Welcome to the United Nations | Department of Economic and Social Affairs

FINISHED COPY

 

NINTH ANNUAL MEETING OF THE

INTERNET GOVERNANCE FORUM 2014

ISTANBUL, TURKEY

"CONNECTING CONTINENTS FOR ENHANCED MULTI‑STAKEHOLDER INTERNET GOVERNANCE"

 03 SEPTEMBER 2014

6:30

WS 11

 

LANGUAGES ON THE MOVE: DEPLOYING MULTILINGUALISM IN THE NET 

 

 

***

The following is the output of the real‑time captioning taken during the IGF 2014 Istanbul, Turkey, meetings.  Although it is largely accurate, in some cases it may be incomplete or inaccurate due to inaudible passages or transcription errors. It is posted as an aid to understanding the proceedings at the session, but should not be treated as an authoritative record.

***

 

>> GIOVANNI SEPPIA: Good afternoon, everybody.  We'll start in three minutes.  Thank you.  Sharp. 

Good afternoon, everybody.  And welcome to the workshop No. 11, languages on the move, deploying multilingualism in the Net. 

(technical difficulties)

>> EMILY TAYLOR: ‑‑ the system's hydrogenous networks and languages, and I think we should be thinking in terms of the addressing system in the same way, that diversity in this sense is a great strength.

So what is an IDN, an Internationalized Domain Name?  I think for many people in this room it needs no introduction, but I'm just going to explain it anyway.  You see on the slides two different examples of IDN local language domain names.  Now, for many years it wasn't possible to register domain names in non‑Latin scripts all the way through the chain.  And so you see at the top you have paradigma.EU, you have the Greek script example .EU in Latin script.  That's called, in the terminology, a second‑level IDN.  But from 2009 or so it became possible to register in local languages right through the domain name, and that's what you have at the bottom example.  You have an example from the ‑‑

(technical difficulties)

‑‑ a summary of where we are in terms of the stats and the numbers.  I'm going to talk about IDN's local knowledge domain names as drivers of linguistic diversity and how that's showing through.

(technical difficulties)

‑‑ the evidence when we look at the content ‑‑ the language of content associated with these local language domain names and the patterns of registration and hosting, and they come through very clearly to demonstrate how local they are.  I'll also just touch very briefly on the issue of universal acceptance or, in other words, how these actually work or don't work, and Mark McFadden will deal with that in more detail. 

We'll also cover industry opinions, what the registries and registrars think about these ideas and local language domain names and also lastly ICANN's new gTLD programme and whether that has fulfilled the promise of introducting more linguistic diversity.  Here's a clue.  It hasn't. 

There's an awful lot in the report that I'm not going to cover because of lack of time but I really urge you, if you're interested, to read the record.  There is particularly a focus on the experience in Arab states, which we have colleagues who will cover their experiences and what ICANN is doing in that area as well as local registries and also country case studies.

So brief headline of the numbers.  There are now 6 million domain names in non‑Latin script in the world.  That's an increase of roughly 800,000 since last year.  2% ‑‑ just 2% of the world's names are in Latin script, but the growth is quite phenomenal.  There's been over 200% growth in the last five years, but it doesn't take very much thought to understand that when you're starting with low numbers and reach it, you can ‑‑ you can get into triple‑digit percentage growth quite readily.  In terms of the growth rate of the different levels of the I.D. and local language domain names, the things that are really growing very strongly at the moment are the full IDN.IDN, in other words the Cyrillic script, where you're getting the same uniform language all the way through.  Those are showing growth of over 100% in the last year.  GTLD IDN growth has been 46% in the last year, but that's mainly because a couple of registries have launched and that normally gives a very healthy spike. 

In contrast, the gTL ‑‑ sorry, the country codes IDNs which tend to be at the second level have shrunk a little this year.  We're now aware of 50 registries that are offering Internationalized Domain Names at the second level.  26 all the way through at the top level, and two new detail G's as part of the ICANN programme.  Domain names with content associated with them. 

A lot of those was as a result of the cooperation with VeriSign.  We were looking at the common net ones, we looked at the EU ones and all the ones in the open zone in the gTLDs.  Of those that had Web content associated with them that's what you see on the bottom line.  And this is with the same chart but another bar added.  And so the headline is it's different.  With local language domain names the spread of languages associated them is different from the general picture, and you're getting much more of languages like Chinese, proportionately more Japanese, Russian and so on, Korean as well.  So that tells us something, doesn't it?  I think it does, anyway.

The next slide I want to show is when you look at the registrations of all top level domains in the world this is borrowed from Oxford Net Institute who have done an analysis of the ‑‑ the deeper the red color the greater the Internet population is.  So for the U.S. you've got over 100 million users, the same with China, and the larger the circle is, the more domain names there are registered in that country.  And what they did was they took the top‑level domain, the country codes, they just took the numbers and with the generics, which aren't naturally associated with any country, they took the country of hosting.  So this is a nice ‑‑ I think this is a robust picture.  And what you see I think is North America and northern Europe in brief terms.  China is there really because of the size of its population, not because it's really starting to fulfill its potential.

And we did the same exercise looking at Internationalized Domain Names or local language domain names, and again, the picture is different.  What you see is that these local language domain names are registered in the country where you would expect them ‑‑ the countries which have most need, are least served by the Latin scripts and have most need for linguistic diversity in the navigation system, so Russia, huge.  China, huge.  And you're also getting, you know ‑‑ you're going east and you're going a little bit south as well.

So the other thing that we looked at was how the script of these IDN local language domain names correlates to the language of the content associated with them.  So if there was ‑‑ if there was no relationship whatsoever, you would expect a completely random pattern, so you've got these, you know ‑‑ these domain names in whatever, Russian, Korean, Chinese language.  If there's no relationship whatsoever you just get a jumble of Web site languages associated with them.  But that's not the case at all.  So this chart is looking ‑‑ in the colors ‑‑ is Arabic script.  If you have a Persian language Web site or an Arabic language Web site in our pool, which all have IDNs, they are all associated with Arabic script domain names.  The same goes for Greek, say, or for Russian, Bulgarian.  These ‑‑ these language ‑‑ Ukrainian are only associated with Cyrillic script domain names and no others.  The same ‑‑ Korean.  Korean is only associated with Hangul script domain names.  There is a really remarkable correlation.  It's a remarkable correlation.  It's not a remarkable fact because anybody would be able to tell you this, but it's remarkable that the data does ‑‑ does show it.  Where you start to get random patterns are over on the other side of the chart where you're getting into kind of English language.  That's associated with all sorts of scripts, domain names, and you can kind of understand why. 

The same with French, that these are often used as the languages of translation on a site.  But ‑‑ and then there are others which I think that the others like Portuguese and Danish, I think that those are more a product of automated translation tools not quite getting it right, although they're fantastically accurate in most cases.  In a minority of cases they're startlingly inaccurate.  So when we look at the .eu names, same patterns.  You have the Cyrillic script domain names associated with Bulgarian and Russian sites, Greek in Greek and everything else with Latin scripting, exactly what you would expect. 

