is really very d
ifficult to
predict the demand. So I have been hearing different feedback from the various language
communities so for some people, the variant issue is very important and people say there is no
way ID N could make it to the top without having vai, ryans a
n that is more required for G T L
Des more than for cc L Des. For others it could be other stuff like, again, the applications and the
Facebook and all the usability aspects so the short answer it is very difficult to predict, to put it
this way.

>>: Thank

you maybe we hear a long answer from very sound representative and thanks but for
taking the floor, really appreciate that as many.com, ID N, thank you very much.

>>: The question really is a difficult question because right now when you talk about 116
ap
plications there is supplier demand. We certainly have a lot of people who would M like to
offer those products and just an example of what we have.com, an.net, we have domain names,
Chinese, Korean, a tremendous number in Germany and a tremendous number i
n Turkey which
is a surprise to us, so it is interesting. But when we go out into demand studies there is not a lot
of demand. There is a lot of preference. We do a lot of prefrlns stuthis as well to say you want
a.com,.C N,.Chinese, cc L D Chinese, G T L
D and so there is a lot of prefrns right now there is not
a lot of trust because a lot of people are unaccustomed to seeing them, they say I don't know
what that is I wouldn't want that I don't trust it, it is not meaningful, incite of
China.commeaningful
in China. We struggle with demand. We have more anecdotal demand
than we do hard data that says if you put out a., something in hundred gfal or in Arabic that you
would Avenue these kind of registrations. People would begin to use them but the challenge is

one of the things you were talking about with universal spheaptance is we have not invested, we
have done a good jot in investing in tools I was sitting the other day thinking about this, as I
looked at my E trade account I sat there thinking what if I A
an ID N and I had to get my pass
word I couldn't get it because I couldn't type in (inaudible).comI couldn't type it in so we have
more to do and I think one of the great things that I can could do fuss is to help facilitate the
development or commission t
he development of tools that people could pick up and say I will
plug this into my Website because people don't know how to come forward, the R Cs are built,
people know what they should do but people don't always, are not always able to code
themselves as

tools I think there is a way we could actually do some of those things with some
of the money we are sit sitting on in terms of application fees to help facilitate than actually
drive it. We can't just have victory and say we have satisfied the suppliers.

We really have to
figure out to get real options. So.

>>: Thanks a lot, extremely interesting and we will see in the coming years and it is okay
--

>>: Mr Chairman I apologise. It turns out that I have an appointment at 4 o'clock with Minister
Vasey and I

need to excuse myself. But I did want to respond to something that occurred to me
as you were talking about domain names and demand. There is obviously in the ID Ns in a
country where languages are present and we heard from Emily about the strong correlat
ions
there. It also occurred to me that for people who want to market in places where those
languages are spoken, products and services, regardless of where the supply is we would also be
interested in having domains of this type, assuming that those domai
ns were more readily
recognised by the target so if I wanted to market a product let us say in Saudi I would probably
be smart to market it written in Arabic rather than written in English. On the presumption that
the target market is more likely to recogn
ise and even discover the product if it is described in
Arabic and the domain names might simply similarly be more easily recognised but of course I
am making all this up, I don't know if it is true.

>>: Well I think there is a great opportunity in those m
arket places to grow dand. It is just right
now there is not a lot of current demand. When we do the studies. Because there is not trust in
what you are selling. I think the Chinese market place has more experience from that standpoint
because C N N has su
mmary. Jnly resolving domain names they can probe within their borz, so
the experiencerience is there an it is growing in those areas, in other areas they don't have that
same opportunity.

>>: Okay I would like to thank you for this interesting, let us say

appendix of our discussion and
I would like to say also thank you so much to vint for participating. Also thank you by here and
we expect to finish in about ten, 15 minutes from now and so thanks again to the speakers who
are leafing now. I would like to
give the floor to Markus ptu Mr of I sa Co who joined this panel
and Isaac there has been a study developed by Isaac UNESCO and OECD which is mentioned in
the world report and it is a study that correlates basicallily the presence of local infrastructure
a
n local language content. I would like to see what is the perspective of the Internet, hear what
is the perspective of the Internet society about ID Ns and multilinguaism and how the Internet
society could compliment the current efforts to promote multilin
gualism and also ID Ns, thank
you Markus.

>>:
Markus Kummer:

Thank you and my apologies for coming late to this work sop and listened
with great interest, it is a fascinating presentations and I have to say I don't really have to at that
much, definitely NO
ANT expert on ID N and the Internet society has not marked much on ITN
but we have indeed Co produced a study with OECD and UNESCO on the relationship between
local conference, Internet development and access prices but before that I would like to look,
ta
ke a little bit of a step back and say it was a great leap forward, the introduction of ID N, when
you think back to the information society it was one of the big issues, the multi
-
lingualsation of
the Internet and without any doubt with the introduction o
f the ID N you make it possible for
people with another language than English as a native language to access the Internet easier an
yes, many people speak English as a couldn't language but there are millions out there who
don't have a second language apar
t from their local language and of course we as Internet
society welcome every step that enhances diversity on the Internet and listening to this
discussion actually it is very good illustration of enhanced co
-
operation. We had yesterday one
day event on e
nhanced co
-
operation and hear the co
-
operation you see between UNESCO and I
can community, is an example of how the Internet actually works, people working together and
that is enhanced co
-
operation and worth underlining. Our study basically proves what we

already knew: develop local content and keep the traffic local, that is how we develop the
Internet and indeed the study finds empirical evidence that this indeed is the case. There is of
course the philosophical aspect that local content is the heritage
of the society and it is
important for their culture, for their societal development and most people are actually mostly
in interest in content in their own language wherever they are from and I am sure you all
experience that when you travel abroad, you a
re most interested actually in your local paper or
whatever, wherever you are from an obviously their ID Ns are the key to go further and we as a
small aside we would have hoped that new G T L D process would have produced more
proposals in ID N precisely
to enhance diversity so we hope there will be more to come and we
also hope that the ccs will help to spread the message an to make the interest an see a potential
interest for local communities to have their own G T L D in ID N.

