FINISHED FILE
EIGHTH INTERNET GOVERNANCE FORUM
BALI, INDONESIA
BUILDING BRIDGES - ENHANCING MULTI-STAKEHOLDER COOPERATION FOR GROWTH AND SUSTAINABLE DEVELOPMENT
OCTOBER 24, 2012
11:00 AM
WORKSHOP 88
BUILDING BRIDGES TO ONLINE MULTILINGUALISM
* * *
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.
* * *
>> GIOVANNI SEPPIA: Good morning, everybody. We are almost ready to start. We have a remote speaker whom we are trying to reach. It's coffee break time so I guess he might -- sorry? Not anymore but I guess people are mingling around. I'm very happy to present to you for the second time as moderator this great panel that will speak about Building Bridges to Online Multilingualism.
While I'm closing this door, just telling you why it is so important for you to -- mission impossible -- to be involved in multilingualism. This is the register operator and we have provided and we provide support in the 24 European Union languages. We have been granted a new language on the 1st of July, Croatian since they joined the EU, so we are now providing support in 24 languages and also applied for the dotu string in Cyrillic and Greek.
We have started a partnership with UNESCO three years ago and formalized with a Memorandum of Understanding in the first quarter of this year. This partnership aims to start investigating more multilingualism and what can be done on different levels to promote multilingualism online to make sure people are given enough capacity to access Internet in their own language and express themselves in their own language on Internet.
That's why we have started this and this is the third year we published the World Report and as really does become environment friendly we have not printed but put it in this memory sticks that you can find here close to the projector on your right and in the memory stick you can find the work we have done this year for the World Report 2013 and it's monumental because we were lucky enough to have other partners with Verisign. Verisign provided us with enormous amount of data which we have analyzed to analyze and better assess the penetration and if IDNs are really used.
What are IDNs? It's a shorthand, for those of you who don't know, and it stands for Internationalized Domain Names, Internet started with the Latin script but then at a certain moment there was a need to allow people and communities not speaking in not having their language in Latin to be more connected and to take advantage of the Internet and the Technical Community started work on protocols to allow people to get online by using domain names that have the second level and more recently at the top level scripts using characters that belong to scripts that are non-Latin scripts.
So we have recently heard few days ago, two days ago at the beginning of the IGF meeting that I can has proposed three new GTLDs for Delegation and those that have been proposed are in non-latin script and more specifically is Arabic word for web, Russian word for online and the Chinese word for game.
And those have been proposed by ICANN for the Delegation. Those will be the first non-Latin characters that are used in generic top-level domains.
As I said this report takes advantage of this partnership we have established with Verisign to get more data about the use of IDNs especially in the dot com and dot net environment and that has provided us a great picture of what and how IDNs are used.
We have a great panel today made up of people who are differently engaged in the industry and can contribute to some of the questions that this panel should respond, questions like what can we do all to promote IDNs? What kind of measures should be taken to facilitate the access to the Internet for those communities who are not speaking with languages that are linked to the Latin script?
In the next 12 months there will be many changes in the Internet because there will be a lot of new detail coming up in the market. Many will be non-Latin GTLDs and that will change the GTLD but also entire Domain Name System landscape. We'll see major changes in next five to ten years.
Going back to the panel I'd like to start to introduce the first speaker and like, first of all, to thank all the speakers. We have also new speaker who is not in the online programme and speaker who can provide the view, registrar's perspective but when talking about registration of domain names most of the registries were let's say backbone, behind-the-scenes architect of the management of the domain names that don't sell directly domain names but they go through network of accredited registers. We are lucky enough to have one registrar today with us, not so many registrars, probably just you can count them on one hand registrars participating this year in IGF. We are lucky enough to have one with us today who can give you the perspective of registrar, what can be done at the registrar level.
That said, I'd like to introduce the first speaker. Emily Taylor, she's author, main author of the report as previous two years' report. She has started this cooperation with -- to develop the daily report on IDNs and this year as I said she has gone through quite monumental work because of the amount of data that we have received from Verisign. She is also has been involved in different Internet, say, areas including with review and most recently the second accountability and transparency review team of the ICANN policy development process.
I'd like to leave the floor to Emily. She will speak and introduce the report with a lot of data and for those of you who are interested also in seeing the report, online report, the report is going to be published today on the URI.U site. If I can just have one screen with the slides, please.
Thank you.