The same correlation is found when you look at where these domain names are hosted.  Now, a few years ago the Internet Society‑UNESCO and the OECD collaborated on a project which showed there was a very strong correlation between where ‑‑ between local servers and local language content.  Waaah!  Sorry.  And what we've done here is looked at the script of an IDN and the country where that IDN is hosted.  So the orangey kind of figure ‑‑ the orangey color that you see over in North America indicates mixed script, so no script is a winner, and that's ‑‑ you can see in places like Germany, U.S., Canada, where you have very strong international hosting countries.  But just look at the story over to the east of the map.  You see Cyrillic script domain names.  This is looking primarily in dot‑com and the open zone files.  We didn't have access to the ccTLDs but the Cyrillic script names they're hosted in Russia or hosted in lesser CC IDN countries.  Hundred script, China, you see Heragon, in Japan you see Hangul only in Korea, that's it.  Thai in Thailand, that's it.  So you've got this really strong association ‑‑ people are hosting where they are registering their local language domain names.

So the next topic we'll be looking at is, you know, how do these things work, how well are they working, so‑called to universal acceptance, and that we define as the relative ease of use, predictability and memorability of IDNs in application servers.  This is a key issue because domain names don't have to do an awful lot.  They have to work on Web pages, in email.  They have to work in other contexts but if they don't work in those then there is not an awful lot of point in having them and it's a difficult problem to solve.  It goes right through the Internet stack and is actually a wider issue than we've been thinking as we've been working on it in the last couple of years.  There's some notable progress in the last year.  For example, Google has announced that gmail is now supporting sending and receiving internationalized email addresses and CN ‑‑ has done a lot of work in open source but much remains to be done.

When we look at usage, and looking at .eu INs, and I think this is a good message ‑‑ it might be the universal acceptance or the impact of mixed script domain names because under .eu, the Latin script, you've got three scripts happening.  You've got Latin with accents, Cyrillic and Greek, and you can see immediately that Cyrillic and Greek, the mixed script .eu domain names are less used as they involve changing keyboards.  You know, they're ‑‑ generally speaking the usage rates of IDNs is lower than general.  But human beings being what they are, creative and good at working around things, are adapting even to limited functionality, and I think that this for me really shows the utility and the appeal of local language domain names, and this was provided to me by Irina from the Russian registry, is that people are using the dot RF domain names in advertising because people remember them.  They understand them and they remember them, so they see them as they're passing by, they see them really fleetingly, they remember them, they go on‑line and like the green truck in this picture is going to redirect.  So the people providing it use the domain names for what it is, very memorable to the local user and then use the ASCII domain names as the primary domain through a redirection.

Industry opinions, I'm not going to go into in much detail.  We send out a survey each year.  Generally people are getting more depressed.  That's about it.  The confidence levels are getting lower each year, which I think is a pity, because if the industry not fully confident in these as a product, that will impact on ‑‑ on ‑‑ on their choices for investment and on, you know, their belief in it, on impact.  We seem to be locked in a negative cycle.  You've heard me saying this for a couple of years.  Lower user uptake is a result of poor user experience, and if you don't have ‑‑ if you don't see these around, then not many people know they exist and so that perpetuates this negative cycle.

The new gTLD programme, nearly 2,000 applications.  It was ‑‑ it was supposed to be meeting unmet needs, multilingualism is obviously an unmet need.  Only 100 or so of these 2,000 applications were in non‑Latin scripts, but I went through and looked at the meeting of the new gTLD endings, I can't believe I did this, but I did it, so that excluded brands and excluded meaningless abbreviations.  But we've got on the top languages in the world with over 50 million speakers, and you can see a nice diverse and fairly evenly spread picture.  And the bottom layer is the results of this, the meaning of new gTLDs in the ICANN programme and in brief it means that 90% are either in English or meaningful in English.  So, in fact, it's actually less diverse.

Before we can get mass uptake a lot of things need to be in place, and universal acceptance is identified by our industry people as the key issue as well as user awareness.  So there is continued growth.  IDNs ‑‑ local language domain names definitely support linguistic diversity and cognitive factors that Giovanni referred to, like ease of recalled context, understanding, they really support the need for local language domain names.  Universal acceptance is critical, but it's ‑‑ and it's slowly improving.  There's more to be done, but there is huge growth potential in this area.  Huge.  And so we must keep working at the different layers of the stack to realise this potential.  Thank you.

>> GIOVANNI SEPPIA: Thanks a lot, Emily, and before leaving the floor to Pat Kane of VeriSign, I'd like to ask you if you have any question about what Emily has just introduced, including, you know, you can complement what Emily just said, as she mentioned several registries, so feel free to step in the discussion and take the floor, just raise your hand.  I see there's a question from Peter.  Please ‑‑ can you please introduce yourself as well for the remote speakers and for the scribes?

>> Thank you, Giovanni, my name is Peter Van Roste from CENTR, and congratulations.  I think it's a really good report and not just because Giovanni suggested that we point that out (laughter).  And it's a very ‑‑ it's a very useful yearly update. 

I really appreciate that.  I have one question.  I was fascinated by the reduction of diversity on the new gTLD programme.  So what happened between application time and the outcome of that process?  Is it ‑‑ all the IDNs supplying for dot Web in Chinese ‑‑ so from 10 being reduced to 1 or what happened there?

>> EMILY TAYLOR: What I did was I looked at the applications rather than how they've gone through, because I felt that that was actually quite meaningful in terms of what people felt was going to be appealing to the market.  So you're right, there were like ten or more applications for .web or .app or so on.  And so each of those counted as one.  So this doesn't show you what we're getting at the other side, because it's still too early to tell at that stage, so this is just simply looking at the ‑‑ the applications themselves.

>> GIOVANNI SEPPIA: Yes, please.

>> AUDIENCE: My name is John Dada from Nigeria.  I find this study very fascinating, and more fascinating to think that linguistic diversity can be a strength.  My personal experience is it's been used very divisively in places.  Are you familiar with a study done in other languages, by ‑‑ African languages, for example, by UNESCO anywhere else?

>> EMILY TAYLOR: Not on the ccTLDs, and I'm hoping that as this project continues that through the cooperation of the African ccTLD registry we can get more data because we're very reliant on personal relationships and getting the data from people individually.

>> INDRAJIT BANERJEE: I think your comment was interesting as far as UNESCO is concerned, although we partner with EURid for the support every year and we find it extremely valuable, we also have the only normative instrument in the world on the ‑‑ it is called a recommendation on the promotion and use of multilingualism in cyberspace, and the reason for which in spite of all the difficulties we encounter in getting a normative instrument accepted in UNESCO, or any other agency, for that matter ‑‑ for example, it took us ten years or more to get the universal declaration on cultural diversity adopted, ten years to recognize that cultural diversity is important ‑‑ and so until the multilingualism we have this binding normative instrument, and what we do is we ask all ‑‑ because it's a normative instrument we have the possibility of requesting all our members to ‑‑ including African Member States, to submit a comprehensive report on what have they done to promote local content, multilingualism in cyberspace in their respective countries, and that report gives us a fairly good sense, and very it's clear that there are great disparities.  Some countries are really active in promoting local languages, other countries, even within Africa, are not doing as much.  So that gives us, then, a chance of going back to our member countries and saying that, why are you are not doing enough?  You know, it's a binding instrument. 

In fact, in line with this, if Giovanni will permit me, we also have a very interesting platform called the atlas of worst endangered languages.  This is an amazing tool.  For the moment we have 2,000 languages which are documented in the tool.  It can give a fairly accurate picture of which other languages ‑‑ which are in the threat, in which parts of the world.  It can also give you a fairly good indication as to which languages will disappear, because already as per recent reports, 43% of the 6,000 existing languages will be ‑‑ will be gone by the end of the century.  So that's almost half the languages gone. 