Yes and what the study fin
ds basically is, goes hand in hand with building Internet exchange
points, developing local content and that in the end brings down the prices so it is a virtual
circle and now what is first is the question where to start but I think you have to start at a
ll
levels. Yes you need to have Internet exchange points to bring down prices and yes, you have to
encourage local content and yes to have to encourage ID N and we as Internet society we look,
we have done in the past and workshops where CCT L Des in devel
oping countries and we see
that as a good way to move forward. CCT L Des are usually sort of the hub of the local Internet
community and they are can use these workshops for explaining, for creating interest for
reaching out. In the end I think the proof o
f the pudding of the new G T L D programme will be
has it not enhanced diversity on the Internet and there we definitely are interested in
contributing to diversity. This are my actual remarks.

>>: Thanks a lot thanks a lot Markus, yes please, there is a m
icrophone there.

>>: Thank you very much. I am BB N R from mo mo ro co, I appreciate Emily about ID N, about
IIP P X, I G F, I meet you there so I think that although some technical challenges have been
already solved, there is still more to do, this goal
of multi
-
lingual Internet where values like
transparency, tolerance, mutual understanding are the main characteristics. ID N rz are very
good initiative by I can and since we have ID N.ID N, so at the first level, it has solved many
problems for no Latin s
cripts so, for example, as another big Internet user I can find my way
around the Web with my own language. For example when I want to access to Al ja si ra
Website, for example, there are many possibilities. These words, using Latin characters, it can
be
Al ja si ra with two E or one E or I if I have a French by ground so because it is another big
word that has been written using no Arabic scripts. However, there is one Arabic spel for this
word which is Al Jessie ra nokter net so in Arabic it is one possi
bility. However, I think that
registration of ID Ns have been the first, second level. It has become possible to reach ITNs and
in terms of the country top level domains, like.sal dya or.Masa, it is great steps towards multi
-
lingual. However, I think that
we suilfts still have, we still need some other applications or
service to be multi
-
lingual like e
-
mail so since when other Internet service like e
-
mail will be
possible with no Latin script I think we will surely be in the era of truly multi
-
lingual Inter
net so
what do we think about these other services apart from Web or ID Ns in general. Thank you.

>>: I think that I know that we are very, very short of time. Was there something from the
remote that, did you want to just take the floor for that? Se drik
did you want to take the floor
for the remote participation questions?

>>: We need a microphone to speak.

>>: While we are get in the microphone
--

okay.

>>: Question to Mr Ja

nis, in the next generation of the Internet going to foai cuss more on the
langu
age aspect than the technical issues.

>>:
Janis Karklins :
I am not sure that I really understood the question. I think if we look back in
about 10 years, starting from our engagement in the world summit, the landscape of issues we
are talking about is cha
nging and in 2003 to 2005 until maybe 2007, the access and
infrastructure development issues were really on the top of the agenda and it was basically the
only concern if I may say of Internet users, particularly in Africa, in Latin America and in Asia.
To
day with the boost of mobile technologies, we see that less and less people are concerned
about access. We are still there concerned about prices of access, what they have to pay but not
that much that they would not have possibility to access Internet but

what we see on the rise,
are the questions related to actual use of Internet and in that sense the answer is yes, I think in
the future we will be talking much more about use of Internet in general terms rather than
access to the Internet because that iss
ue is about to be resolved for the good.

>>: Thanks a lot, Janice and I like to give the floor to Emily for a quick fine word.

>>: Thank you very much Giovanni and thank you to all of the panelists and to, as we started
late inevitably everything has got s
cfee standard I have a feeling that we could continue our
discussions in a much more interactive way hat meek more time so thank you fewer patients on
that. I think odd we have heard today is that the first steps in creating a more multi
-
lingual
addressing

system are there but the journey is certainly not over, the house does not have the
roof on, to mix my metaphors horribly. We seem to have hoant in on the user experience as the
next great challenge, the difference between preference and trust and the Gul
f that lies between
the two, the role of the CCT L D registry as providing a local hub that can enable and foster a
really trieving environment for multilittle gyalism I think in the next few years we will see
whether these stand or fall. There is
--

there

are great challenges but I think we should also as
Janis encourages us to do we should step back and see an awful lot of ground has been covered
in a very, very short time tnks to the goodwill and co
-
operation of many multiple parties so
thank you very mu
ch.

>>: Yes, Janis, final word from Janis.

>>: Thank you and I would like just to say how important it is to continue this analysis and dai
gathering. Data gathering probably is the most difficult part of this exercise and I would like to
encourage all tho
se who are interested and all those who have who are working in this area and
collecting either empi ri L C data or just a survey of public opinion maybe to join us in this effort
an share that information to the extent possible that we can make this analy
sis and conclusions
on the basis of most possible information and certainly our intention is to continue this work
also in the future. We will, as I said, present the report to UNESCO member states in February at
the end of February during the business rev
iew conference and I hope that in the meantime we
will already start our advocacy work based on the findings of this rep so thank you very much
for your interest an see you next year.

>>: Thanks a lot Janis, thanks again, Markus, Janis, Emily, to contribut
e with the appendix and
also to se drik for the remote participation, tanks a lot everybody.
--

thanks a lot everybody.