>> EMILY TAYLOR: Thank you very much, Giovanni.
We just are getting the slides up onto the screen. You're gonna drive, right?
(Laughter)
>> GIOVANNI SEPPIA: I'll do my best.
>> EMILY TAYLOR: It's a great pleasure to be here introducing this year's World Report, and as Giovanni said, we have had the collaboration and cooperation of many people and Giovanni mentioned Verisign but I'd like to also knowledge CNNIC and Russian ccTLD whose representative is in the room including the center community as well as many others.
Next slide. The data sample we have been looking at representing 90% of the world's registered domain names and the special focus this year has been on usability and also how domain -- how IDNs are being used in practice. We have looked in particular at the Asia and Pacific region but also country case studies which we've been knowledge up over the years which cover Arab states and also Russian Federation. Then there's facts and figures as you would expect in this sort of review as well as industry opinions.
The next slide, please.
Just I don't know if that's at all visible to you on the screen because we've got sort of rather nice light green on white but the really this slide just highlights some of the major findings of the report which is that out of over 250 million domain names registered in the world we found 5.1 million IDNs, Internationalized Domain Names, so that's just 2% of the world's registered domain names. When you think about the comparison between offline and online life and number of people who are unfamiliar with Latin script in the world, that's a rather startlingly low percentage in my opinion. In .eu, a URI report, EU offers internationalized domain names in Latin and also Greek and Cyrillic and had at the end of 2012 61,000 internationalized domain names.
So we have been looking at in the usability study and five most popular browsers, ones we use on desktops, are now in their most recent versions supporting IDNs pretty well but the same is not true in mobile devices or in Web-based applications either.
So we found that 92% of the world's most popular websites really don't recognize IDNs, for example, in the creation of user accounts, if you are asked to give an e-mail address all -- pretty much all of those don't recognize IDNs and also don't recognize IDNs as links. So in this modern world where we are what we share on social networks if you type in a URL which is in non-Latin script, for the most part that won't actually just function.
That said, that's the kind of dreary bad news. But on the hopeful side, there is a very, very strong, near-perfect correlation between the script of the IDN what language that is leading you to expect, and if you follow it through, there it is in the language of the website. That's I think very important and we can start to say that IDNs have a very important part to play in the multilingual ecosystem. For completeness, the first IDN e-mail was sent successfully in 2012.
So next slide. We have been looking very much at usability and usage this year. So the -- we had to define usage in some way. We defined it as relative level of ease of use predictability and memorability of IDNs in Internet services applications so we were looking at browsers, and the main finding is it's getting better in traditional browsers but still very poor in mobile devices. Again, when you think that not just in developing countries or mobile devices that the device of choice increasingly for accessing the Internet but also in developed countries, too, this is an important gap to fill. Web-based services we actually checked across Cyrillic, Arabic and Hmong across 13 of the world's most popular websites and looked at e-mail across traditional e-mail providers and mobile applications as well.
So what did we find? There's actually -- when you look at Web-based services, there is a difference in the way that they treat full IDNs so that is something that Cyrillic script both in the main bit and the ending, and in the hybrid IDNs, by which I mean, say a Chinese script.com so you have two scripts in play in the same domain name.
Generally speaking, the treatment of hybrid domains is a little bit more developed but you still see, I don't know if you can see the little screenshot there. We have got Twitter if you put in hybrid domain name as a link you can do that, but what it will show to your followers is a bunch of meaningless letters and numbers, the underlying code. In terms of usability and memorability, not fantastic. And as user identifiers so to set up a Facebook account, to set up an account in Twitter or any of the LinkedIn, forget it, you can't use an e-mail address at this time. That's a very important gap to fill.
So as I mentioned, the first IDN e-mail was sent in 2012. When you consider the IDNs have actually been on the market and for sale for 12 years now, it's perhaps a surprise that the first IDN e-mail took 12 years to be sent and it just does highlight the gap in usability we have been finding.
Speaking to colleagues in registries that have deployed IDNs they tell us providers are now able to send IDN e-mails very successfully but because an e-mail will have to travel through so many intermediate steps in order to arrive at its destination. If a single one of those can't handle the IDN it will bar it and fail and this provides a disincentive for providers to offer this service because it's a very difficult support mode to analyze, to identify, because as far as the user is concerned it will have failed and it's probably the provider's fault. Until that chain is working end-to-end, it will deter people from providing it. But on the bright side, within China, 600 million users are offering full internationalization so it's coming but if you look all the big name providers, Yahoo!, Microsoft Outlook, Gmail, Google and iPhone, none of these are working with IDN e-mails.