So for us to really get back to what ‑‑ Emily's excellent presentation and report, is that for us we see IDNs as a building block, and for me personally I feel, coming myself from a developing country, that it also gives you a sense of ownership, that I have my country's name on the ‑‑ on the site.  But then again, as Emily pointed out, there are still lots of gaps.  I think on the software side, on the services side, they are not really tuned to take on IDN, and I think that would be the next challenge, and I also feel that having an IDN is okay, but if you don't populate it with rich local content, it's just a name.  Thank you.

>> GIOVANNI SEPPIA: Thanks a lot.  Please.

>> AUDIENCE: Good afternoon, all of you.  I just wanted to pick the last point that Emily had shown, that there's a huge ‑‑

>> GIOVANNI SEPPIA: Could you just say your name and ‑‑

>> I am Sayed Cosi from India.  One of the last points in the last slide I saw was a huge potential for local language content.  The scope is wide enough.  And it was simply ‑‑ last week, I guess ‑‑ yes, last week only comment of India has come up with .harat that is ‑‑ that domain ‑‑ it has come up.  So since, you know, we are ‑‑ two things.  One, we are very excited because we have now our own domain name in our own local language and all, but at the same time we also can see the challenge of how to populate it in terms of ‑‑ because after 65 years of independence we have seen the other domains taking the space of the digital identities, digital presence and all, with content, populating content and that exercise is gone ‑‑ a huge exercise has already gone over in terms of content, domains and all.  And India dot ‑‑ domain is in English. 

So the challenges we can see because I know for ‑‑ for every stakeholder is how to populate this domain, especially the new one that is the new government has come out with this domain and all.  So what you see as an expert yourself, since we have seen this entire scenario coming up, how do you see it, as a policy challenge, as an implementation challenge?  So can you ‑‑ what are the two things that you see that also we need to focus as a government where multi‑stakeholders can take a role in populating those domains at a national level also.

>> EMILY TAYLOR: I think that the two issues that are primarily important is the state of the general ‑‑ you know, there are numerous factors that go to mass uptake, and if you're just talking about domain names in general, they will be things like how large the population is.  You're in good ‑‑ things like broadband penetration, literacy levels, a whole raft of factors go into that, but then there's also that the ‑‑ the way that the registries are run themselves, how ‑‑ if you want volumes, there are lots of reasons to not have liberal policies, which are perfectly fine, about but if you want high volumes, low price liberal volumes for registration, very strong network of local registrars are essential, and the sort of ‑‑ something that we've been talking about in context of the Arab states where there is ‑‑ you know, there is great potential to be fulfilled.  There is ‑‑ there are other factors that are ‑‑ that are inhibiting uptake, such as general immaturity in sort of the downstream markets, the cost of creating content. 

That's probably a longer answer than you wanted, but the ‑‑ you know, the ‑‑ the main thing is to have the domain name and to have it usable, and I think that that's going to kind of be covered in others' comments.  But before these domain names can be used seamlessly in every way that you would expect to use another domain name, then the uptake is going to be inhibited, and as long as you're expecting that and prepared as other country codes ‑‑ as policy makers, advocacy, what goes across all of the country code registries who have adopted IDNs, local language domain names, is that they are absolutely tireless in advocating them and badgering people, annoying people to make them work better and to really get the uptake working.

>> GIOVANNI SEPPIA: Thank you, Emily.  Yes, please, here.

>> Thank you.  Thank you, Emily, thank you, Giovanni.  May I introduce an illustrative example using (?) complement, Emily presented regarding the relationship between local content and IDNs.  We have an interesting example.  My observation on it is .jp has two spaces where IDNs can be registered, as well as ASCII labels.  One is at the second level in the form of something.jp.  And the other is at the third level in the form of something.prefecture name.jp, such as (?) Tokyo.jp. 

As to the second level something.jp registration, 13% of the registrations are made in IDN format.  On the other hand, 25% of the registration are in IDN format, when registered as third level something.prefecture name.jp.  If I would explain the situation pretty roughly, I would say local content under the prefecture name ‑‑ local content can be more suitably presented in IDN label than ASCII label.  Further rough guess may conclude that if more local content is coming in the near future, for example, from Arabic or from China or from India, more IDNs will be used, of course under the condition that the IDNs can be technically used.  Thank you.

>> GIOVANNI SEPPIA: Thanks a lot.  Really good point.  Thank you.  Yes?

>> MANAL ISMAIL: Thank you, Giovanni.  This is Manal from Egypt, and thanks, Emily, for the very informative report.  Just to reflect on what you said on the Arab region and of course I cannot claim speaking on behalf of all the other region, but at least many of them, and Egypt is one, I think challenges that has to do with the uptake of the IDNs, primarily starts from having a weak DNS ecosystem, or industry in general, even for the ASCII domain names, so we don't have the ecosystem in place in terms of policies, dispute resolution mechanisms for scripts like Arabic, where you do not have this expertise really in other developed countries or Latin‑speaking, English‑speaking countries, so you have to develop your own expertise in things like this.  So it's the ASCII DNS ecosystem, the universal acceptance and user experience, related industries such as local content, of course, ePayment, penetration of ePayment on‑line services and things like that, and finally penetrating new markets who are really suffering the language ‑‑ they have a language barrier and might not even be on‑line as we speak.  So they are really new market, new potential market that we have to work on, even bringing them on‑line, and they would constitute a good portion of these new potential markets.  Thank you.

>> EMILY TAYLOR: Thank you, Manal.  I think those are excellent points, and for me it's very interesting seeing almost a diverging path.  In the Arab states where you have a real burgeoning of local language content, particularly when people are using social media, that ‑‑ there's a fantastic study coming out of Dubai on the social media in Arab states, and showing that people, of course, love to make their comments in their local language.  It's, again, not really rocket science.  You would expect it, but there's an enormous enthusiasm for it and a lot of local content being created, and it makes me worry that if we can't actually bring the domain name system back into line with the rapid rate of creativity that's going on for content, that people will seek other solutions, and that would, I think, be very detrimental to the unified Internet.

>> GIOVANNI SEPPIA: Thank you, Emily.  Please, Ram.

>> Thank you.  This is Ram Mohan, I'm with (?), also on the board of ICANN.  In India one of the things that I've been observing is that you see English language or ASCII‑based domain names with local content, and it appears at least in that region that the trend is that the gateway to get to the content is almost irrelevant, because you can type in that ‑‑ whatever you're looking for, you can type that into a search engine in your local language, and, you know, Google, for example, will bring back the correct site, even if the site is actually written in ASCII.  So there seems to be some level of divergence between content in the language and the actual identifier for the ‑‑ the place to go to that content also needing to be in that language.

>> EMILY TAYLOR: I think that's absolutely right.  And one of the ‑‑ it's quite an irony, really, because one of the reasons that the domain name system developed in the first place was because domain names were much more memorable and meaningful for individuals than the numbers that they translate to, and what you're describing is almost a sort of a similar work‑around, is that search is kind of stepping in.  However, I do think that domain names are so ubiquitous, is perhaps one of the things that comes through in discussions on universal acceptance or how these are used, is that, you know ‑‑ that domain names are, in fact, everywhere. 

When you're trying to adapt to include local languages, you have to think about search results, you have to think about user names and user accounts, certification, DNS policy, the works.  I hear this argument about search a lot.  I'm not completely convinced is that it is a substitute, though, because I think what it does is it relegates people who don't have that understanding of the ASCII to having a second best in a way.  You saw with the Russian vans, that there's such an utility for having something direct and navigation, and that while search is fantastic and the technology is amazing in getting to where you didn't think you knew you wanted to go, where you know exactly where you want to go, it's really important to be able to remember and type in a name, in my opinion.

>> GIOVANNI SEPPIA: Yes, please.

>> AUDIENCE: Thank you.  My name is Satish Babu.  I'm from India as well.  I have three quick comments.  The first is the fact that we have taken so long to operationalize IDNs, that content has actually gone ahead.  For many communities in India the content is already there, using Latin labels.  And to get them to come back to ‑‑ I mean, there is a low incentive to get them to come to IDNs.  So this is one issue that, you know, I see ‑‑ the main content provide is that newspapers, blogs, the government, and most of all Wikipedia, and all of them have actually been ‑‑ have been quite some time with Latin labels.  So there's low incentive to now get into a new regime with IDNs.  I don't know how we combat that.  The second is the issue of usability.  We have complex scripts, which are very difficult to I don't, especially on tablets and mobile phones.  It's very hard to kind of squeeze these keyboards on to these devices.  So that's another problem.  I don't know how we're going to fully overcome that.  Especially considering the fact that the young people are using SMS in the Latin script.  They are not moving into any of the ‑‑ the so‑called local scripts for SMS.  So it is ‑‑ the next generation will have even more problems in using the scripts.  It's not a good thing culturally, but that's where we are.

And finally I also wonder at times, although I fully see the reason for IDNs, whether we are fragmenting the Internet further with IDNs.  Thank you.

>> GIOVANNI SEPPIA: Thanks a lot.  Would you like to address ‑‑ no, just because there is apparently two more people would like to take the floor, remote participants, and so, you know, time is not on our side.  So yeah, please.

>> In fact, this is coming from India, the third Indian speaker to take the floor.  You can see we are a democracy.  (laughter) And a very lively one at that, but the point is both Mr. Ramon and the legal speaker, I didn't get your name ‑‑ I think it's very important that we do not look at this issue in a homogenized manner.  A country like India functions very differently from any other country in the world.  For example, a huge percentage of my generation cannot write (?), so for us if you tell us it's a script issue, et cetera, the one you mentioned, it's critical.  I mean, a lot of people are writing to me in Bengali, which is my mother tongue, but using Latin script.  So we've created a new language of sorts.

Second point to respond to first issue is, like Ms. Ramon said, I think while waiting for IDNs to come, we have become also very innovative, using Latin script.  We have put all our local content up and there is a lot of catch‑up to do. 

And therefore the last point that you make, again, I take it not only from fragmenting the Web perspective, but what will be the dividends today when people have moved ahead, ignore their favorite Web sites and get all their local content ‑‑ what will be dividend in this context?  Will we really have moved so far ahead with I don't know, 100 million users in India today, in terms of getting ‑‑ moving ahead with this idea. 

So these are good questions but very specific to India, and with, I don't know, several hundred or several thousand languages, the matters become far more complicated, plus it's going to be a large pool of English‑speaking people, and that again makes things very different.

>> GIOVANNI SEPPIA: Thanks a lot.  There is a question from the remote participation. 

>> AUDIENCE: I'm Cafu from Ghana.  With a number of Africans not literate in the English language, how can African languages be incorporated into the current IDN scripts to enable developers to build local apps that are relevant to local users? 

>> GIOVANNI SEPPIA: We have Fahd from ICANN who would like to take up the question and eventually Mark for a more technical perspective.

>> FAHD BATAYNEH: Thank you, Giovanni.  Good afternoon, everybody.  My name is Fahd Batayneh, and I work for ICANN.  One of the things I do at ICANN, I actually work as a stakeholder engagement coordinator for the Middle East, and actually in the Middle East we have a strategy, a Middle East strategy, and one of the outcomes of the strategy was the birth of a task force on Arabic script IDNs.  And I would answer to the gentleman from Ghana, I can speak about languages ‑‑ let's say languages ‑‑ African languages that you would use the Arabic script.  So I believe in Africa there is a huge number of languages that actually use the Arabic script ‑‑ I mean, there is a huge list of names that one would not imagine, actually.  And so how could Africa get more involved in, let's say, the multilingual Internet and probably domain names, let's say Internationalized Domain Names, I would say is to be part of such initiatives.  So actually within the task force on Arabic script IDNs we are suffering from actually getting folks from Africa to actually be part of the discussions related to the Arabic script.  I guess it would be the same case for, let's say, other languages that are using other scripts.

>> MARK McFADDEN: This is Mark McFadden.  I wrote part of the universal acceptance part of this, and what I would also say to the remote participant from Ghana is that there is an immediate opportunity at the ccTLD level to have an impact in the near future in terms of getting other languages into ‑‑ into IDNs.  I think that that is a significant short‑term opportunity. 

One of the things is in terms of IDN applications in the new gTLD round we saw very few from Africa, and I would think on a policy basis working in ICANN environment, whether it's through the GAC or through the GNSO, ensuring there are mechanisms in place to give, if not enhanced treatment, the ability for the African community and the people who have languages that are underrepresented in the IDN community a way to have better results in the second ‑‑ or future rounds of new gTLDs.  I think that would also be an important development, because as we see in this year's report, the IDN component of the new gTLD programme was fairly disappointing, both in terms of the languages represented and also, as Emily pointed out in her remarks, the meanings of the strings that are represented. 

And so finally for the person from Ghana, I think engagement in the ICANN process for future new gTLD programs is essential.

>> GIOVANNI SEPPIA: Thanks a lot, Mark.  Yes, please.

>> AUDIENCE: My name is Parag Mehtani from Sudan.  In fact, I am in the FTLD ‑‑ I am the head of that working group.  So my responsibility more is about developing using of the IDN in Africa.  So, in fact, there is ‑‑ I also am a member of the Arabic DP, the (?) planning for Arabic language.  So my concern is I am focusing on using Arabic script in Africa and ‑‑ in African languages.  In fact, there is more than 40 languages in Africa using Arabic script.  Sometimes they are not the official script for the country, but it's all that is there and it's already used by a lot of people there.  Some problems were (?) problems in Africa, sometimes some characters are using, for example, in Arabic script, they are not even in the Unicode, so it's very ‑‑ the Arab Latin ‑‑ very behind.  They need to approve these characters and Unicode (?) think about it, about IDN.  And that's my comment.

>> EMILY TAYLOR: Thank you for that comment.  If I could make just very quickly, I think that your final remark about some of the characters used in African languages are not able to ‑‑ they're not present in Unicode, which means that they're not possible to reflect in a domain name at the moment.  I think this highlights the importance of initiatives like the task force on Arabic IDNs which Fahd spoke about and I'm lurking on that list and can see the care that all the participants are taking to ensure that they are actually scanning all of the languages that they need to be thinking about.  But of course it relies on the sum of knowledge of the participants. 

But this is an incredibly important piece of work, because, okay, it's going to be generating rules to create domain names, but because a lot of the participants are actually managers of domain name registries, you can anticipate that those rules will then be ‑‑ you know, they will be pulled across into the registry policies and will ensure that the domain names will work outside of national borders or outside of linguistic borders so that they will resolve universally.  Thank you.

>> GIOVANNI SEPPIA: Thank you, and I'd like now to give the floor finally to Pat Kane of VeriSign and immediately afterwards to start out and we may take more questions from the floor.  Thank you.