We found that.
So next slide please. Let's look at the sort of usage because you register a domain name but you don't have to use it. It will still be there. Generally speaking, usage of IDN e-mails lags behind the more traditional counterparts so we found there are two hybrids in this slide looking at 30 ish%, 30 to 40% of use in websites plus another 12 to 15% in redirects versus Russian fully IDN domain where they made huge strides in the last year in improving the instance of usage. So trends are going in the right direction but still a long way to go.
Next slide, please.
We talked about correlation between language and website and the associated IDN and it really is quite amazing to see how far these match. When you look at a Greek script -- .eu, the only languages you see are Greek, English and there's a hodgepodge of other languages but really very few. Cyrillic script you're looking at English, Bulgarian in the EU, Russian, Portugese for some reason, I don't understand, that comes through in both .com and .eu and I have no idea why and then Latin script as you would expect with a script that's used across multiple languages within Europe, you have got an array of languages. German features very strongly and so do many others.
So next slide, please.
This is a little bit -- this is one to study afterwards but take it from me when you look at the dot com and net IDNs were you have a data sample about 100,000 so very nice sample here. There is an almost perfect correlation between the language of the website and the script of the domain so taking Chinese, the first line, we have got 40, nearly 41,000 in that script, in the hand script and about 41,000 websites in that language. So there's over 99% correlation. This is an extremely hopeful and exciting correlation that shows IDNs' role as signaling content in a particular language. Again we saw quite a high level of redirects but when you look at the script of the redirect -- and Korean is my favorite, 95% of Korean script IDNs that go somewhere else, they go to web content in the Korean language. Similar correlations are found across others.
Another study by UNESCO, ISOC and OECD found a strong correlation between local content and location of servers and so we looked sort of taking this we looked at the country of hosting so the pie chart on the left shows the sort of country of hosting for the general EU registrars. It is showing a distribution. On the right you show the distribution of hosting of internationalized domain names and it's different and in particular you see more are hosted in Germany where there is a lot of, you know, diacritics in the Latin script but you also see Bulgaria and Greece come in as part of the top 10 countries of hosting of IDNs whereas they are not featured at all in the top countries of hosting of other IDNs -- sorry -- of other .eu domain names.
Again, we have some other interesting -- in my view interesting information about the content is kept locally, it's hosted locally, it starts to build up a virtual circle of local language. I'm not going into detail on Asia-Pacific because we have speakers from the region both remotely and here who will talk about their own experiences but just to highlight quickly we have developed a matrix which has on the one side country and language factors and on the other side the way that the country code, ccTLD itself, all those contribute to making conditions good or otherwise for the deployment of IDNs, but just to say in our study even the countries in that top golden right-hand quadrant, China, Republic of Korea and Russian Federation really still in my opinion although doing remarkable work are still very far from achieving the potential of those markets, those users.
On facts and figures, generally there is growth year-on-year but in fact that masks that some of the larger IDN registries have in fact shrunk over the 12-month period. The tiny little line at the bottom you can see is the Arab states and I think we just have -- I won't go into that but it's really quite remarkable how far, you know, the scale of deployment in the Arab states is below that of the rest of the world in volumes; however, in terms of percentage growth it's actually pretty healthy.
The growth rates in a nutshell, they are much more peaky, much more volatile in IDNs than they are. Growth rates in the whole register are now fairly level around 5 to 10% whereas they are all over the place with IDNs.
Next slide. We asked registries and this year for the first time registrars -- and we have Peter here as a welcomed super subject expert on the panel -- we asked them a series of questions but just two I want to highlight here. How do you think IDN uptake is doing versus your expectations? In the main, it is fairly in the middle. I think it's sort of, average is for both registries and registrars, still in the 2.something, you know so they're not delighted but their expectations are probably fairly low as well in some cases.
Next slide we ask: How do you think end user awareness is? Again, I think that probably again they're fairly similar, registries and registrars think about the same and it is fairly stable over the years. But generally speaking, they don't think that users know about IDNs and I agree with them, to be honest. Next slide. So what do we learn from all of this whistle-stop tour? IDNs clearly have a very important role in supporting the growth of multilingual content which is actually what this is all about. This isn't just counting domain names. It's about how people navigate the Internet. When you think of we're familiar with Latin script, when you visit a country where you are going through an airport and you don't understand the script of any of the signs, how do you know where you are? How do you know where you've been?