>> PAT KANE: Thank you, Giovanni.  I do want to thank Giovanni and Emily for giving us the opportunity to work with them on this project.  It's really been a pleasure, and one of the really great things about data is the more data that you find, the more questions you have about the data.  I know that when I ‑‑ when I read one of the early versions of the document, some of the work that Mark had done, we went through and we said, well, what about this?  Do we have this data? 

Sure.  I am now in control. 

So, you know, one of the things that ‑‑ about the data, we asked a lot of questions about it, and Emily was going, I don't think we have time and we got to cut it this way and do that.  So we've got plenty of work we're going to do for next year's version of the documents, so I'm looking forward to continuing working with both Emily and Giovanni on this.

So I'm not going to take a lot of time, just a couple of slides.  One, this is the disappointing slide.  This is the slide that has taken us forever to get to where we are today.  I can remember back in 1998 and 2000 when we had an argument about IDNs and the DNS itself as to whether we were going to have ‑‑ remain with ASCII in the DNS or actually use something like UTF‑8 or UTF‑16, which actually would have made this problem that we have with universal acceptance almost completely go away. 

And for those of you who don't really know what ASCII is, it's the American Standard Code for Information Interchange.  We're going to geek out on this all day today. 

So, as Emily talked about, there are 6 million domains, 6 million IDNs today, which the slide is not updated for that.  We launched in dot‑com, November of 2000 ‑‑ we launched IDNs, and so the state of affairs for IDNs has not change dramatically since then and it's really the universal acceptance, and we're trying to do a lot of work within organisations like W3C to put together an open source set of objects and tools so people can actually build Web sites.  I mean, it's great that mail works.  It's great that browsers work, you know, that's navigating, but when you want people to use as their primary email address an IDN, you want to be able to use that as your user I.D. when you go to social networks.  So when you put in your user I.D. and you type in, you know, XN‑‑BR549.com, what am I doing, right?  And so there's still a long way to go and we continue to work on that.

What are also important dates is the ICANN new detail (?) programme.  It was approved at the Paris ICANN meeting in June of 2008, back when we didn't have a whole lot of IDN traffic and we had smaller populations around the globe.  VeriSign applied for 12 transliterations of .com and .net in local scripts and every year I've had to go back and rejustify my budget for that.  And when I do that I've got to go in and talk about what the market size looks like. 

And so the most disappointing thing this past year is when we went into China to where ‑‑ six years ago China had an on‑line population of about 125, 130 million, where today it's over 650 million is their on‑line population today.  We went through and did the market study, the focus study and the answer was, those things are for old people.  So people ‑‑ to the point that's been made a couple times already, people are very, very comfortable using as their identifier an ASCII domain name, an English character‑based domain name.  So, yes, disappointing.

So just to follow up on one of the other slides that Emily talked about, 2% of the IDNs today are IDNs.  1% of those ‑‑ or 50% of those domain names are non‑Latin.  So when you say Latin, they are extended Latin characters so that includes the enyes, the sedias, accented vowels, umlauts, those characters ‑‑ 50% of the 6 million IDNs are some Latin‑based language. 

We'll jump to this.  When we talk about growth ‑‑ two slides about dot‑com specifically.  This is the growth of active dot‑com domain names that are IDNs.  So we've got over ‑‑ we have over 1.2 million IDNs and almost 700,000 of them ‑‑ this is 2013, so over 700,000 result in active live content, which is good.  That's a good percentage in terms of what are being used.  But the one that I am excited about, and I use this one a lot with my CFO, is when we talk about how domains are being used that are actually content.  So this is over a nine‑month period in 2013, for three different scripts.  And so for ‑‑ and this is done as a percentage.  So as you take a look at Hebrew, this would be the complete set of Hebrew domain names that are IDNs ‑‑ I'm sorry, Hebrew content in dot com.  So of those that are using an IDN, that's that.  And those that are using ASCII‑based domain name are these, and this column here shows domain names used as a redirect.  So basically these again are redirected to Hebrew content. 

And the important thing to look here is it's growing.  So in this nine‑month period ‑‑ this is really towards the end of last year, new gTLD contracts are being signed, people are getting excited about the programme, so even though we have a lot of people that are comfortable, younger populations are comfortable with the ASCII scripts and not gravitating ‑‑ especially in larger markets, they're not gravitating towards the IDN visually, we do see more utilization.  And to Emily's point earlier, more utilization drives more awareness, drives more adoption, drives more utilization.  It comes a becomes circle at that point in time. 

I'm really excited about the Korean market.  When we go into Korea, Korea is one of the places that is real ‑‑ I mean, everybody is really passionate about their language.  One thing I've learned in doing this for 14 or 15 years ‑‑ and just one point of note, when you carry carp in the room to actually have the original IDN guidelines group here today.  Anyway, how are you doing, Mohamed?

I don't know where I was going, but anyway ‑‑ so Korea, the reason I'm compassionate about Korea is Korea hates English.  They hate seeing the English language, English characters all over the place.  So we're starting to see the IDNs ‑‑ is that right?

>> EMILY TAYLOR: Yeah.  (laughter)

>> PAT KANE: Every time I go there I talk in English for about 30 minutes and people go, we're tired of hearing English, and I sit and drink a beer.  It's not a bad thing but it's what I do. 

But this is very exciting, but you see so much more usage in dot‑com domain names that have Korean content using an IDN.  So the utilization statistics are showing that there is growth and there is improvement in the legacy operations.  With the new gTLD programme we have two or three Chinese new gTLDs that are out now.  They're seeing growth.  A large percentage of their domain names today are reserved for the government, but we're seeing utilization and we're seeing growth.  I'm going to be excited when Gong Su and Hong Lo come on line because they've been regionally resolving domain names inside of China for a long, long period of time, and we should see some instant adoption at that point in time.  That's really about it, I think.  Oh, universal acceptance, it's still a big problem.  And that's it.  (laughter)

>> GIOVANNI SEPIA:  Sarah?

>> SARAH FALVEY:  Do I have to follow that?  Sarah Falvey.  I'm at Google and I don't have slides.  I'm just going to sort of talk about sort of what we're seeing, and I'm really excited about IDNs.  I don't ‑‑ this is a depressing group.  This is ‑‑ you guys, I think it's going to change, so I don't think we should be depressed.  And I think it's for a couple of reasons.  I think, you know ‑‑ I was on ‑‑ I've been on this panel, I feel like, every year for five years now, and when we started talking about IDNs ‑‑ huh?

>> AUDIENCE: (?)

>> SARAH FALVEY: When we started talking about IDNs I think it was the fault of our community because they were fast‑tracked within ICANN, we developed technical standards around them, and I think most of us thought, oh, my God, the work is done, this is awesome, this is so great.  And it took, I think, a lot of conversation with folks in various regions to ‑‑ to realise that there is this adoption issue that has really been the barrier for why IDNs aren't being used more broadly.  And I do credit, frankly, the IGF with a lot of talking about adoption in all the different cases and all the different problems with it, because it's really spurred, I think, a lot of different working groups and things like that getting going.  I can speak personally from Google's perspective, you know, as it was mentioned earlier, we just announced that IDNs will resolve in gmail, and a lot of that is because, you know, we heard people talking about it in forums like this, and we sort of realized kind of what the problem is, and we're going to be ‑‑ we're still doing a lot more work. 

There's a lot more announcements that will be coming in terms of how far IDNs will work in our products and obviously we're just one company but I will say one of the great parts about the new gTLD programme is even though there weren't a lot of IDNs that were applied for, there were a lot of companies that ‑‑ and Civil Society groups and governments who maybe didn't realise kind of the adoption issues that are now intimately involved in domain names in a way they never have been before, and they're definitely paying attention, I think, a lot more than they ever have.

We're one of the companies that applied for one of the generic IDNs.  We have dot me now which launched in Japan earlier this year, and we've just seen its full ASCII registrations, we've seen lots of interest in if in Japan and we've seen lots of stuff going on with that IDN.  And so I think ‑‑ so there's the technical side of things where we need to have ‑‑ we need 206 them work universally.  But I think the other side of things that we're going to see is we're going ‑‑ there's never been a lot of marketing around IDNs, so all the discussions that people had about, well, they'd rather just work in ASCII.  Well, do we know that or does it ‑‑ is it familiar?  For good or bad reasons, it may just be familiar, and there's never been a lot of discussion introducing local communities into what ‑‑ what an IDN is, how they can use it, how it might be beneficial for their community. 

And I think with a lot of these IDNs now coming on‑line and even with generic ASCII TLDs coming on line, one of the things is you'll see a lot more marketing go into talking about IDNs, and the second thing you're going to see is user behavior is going to change regardless. 

So I think we've all been brought up in this dot‑com world and we're all ‑‑ and it was great for a really long time.  There's nothing wrong with dot‑com, but one of the things you're going to see is there's all these different gTLDs out there, where they're in ASCII or they're in IDN and cc's and ccIDNs, and there's just going to be this mix of things going on, and so it's going to cause users ‑‑ it's going to create a bit of a chaotic situation which will cause users to change their behavior, and they're going to start thinking about maybe they don't want a dot‑com, maybe they want something else, maybe they want an IDN and it's going to make more sense, I think, to users, because the whole market is going to change.  And some will be successful and some won't, but I think it will just cause people to think about it differently.

And so I think that's ‑‑ I mean, I'm really excited about IDNs.  I think they're going to be great, and to some of the other questions that people had about not being able to ‑‑ having to switch keyboards and things like that, there's also really great things going on‑line around language, around virtual keyboards. 

You know, Google we have one that you can download ‑‑ it's a chrome extension, and it really makes using IDNs, switching from ASCII to ‑‑ to IDNs almost seamless.  And so I think as people realise that there are more tools out there to help use IDNs and to get on‑line that way, I think it's going to ‑‑ I think they're going to be really cool and I think people are going to do really interesting things with them.  So I don't know, I'm really excited.

>> GIOVANNI SEPPIA: Thank you, Sarah.  Apart from being cool, do you have any plan you may like to share with us to ‑‑ at your end to help us, the registry managers, to market IDNs?

>> SARAH FALVEY: Well, I mean ‑‑ I don't know if we ‑‑ I'm not on that side of things, obviously.  I do know that we're obviously marketing our IDN and things like that and I know that we've partnered with other groups.  So, for example, we partnered with multiple other registries to develop a sort of test space so that other ‑‑ so that users could go in and test to see if their stuff was going to resolve or not and kind of to help figure out what some of the problems were.  So we've been working with lots of registries to figure out sort of the technical side of adoption things.  And I think the partnering around the marketing is going to come.  We just ‑‑ there's barely any out there right now.  Ours just delegated earlier this year.  So....

>> GIOVANNI SEPPIA: Thank you, Sarah.  And I understand ‑‑ I have a question from the remote participation.  But before taking that, if it's possible to give the floor for some minutes to Mark McFadden, who has been the author of the section on the report about universal acceptance, which follows quite well the intervention of Sarah.  Thank you, Mark.

>> MARK McFADDEN: Thanks.  So I'm the balancing act with Sarah.  I'm the one who's depressed.  (laughter) And I'm the one that's getting the extra work from Pat, which is good.  Our feeling, having done this for a few years in terms of universal acceptance, is one of the things that happened with IDNs quite clearly was there was a focus of using IDNs on the Web first, because that was ‑‑ that was the universal interface to a lot of applications, a lot of on‑line and social services.  And what we see in progress through the years is that the Web is doing ‑‑ is really doing fairly well.  For instance, browsers as a technology do a fairly good job, not universally consistent but a fairly good job of supporting IDNs, but what we've recognized over the years is that there's far more to it than simply making sure that browsers successfully support IDNs. 

And, in fact, another way to think about universal acceptance is to say, what we'd like to have, what we'd like to finally end up with is an environment where we could use IDNs anywhere that we had previously used a legacy domain name.  That's really where we would like to end up.  And in order to get there one of the things that we have to understand is that there is a vast environment in which domain names are used.  It's not simply the Web, and it's not simply email.  There's a huge number of environments, technical environments, in which IDNs should be able to be used, which they currently can't be.

In this year's report there's a beginning look at some of those things.  For instance, things like digital certificates, policy information that's supported inside of the DNS, things that we use to combat spam.  We use the DNS in many, many more ways than simply translating a string into an Internet address.  Those days ‑‑ those days where we used the DNS for that are long gone.  We I don't the DNS for much, much more than that.

And so one of the things that when Sarah talks about her being excited about it, I would share some of that excitement in the sense that I think that those ‑‑ the technical solutions that are barriers for universal acceptance are first of all becoming better known, partly because of this work but also partly because of other people's work.  And then second of all, those barriers are indeed getting worked on.  I see real progress there.

Now, there are things that are real problems, very difficult problems.  For instance, Pat talked about using IDNs and using international email addresses as user names.  We almost invariably when we sign up for on‑line services use our email address as a way to identify ourselves.  That's fairly primitive, and Pat will tell you that, that that's a fairly primitive way to identify yourself but it's so common that we're used to it.  And, in fact, there's an underlying security problem that goes with it.  Imagine that we could get to a day where social networking allows us to actually use scripts that are native to our language, right?  That would be ‑‑ that would be an area of real promise, and we know that, for instance, social networking organisations, people who provide support for those, are working on that problem, but I can tell you, and Pat took these words right out of my mouth, it's a very difficult problem.  You would say to yourself, oh, well, gee, just change the database to support the unite code.  It's not simple, the actual identifier is embedded in social networking service far beyond the identification components. 

So it's very difficult for these organisations to make progress.  When we hear that Google has actually been able to make mail support nationalized email addresses one of the back stories there should be how much work it took, and it actually ‑‑ it's a significant effort.  They should be thanked and given a pat on the back for the work they did.  It's a significant effort.  There's an organisation that thought that effort was worth making.  What we're seeing now and I think again sort of taking the same medicine that Sarah is on here, being optimistic about things, I would say what we see in this year's report is that the user acceptance problems, the universal acceptance problems are very, very difficult. 

In this particular year we haven't made a lot of progress, but when our interviews went with major software developers, major operators, what we hear over and over again is that they recognize the problem, and they're working on it the problem. 

And so if I were to finish with a positive thought, that's what my positive thought would be.  The technical problems underlying universal acceptance are very difficult.  They are one of the barriers to IDNs being accepted, but people know it, are aware of it and are starting to work on it.

>> GIOVANNI SEPPIA: Thanks a lot, Mark.  I'd like ‑‑ yes?  One of the ‑‑ I'd like first to get the remote participation question.

>> In Bangladesh we find that IDNs are not as relevant to local content.  The user behavior is well entrenched in Latin script, (?) content.  So I question Ms. Taylor's contention that IDNs are still more important than search or apps or other innovations.