Deployment of IDNs is increasing and there is progress on several fronts but there is much more progress to make. The Director General of UNESCO made a statement this year encouraging the Technical Community to work even harder to make deployment a reality. I think in the next year we'll see deployment of perhaps 100 new GTLDs in different scripts and I believe that will provide an impetus for investment by areas of the industry that are definitely capable of supporting this but have had other priorities.
I would also just like to highlight a personal observation, that the advocacy of the registries, whether CNNIC, ccTLD, Korean Registry, they have really been tireless in advocating the adoption of IDNs and lobbying major providers to support them better. So I think without them it would be a lot grimmer the picture but I think at the present we have a fairly negative cycle because there is low uptake and so because there is not a lot of drive from users saying we want this better, it's fairly sustained over the years. That again just deters people from using them which again does not really lead to high user awareness. So we are trapped in that at the moment and something needs to change.
We kind of adopted the hierarchy of needs as a model to try to understand where we are and what needs to happen and I think we're really still in I think what Masloff would call hygiene factors. They basically don't work very well at the moment. Until they work you're not going to get the new market offerings, you're not going to get the early adopters and mass adoption so there needs to be important foundational work that will drive wider uptake. Potential is definitely there.
So there are encouraging signs but more needs to be done. Thank you very much.
>> GIOVANNI SEPPIA: Thank you, Emily. Thanks a lot.
We have really spent a lot of time with that tool, analyzing and assessing all the different surveys and IDN resources that we were looking into to have a better understanding of how much IDNs are used and perceived.
Next panelist is a great supporter of this study which was started three years ago. Janis Karklins, Assistant Director General at UNESCO, diplomatic career and only he is Latin ambassador to the United Nations. He was a Latin ambassador but also is well known in the Internet community because he has served as the Chairman of the GOC for three years. So I would like to leave the floor to Janis and again thanks a lot for the support and this great partnership that was established with URID.
Thank you, Janis.
>> JANIS KARKLINS: Thank you, Giovanni.
I must tell you that the pleasure is mutual, if I may say. Certainly I have personal interest in supporting this, and that stems from the recent history when I remember myself and a few other ICANN and CC community folks were cornered by representatives of Arab ccTLDs in the corner of the room and where they asked us if you don't -- if you will not do something with an IDNs, we will talk about alternative route. That was good encouragement for ICANN to launch the process of what is now known IDNCC -- fast track and this is also reason why or that gives us information in this analytical study on the IDN uptake.
We see certain trends from the previous year and more information we will gather, more kind of better analysis we will have. For UNESCO, this studies very important because that goes in direction of implementation of the recommendations of use of multilingualism in cyberspace which were adopted in 2003 and we're reporting on a regular basis to Member States on the achievements and clearly the study provides us with a lot of very valuable information.
As you mentioned, or as Emily mentioned, the general of UNESCO based on findings of 2012 study made public statement where she outlined importance of technical, work of Technical Community in moving this development of IDNs forward and encouraged to continue in that path.
Myself, I wrote -- based on that public statement I wrote letters to Chairman of ITF and IAB, asking them to draw attention of respective constituencies and now when I spoke with Yari, he said that we may expect some attention from their side and particularly attention to those engineers who are working on development of applications and browsers and what is identified in the report as remaining technical challenge.
Again, Director General also congratulated Technical Community with successful e-mailing or application of IDN e-mail protocol and again the question is now just time and how fast and widespread this protocol will be used by service providers. Whether IDNs -- or how to break this vicious circle as Emily outlined in her presentation and improve user experience.
When we look to current IDNs in the route, they are very local, if one can say, big operators running local service. Nevertheless, big internationals are mostly in Anglophone territory. Now when companies like Verisign will be say commercially interested in good uptake of their IDN GTLDs, I think that will increase a lot pressure to developers and I suspect that or I predict maybe in next couple years we will really see dramatic improvement of user experience.
That is inevitable when big money -- not substitute but complement the national and identical -- identity drives behind use of IDNs. I think that will be a stimulus for further technological development.
From our side at UNESCO we are talking to our Delegations through or by using this report, telling them that there are also things on the organisational side that needs to be done, and we will continue doing this work from our perspective.
I will maybe stop here but before that, from my side I also would like to thank Yuri for supporting this activity and studying and I hope that in the future we will go from one even more interesting to another even more interesting study and certainly that will be useful material for all those who are interested in development of multilingualism in the Internet. Thank you.
>> GIOVANNI SEPPIA: Thanks a lot, Janis, really, a great pleasure for us as well to work with you these three years on the report and look forward to continuing the work.
Next speaker I'd like to introduce and we are going to have three speakers now coming from three different -- still the Internet family but three different sectors. First one is Hongbin and Hongbin Zhu is Senior International Strategies at -- since 2010, also responsible for the new GTLD application and Internet Governance at CNIC.
I must say at the end of the meeting we were so lucky to meet Hongbin in his own environment in the registry and it was a great experience because we were really sustain into a world which is very, very special for the way you handle these huge registry and the way you manage the data, you treat the data, and the way you try to promote the Internet and toe main name in China.
I want to compliment you. Really a great visit at the end of the ICANN meeting in Beijing. I'd like to leave the floor to you and then to your presentation up on the screen.
>> HONGBIN ZHU: Thank you, Giovanni.
Good morning, I'm Hongbin Zhu and I come from CNIC. I'm in charge of the IDN research and universal acceptance in CNIC and now I'm wanting to introduce a little bit about how the Chinese going in China.
Before I start to introduce about IDNs I want to talk a little bit about the Chinese language usage on the Internet in China. Actually we have a lot of Internet users in China which is now almost 600 million users actually using the Internet in China. We have a lot of local content developers which are very famous among Chinese Internet users. On the left side of my presentation is some software developers, they all have larger user basis so as you can see in this numbers the Chinese language content is more closer -- closer to our users and according to another research I found on the Internet the Chinese language is now the second biggest largest, second largest language on the Internet which is almost -- just the next to English language.
The study also said about that the Chinese language will become the first language on the Internet in maybe in the next ten years so I think it is huge market but we are thinking about the Chinese ideas, the thing is a different thing where we need to do a lot of work for that.
Next slide, please.
Actually our study of Chinese idea has started from the end of the 1990s and we have a cooperation with a lot of stakeholders to research on the IDNs including our governments and the academia and in 2000 we established an organisation called Chinese Domain Name Consortium. This organisation is a multi-stakeholder organisation which provides registries -- user groups and they all provide some recommendation about user experience about IDNs and also technology standards, what it's gonna be, what does user expect, and in 2004, the Chinese IFS got published and more so in 2005 as -- yes, it's also got published and remarkable milestone of ours is that in 2010 our first IDN TLDs get Delegates in the route which is -- simplified TLD and traditional TLD delegate gathered to us, promising because I would talk about simplified Chinese and -- later but it's best practice for us because we see each simplified Chinese and traditional Chinese almost the same.
Unfortunately, the number of legislations is declining after delegate to the route. Major problem is user experience. The end users, most of them, user doesn't know about IDNs, just registrar trying to sell them. So we need to improve the user awareness so we can come to the next maybe climb for our TLD.
Next slide, please. I will talk about Chinese -- issue. As you know in Chinese community we have true writing system, simplified and traditional Chinese. Used in the same community and they can sometimes they can exchange or interchangeable so user can mostly be get very easy to get confused about the contract use so we have developed a policy which is called Combined Delegation of Combined Registration Policy. This policy actually mandates that the variance and original characters belong to the same registrant and they resort to same website. This is our policy and this is guaranteed for the security of the users.
But now I think the -- not being recognized by ICANN or other registries so we are still pushing for this issue to ICANN and to the other communities with respect to that issue.
Next, please. A promising success is e-mail address. We started e-mail address internationalization research in 2004 and in last year we actually sent the first Chinese e-mail from Beijing to our community members in Taiwan and Singapore and other community members and this success is actually a tribute to the joint efforts of our community members which included -- Dominic and sing -- and dot Asia and other -- also the Korean registry KISA and JPIS. They all have made a great contribution to this work and we actually have step into the Intergovernmental organisation which is APAC to promote acceptance of the e-mail address. This is a very promising phenomenon.
Next slide, please.