>> EMILY TAYLOR: Thank you for that question.  I think that it sort of goes back to the remarks made by Banerjee of UNESCO just earlier, is that users have adapted their behavior to deal with lack of usability in some context.  I think it also ‑‑ also I would emphasize that the ‑‑ you know, the ubiquity of domain names and the fact that it's ‑‑ it is an important fact in local content, but as Mark McFadden has just emphasized, that domain names are used in a variety of different contexts and need to be internationalized across that.  I'm not saying, of course, that local language content doesn't exist in other contexts. 

But if we looked at the evidence of the global Internet and the real strength and overrepresentation, I would say, of English language as the primary language of Web content, and then you look at the patterns that IDNs reveal, you see something different, and that should lead us to think, well, why is that, and why is that differently?  Is it because actually people want to I don't English language as the primary language of Web content throughout the world?  And in many cases I've heard people say, well, actually I want to reach a global audience and it makes sense for me to use English in some contexts. 

But you also see, for example, I talked about the content creation in Arab states, which is happening a lot in social media, that when people kind of relax into their mother tongue, they really create content in local languages.  So I wouldn't see it as a binary decision and I wouldn't say that ASCII domain names have no role in encouraging local language content.  There are multiple factors, but what we have seen is that IDNs and local language content seem to really go hand in hand. 

So again, to borrow from what Banerjee said, they can be used as building blocks and you can't build a house with one building block, but particularly if you're at the lower levels you really need each and every one in order to create a strong foundation.

>> GIOVANNI SEPPIA: Thank you, Emily.  Next one?  Please, yes, you.

>> I'm dot‑eu.  I'll be short.  There's some good news and some bad news.  The good news is we talked many times before and many years that the drivers (?) acceptance will be the large Internet corporations with implementing the Internet sort.  So we have good news that gmail is finally doing something ‑‑ well, it's not yet, okay?  It's working part‑time, let me put it like this.  But it's really good news because there will be more followers, definitely, after such up ‑‑ the platform supports the IDN.  And the bad news if you look at my badge, there are two question marks, and it means dot RF.  So we ‑‑ the bad news that we have a lot of legacy software, including the United Nations one, which is really weird for me.  It doesn't support IDNs, on Unicode, let me put it like this.  Thank you.

>> GIOVANNI SEPPIA: Okay.  We'll make sure that the Secretariat of the IGF addresses that as soon as possible.  Irina, do you want to complement what he just said?

>> IRINA DANELIA: Actually, I want to follow the previous question and Emily's comments, and it was mentioned that users have adopted their behavior to non‑IDN email addresses, and I want that now businesses are trying to change these habits and to ‑‑ because the IDN addresses ‑‑ they probably have ‑‑ probably their biggest value currently, in current situation, in current world, is the value for businesses, getting them additional ability to promote themselves if they go on‑line.  That example with their green track, and there are many, many other examples.  If I'm doing a business, if I want my customers to notice me, if I'm going on‑line, I want a good domain name, and in all ASCII world good domain names are occupied.  They are used.  But new IDN world, really gives me this ability to be noticed, to get ‑‑ to be remarkable.  That observation is that then IDNs, if we are talking about ‑‑ IDNs, they really need government support. 

If the government supports this initiative, it can help increase their awareness many, many times.  No private company will get ‑‑ has that money for advertising, for building awareness as government can give you.  And as a good example I just noticed in my country we are having the day of our city, of Moscow, capital, it's going to be a big celebration, and there are different announcements in the radio that ‑‑ like we have ‑‑ we'll have this event and this event and this event, and for details please go to the Web site, which is I love Moscow.RF.  That's probably the biggest advertising and biggest promotion that IDNs can ever get.

>> GIOVANNI SEPPIA: Thank you, I Irina.  Manal, please.

>> MANAL ISMAIL: Thank you, and thank you all for the presentations.  Actually it makes one optimistic to hear that the universal acceptance issue is recognized and is being worked on, because just to iterate the importance of universal acceptance to users, normally users are users, so if you're using the domain names in ASCII and you don't have a problem, you won't really care to switch to something that's not working equally smooth.  And if you have a language barrier, then it becomes more problematic for you, and we are trying to resolve a language barrier issue and introduce more of a technical barrier thing, because I fully understand that users adapt, and we can tell them new habits, but you cannot really change everything.  I mean, we keep telling them, you won't use space, you have to use a dash.  So ‑‑ and then you will have XN dash dash so please don't panic and continue.  So it becomes ‑‑ when it becomes too many things to learn, it might be somehow creating more barriers. 

So I believe it's a big issue, and it's most important for all users in general, to really convince those who don't have language barrier to come back to their own languages.  It has to really be working very smooth and similarly ‑‑ or more importantly to those who have the language barrier.  Thank you.

>> GIOVANNI SEPPIA: Thanks a lot, Manal.  Ongid, please.

>> First I want to give my thanks to Emily because you have written a very good report, which has covered a lot of information about China.  And we do want to let everybody know we have put great efforts on the IDN development and more so the acceptance issues, and also ‑‑ I also have read the reports given by Pat, and I think it has given us a lot of insights, although I am a little bit disappointed that there is no .CN on your reports, there's only .TW.  Yeah.  (laughter) And ‑‑ because we are ‑‑ we actually covered (?) to develop the IDN issues, and this year is actually a very promising year for us with regard to the IDN development, because we have many software developers to support the IDN emails.  We are trying so hard to push forward IDN emails to the global communities, especially with regard to (?) and Google.  And this is our great achievements. 

But when we are making this progress, we also are looking to the drawbacks to our efforts, and we actually found that in Chinese market we are too overemphasizing the ‑‑ the big companies, yeah, because we are trying to do more marketing to the big companies, and we have little coverage to the startups.  This is what we haven't ‑‑ what we have to ‑‑ we are not so inspired about that.  Because right now we see there is lot of market potential for the startups. 

And in China, actually the Internet market is a little bit strange.  The big gets even bigger and the startups are not ‑‑ fade away quite quickly.  So ‑‑ because the big companies, they have ‑‑ they already have very nice ASCII domain names, they don't have the incentives to push forward for the non‑ASCII ones.  And the users already know, the (?) ASCII domain names.  They don't ‑‑ they don't have the incentive to access the non‑ASCII domain names.  So the Chinese users are already ‑‑ already get adapted to the ASCII domain names, they seem to not like to I don't the non‑ASCII ones. 

And right now we are trying to move forward to focusing on the startups, because when you see the startups, they actually can bring us more innovative products.  If they have such kind of products and they have a direct link with the non‑ASCII domain names, I think that would be very promising message for us pushing towards the non‑ASCII domain names, IDN domains. 

And right now we are going to have a marketing company which ask about 50 companies, that are startups, to give a showcase of how they use non‑ASCII domain names, and we want people to know that the IDN domain names can be used in this way, in an innovative way, and they are not even comparable with the ASCII ones.  Thank you.

>> GIOVANNI SEPPIA: Thanks a lot, and I'd like to thank you again for the really ‑‑ it was a great pleasure to visit Sienica last year, and for me and Emily it was an incredible experience to learn how much you have done, not only at IDN level but also to engage with the local community to promote the domain names in your country.  So thanks again for that.

I'd like to give the floor, if she likes, maybe, to Minjee Kim from Korea.  So you may have time to react to Pat Kane's statement about how much English is presented in Korea, (laughter) and also give us some information about your experience with IDN in Korea.  Thanks a lot.  Minjee.