Browser support is actually okay in China because we have most of the users in China actually use our local browsers and they may use we call it the 360 browsers, these large browser in China, and they use this kind of browser and these actually support the Chinese IDNs so we don't have any problem with the browser issue but mobile, there is some problem with the mobile issue and I want to talk more about e-mail because actually I have experience as to sending e-mails in Chinese to other people. I also have e-mail account in Chinese but problem is that they cannot receive it. Actually developer strategy which is we combine the Chinese e-mail with an ASCII e-mail and if you can receive it, if you cannot receive it, we just have an alternative e-mail to send to you. It's just automatic, yeah, so it's something we can make up for today's phenomenon because there is not very promising. And the search engines is quite supportive. Finally, I want to share some ideas. We had major standard developers for the Chinese e-mails and Chinese IDN standards and we are worked with developers with minor languages actually in use which is language of Mongolia and Uga and that is almost the same as the language of the Uzbekistan and we have also developed IDN solution for them and it's not in wide use but we are trying to push the acceptance of these standards but now we are facing problem that standard is there but no one knows about it. So it's, we still need to do a lot of things to promote the acceptance of the standards and our government is quite supportive for the IDN user because it -- integrity of work language community especially between China and Taiwan and other Committee and we have a close relationship with other community governments and they are quite supportive, especially when we were in APAC meetings.
Another point to raise is that some user experience problems which still need to be solved which I list here including phishing and also some interruption during when you -- through the Chinese IDNs.
Finally, I want to share some of my opinion about the future technology and which can actually promote acceptance of Internet. We are thinking about the voice recognition technology. We actually are researching on this kind of technology. You can actually say or speak the Chinese language domain names and you can access websites. It's good for disabled people and it's also good for people who cannot read English words. That's I think we have make some efforts on developer this kind of new technology we can promote more acceptance of the IDN issues.
Thank you.
>> GIOVANNI SEPPIA: Thanks and again thank you for the valuable contribution to this year's report. Data you provided are really valuable and we really enjoy to enter the world of IDNs, especially on the continent of -- in the Asia-Pacific region.
So I am looking at the remote moderator corner. If Minjung Park of KIZA, Korean Registry has managed to access the remote participation because she was also one of our panelists and she was supposed to have a remote presentation but I understand there was some issues. No connection. Okay. Thank you.
So I would like to leave the floor now to the very last-minute, nice addition to the panel because, nice addition is Peter Larsen, a registrar, -- registrar for many TLD extensions. He's also a member of the advisory Board and I really wanted to have Peter's perspective to understand as I said at the beginning, introduction, most registries, they to not sell domain names directly to the end users. They do sell domain names via network of basically retailers and network of what is in our jargon they are called registrars.
Peter Larsen is the owner and manager of Larsen Data, Danish registrar and he is well known in the community because he's very proactive in participating in ICANN meeting but also in European ccTLD meetings where he has been vocal to make sure that all could hear the registrar's perspective because at the end they are those closer to the end user community.
I would like to leave the floor to Peter and thank him again for his last-minute availability and I'll put up his presentation.
>> PETER LARSEN: Thank you.
Well, it was a last-minute effort so I only have four slides and I don't have that much nice statistics. Next slide, please.
We are based in Copenhagen, Denmark and running free DNS service as one of our registrar competitors and we have included IDN in our service since roughly 2003 or 2000 and this was due to the Danish registry added the IDN, for the special Danish three -- we have and we have a lot of experience with that since because we are also registrar and selling. So we have the end customers contacting us with issues.
Next slide. The top one is a typical Danish domain IDN. I put in a short link for it so it can easily be typed by any of you here but this domain is used as forward to another domain as many of the IDNs in Danish language is because there is very seldom used as a contact point due to all problems we heard from the web browsers and the e-mail issues. If we could, I don't know if the Internet is working but the second link on that page if we could get that up, if that's working. It should be connecting to the Danish Registry and they have a nice overview of the number of domains that is registered since 2000, how many of them that are IDN and how many of them that are Latin.
So you could, if you scroll down you can easily see that the numbers are hovering around 4% or 5% of the TLD which in my perspective is quite good but it should have been quite higher since the Danish letters commonly used in many names and good domain names should have been around 15 or 18%. We have a lot of customers contacting us. Usually after they get the domain and website running and telling us people cannot contact us, e-mail is not working and we have been emphasizing to people they should get both Latin and the IDN version, doing that for promotion and e-mail and mix them nicely. I can recognize a lot of statistics that I heard here today as that is how the end customer is reacting and it's a bad end user experience definitely.
We have seen since 2004 we have the IDN in the Danish TLDs so e-mail issue has been a problem since and it's just recently last year that there seems to be getting a solution to this. We have been telling our customers since 2004 it's coming, it's coming. And I'm afraid that we are going to say that a few more years but I hope the new process will speed up software and I hope that some of the registries or providers, software providers, are, well, helping to open-source projects to understand the ITNs better and definitely getting it pushed a little further.
Just recently Giovanni mentioned some of the new TTLDs was put into the route here a few hours ago actually and one of them is games in Chinese or game I'm not quite sure. If you look that up in your quick browser, only one working out of the box is Firefox of the western web browsers. The rest of them actually go to a search page for Google. There is definitely problems with applications and software people are using still.
My advice would be if you have the possibility as a company, registry or whatever, push your software provider push your open source community, help them provide them funds programming skills to help with IDN. Yes.
>> GIOVANNI SEPPIA: Thank you, Peter, thank you for your last-minute availability.
Before I leave the floor to the last panelist, I would like to just open the floor for questions if there is anybody who has a question and I'll do the microphone guy. Working, yeah, okay.
>> Hi Peter -- general manager of -- European ccTLD platform. I'd like to thank the panel for interesting presentations. Stunning numbers. The zero percent is something that is very hard to get our heads around especially in times that everybody is cheering and clapping for the introduction of another 4 IDNs to the root, I have one --
(Laughter)
-- I have one question but I am -- maybe have better wait until Marco has given his presentation, too, but if you assume that the software market is a pretty well functioning market, there are no restrictions to bringing any type of software to market, wouldn't this logically mean there is simply no demand for IDNs, otherwise, someone would have already a long time ago brought to market 0 a browser, e-mail server that would work perfectly well with IDNs, wouldn't have cost that much, would have been pushed heaply by registries, by cultural ambassadors, just didn't happen maybe because nobody is really asking for IDN. It's a provocative question. I don't think I know the answer to it.
>> GIOVANNI SEPPIA: Thank you, Peter. A bit more provocative than the one I want to ask so anybody like --
>> PETER LARSEN: Over time we have seen plug-ins for e-mail or Outlook and Internet Explorer and some of them were working,
A Dutch company and big company and they are definitely aware of IDN. They have some things working and some are not working so they are aware, implemented but not tested correctly because I don't know if they don't know it should work or how it should work but definitely there is lack of knowledge or interest. Not sure why.
>> EMILY TAYLOR: It's a question that haunts me a lot as I'm doing this study and writing it and of course that is one explanation. It could be but I think there are -- about the Internet is sometimes someone else's problem and he said it's like this great big commune, amazing, wonderful commune where it fails because nobody takes the garbage out. And it is I think there are so many people who could be responsible for making it better but actually the standards work is difficult. These -- fact that it has taken so long for standards to develop is not just because of laziness or fecklessness, it's actually because it's difficult.
Market launched a long way because of other pressures, political pressures and community pressures. Market probably launched a long way before it was ready. I -- I heard Hongbin saying 80% of Chinese people have no concept of reading the Latin alphabet. Just there is a market there. There just has to be a market. Now, you're right, though, if domain name market doesn't step up to the plate that market will access information differently. Without a doubt and they are ready but Domain Name System underlies so many different applications, not just browsers and not just e-mail that in my opinion is necessary and in my opinion I think the market is ready.
>> JANIS KARKLINS: It's not that long time when IDN e-mail standard was adopted so it is not -- IDNs have been around for longer time but standards hasn't been around for so long time. E-mail is one of the most important functionality. As soon as that will start really working properly I believe the market will pick up and then everybody will go forward. I would be also using analogy of IPv 4 and 6, IPv6 is around for many years and there is no need of market. There will be need at one point.
>> HONGBIN ZHU: As other panelists said we have developed IDN e-mail standards and the problem we are facing is that software providers don't accept standards or they are trying to lock in users to their current system so they are -- they have no incentive to develop the next generation standards or the standards which fits the needs of the Chinese language or other international language users so I actually think there are some problems with software industry which is that if they have developed software they have an incentive to stay there, they don't make movements unless someone else or some more powerful actors push them to do that. I think that's my own opinion. Thank you.
>> GIOVANNI SEPPIA: Continuing to provocative. How long do you think it will take for IDNs to become popular? So which is the time span? Like five or 10 years? 15 years? I mean, I announce very proudly new GTLDs up and running and the market is huge because there are going to -- big markets, Russian and Chinese but how long will it take for them as well as for other IDNs to become popular and used?