>> MINJEE KIM: (?) English.  And I want to say just a brief case in Korea.  As you know Korea IDN service is really good, so we just surfaced a little bit about a lot about the universal acceptance so broader system and mobile phone at the IDN service is really good, and improved in the last year, but there is some complaints about EAI, international ‑‑ email address, internationalized.  So in case I have the plan to like fund ‑‑ to the registrar.  For example, if the registrar to set up the email server, a little matching fund.  So if the registrar set up the email server, the universal acceptance will be better because the customers have it better.  So that's a little bit of short plan for the (?).  That's all.  Thank you.

>> GIOVANNI SEPPIA: Thank you so much.  And I think I remember that last year in the ‑‑ the local ‑‑ the regional IGF in Seoul announced that the very first full email experienced was tested.

>> MINJEE KIM: Yes.

>> GIOVANNI SEPPIA: Thank you.  I have a question from the remote participant, and then you, please.

>> AUDIENCE: User behavior on the Internet assumes latency ‑‑ (?) a good set are required to make the case for IDNs.  Just universal acceptance is not enough.  That only pushes the technology forward when user behavior is what must be focused on.

>> EMILY TAYLOR: I think that's absolutely right, and some of the exciting developments that have happened in the last year, which are highlighted in the report is that, you know, there is some evidence of software tools for application developers out there in the mobile market being available.  The thing is that probably not very many people know about them and, you know, that they are ‑‑ they're not mandated ‑‑ or sorry, they're not there by default, that you have to go out and look for them.  But absolutely.  But it is, again, stepping stones.  You can't really start to have multilingualism in addressing if you can't actually support those characters.  Then you can't really have them in cool applications unless they actually work in the standard ways that we would expect to use the email.

I'm really interested in the ‑‑ Mark's ‑‑ the remarks made by Minjee about the focus on startups and small or medium size company.  Because the registry code it's the right size in my opinion because it's there you will get to the people and businesses that are less familiar with the Latin script, that are much better served by local language, and wanting to make, as Irina said, when you're starting your business you want a good domain name and you want something that's memorable and meaningful and that people can type in and get to where they want to be.  And the wonderful example from KISA, the Korean registry, just an incredibly practical intervention which is to match funding to encourage people to get to their email servers compliant.  What a simple and incredibly effective and practical step to take.  So, you know, to get to where we want to be, it is multiple actors taking small steps and working together to get that multilingualism happening. 

>> GIOVANNI SEPPIA: Thank you, Emily.  I'll take the last question from the floor.  Please.

>> AUDIENCE: Thank you very much.  Satish from India again.  Thanks for the encouraging news on the work being done on the technology front.  I'd like to place one specific point for the consideration of the technology people, and this has to do with the disabled and the access to IDNs.  The IDNs are likely to instantly break.  Many of these come to us use regions that the disability are used currently.  Maybe some work on this would be very useful.  Thank you.

>> EMILY TAYLOR: Thank you very much for that intervention.  Fully agree.

>> GIOVANNI SEPPIA: Okay.  Really quick from Fahd.  Thank you.

>> FAHD BATAYNEH: Thank you, Giovanni.  This will be quick.  One of the things we're doing in our region, the Middle East, to promote IDN uptake and probably even developing solutions for the use of IDNs, mainly the issue of universal acceptance is that we are targeting ‑‑ let's say university students where the undergrads or graduate students, as well as entrepreneurs and encouraging them to actually develop software and probably devise solutions for universal acceptance issues related to IDNs.

>> GIOVANNI SEPPIA: Thank you, Fahd, and Mohamed?  Just to balance the regions, because otherwise we'd like also to hear a bit more from the African region.

>> MOHAMED EL-BASHIR: Okay.  I hope I don't really be between you and your lunch and breaks.  I think first I'm happy to be here in this room with four or three of all of the IDN guidelines working here through Pat and Ram.  I'm excited about the issue of global acceptance.  I mean, with all the complexity there it will be resolved.  There's a drive for it.  Commercial drive is there.  There is also consumer demands, and just to give you one example, there is some work has been done for our ‑‑ out of localizing famous software ‑‑ social media applications, like Twitter.  And the moment that Twitter with the support crowd ‑‑ sourcing of individuals in the region introduced RB cash back, that exploded the use of Arabic language in Twitter.  I think now so do ‑‑ I think it's the most active users within Twitter currently.  So I suspect that with Google now step, others will be coming, and then we'll see huge use in terms of the IDNs.

It's normal.  This ‑‑ the registries have started operation in the last two or three years.  I'm running currently independent an ‑‑ basically we are in operation the last two years.  This is a normal progress of R&D and software development.  The DNS industry took, I think, 20 years for some of those applications surrounding the DNS to really be up to the level that we are, so I don't think that's ‑‑ that's a normal progress of resulting things.

I think there was huge expectations initially when we were applying for IDN, and were preparing to launch them, but the reality is there are multiple issues that we need to ‑‑ we need to address.

IDNs ‑‑ IDN is not usually the normal thing, by the way.  Yes, maybe for the legacy situations, that has become the normal practice currently.  But the normal is IDN.IDN.  For example, if you're using Arabic switching keyboards it's not a good thing.  It also is a bad user experience from the other side.  So you need to look at it from the different perspectives as well.  So it's not normal for ‑‑ for an Arabic domain name to be an ASCII ‑‑ an IDN to ASCII.  So that will be ‑‑ the evolution will happen, I'm sure soon we'll have IDN operational.

We are also seeing at the edge are industry, running in Qatar, we're seeing also ‑‑ globally ‑‑ 3 to 4% of the current registrations are IDNs.  And the majority ‑‑ the majority are redirections to current existing Web sites, or redirection to Arabic pages, because also you need to understand that there is loss of entities who does multilingual Web sites, so basically official organisation could have English in Arabic Web site as well.  So the same patterns as Pat and the report identifies already in the IDN ‑‑ still experiencing that.  But I think that eventually it will happen.  It's just a matter of time and a major ‑‑ major players stepping in and resolve issues.  Thank you.

>> GIOVANNI SEPPIA: Thank you so much, Mohamed.  And from your words, I'd like to take some ‑‑ make some concluding remarks.  As you said, IDN.IDN should be the normal way for end users to approach the Internet, and that's really important.  It's really important to educate the end users, to educate also all the different players in this complex market.  So not only registries but also registrars or international organisations, governments, everybody might contribute to make sure that the full IDN experience is enjoyed by the end users.

I heard some sort of mixed feelings, positive, negative, IDN will make it.  There's going to be huge update.  Maybe not.  Some issues still at acceptance level, but I think the majority of us in this room agree that IDN will make it, and as Mohamed just said, it's just a matter of time.  It's just a matter of making sure that people, they get more familiar to a different way of enjoying the Internet.  And as also Mohamed said, the more we do, the more we believe in this, the more players will step in and make sure we're on the same page.

So that said, I'd like to thank all attendees, all the panelists, UNESCO, VeriSign, the original ccTLD organisation for contributing to the report.  Thank you so much.  I'm sure we'll see you next year for a further update and thanks a lot to Emily who coordinated this monumental work of putting together these pieces of this report.  So thanks, everybody.

(Applause)

            

         *******

                    This text is being provided in a rough draft format.  Communication access realtime translation (CART) is provided in order to facilitate communication accessibility and may not be a totally verbatim record of the proceedings.