>> JANIS KARKLINS: I will say 2015.
>> GIOVANNI SEPPIA: Okay. Hongbin?
>> HONGBIN ZHU: In my opinion it might take a little bit longer. Yeah, of course there are still things to be done especially with respect to ICANN's inefficiency in solving the variancy issues so there's a lot of things to be done. We need the international organisation to support best practices promote awareness on users, governments and also civil societies so I think it is about 2020.
>> GIOVANNI SEPPIA: How much of this, because we have been speaking about importance of end user awareness education and how much the different parties in the process should be involved in this awareness process. Do you think it's something that should be a common effort or should mostly stay in the hands of registries, GTLDs, CCTLDs or organisations like ICANN or Google search browser? Who is the most important let's say factor in this educational chain?
>> EMILY TAYLOR: Application providers because they are people who have the gateway to the user experience. They're not part of the conversation, nice to have -- here and I'm really interested to here what Google will say and not meant to be a challenge but Pat Kane of Verisign said that the IDN discussion at the moment is document natured by registries. Only people talking about them, about IDNs and most people in the world have not actually heard of any registries at all. They certainly have heard of the applications that they use every day. If they're not seeing IDNs in links or e-mails or in their online life they won't become familiarized and even when Chinese users are not familiar there is the possibility to have we were looking at search results where you got completely Chinese environment and the only Latin script you see in the URLs were search results and we're speaking with Irina, similar search results in Google but that is improving.
But it's got to be people actually providing that market.
>> I would certainly see ISOC's role in this campaign awareness-racing and maybe capacity-building in countries where IDNs should have much bigger role than they have today.
>> GIOVANNI SEPPIA: I'd like to leave the floor now to the last panelist, Marco Pancini, Senior European Policy Counselor at Google and also member of Google Policy Team in Brazil.
Prior to joining Google he worked for e-bay and i-Bazaar and has been quite involved in last meeting so ICANN level, is much into say policies around Domain Name System as well.
So the floor is yours.
>> MARCO PANCINI: Thanks, Giovanni, for the introduction and I try to give a slightly different perspective and try to maybe see this issue from the our point of view and from the point of view of global player in this world which is trying to improve access to content, improve communication between people speaking different languages.
First of all, I would like to start from very quick and easy analysis I've done of search keyword, most searched keyword is "Facebook." If you see over time from 2007 to today, you see people not typing URL, not searching on the search engine tells you about the way you did a few minutes ago but again if you look at the trend of how these search since the beginning since launch of Facebook, so 2007 to today you can see we have the opportunity to get animation you can see how there was a huge shift from receiving this search from English-speaking countries, there we go, to basically the rest of the world which is telling you a lot of things.
(Pause)
If you see now the viability and interest from -- try it again one last time. Okay. Not working but again if you see the trend you will see how the search number coming from English-speaking was absolutely majority in 2007 and then again it was consistent in 2008 and 2009 but since 2010 you see how the search coming from big countries and other region are becoming majority up to today where Turkey, Tunisia and Venezuela -- this is statistic which is taking into consideration both number of search and population of Internet user making -- assessing our services from the countries now free countries with three different languages and three different alphabet are scoring as three top sources of search in relation to Facebook which shows also the viability of these languages online through the platform which allows publication of content like Facebook and social networks in general.
Second trend I would like to mention is the one relative an important service we launched already few years ago and is becoming more and more important in the Google ecosystem, Google Translate, it supports today 71 languages across the world. Supporting 71 languages gives you a perspective that this is not just a result of group of translators looking to the content available line and making translation, humans. This is machine, machine-to-machine so what we are doing is taking content that is available online and using algorithm which makes content -- mixes it with request made by user through our translation machine and a chorus of this steam based on artificial intelligence again for a specific languages like Chinese, Arabic, shows viability of content is actually becoming even more effective and relevant.
So it is affecting in a positive way the functionality of Google Translate.
Third trend is connected to Google Translate that I would like to mention is in relation to wearable technology. We will see a lot in the future, technology playing an important role in improving communication between individuals in the real world but using the possibility and resources of the online world. If you take a tool like Google Glass, Google Glass has several applications but one of the most interesting ones is it will allow conversation between two individuals in the real world, again, that are speaking in different languages directly by offering realtime translation thanks to the power of the BigData, cloud computing, also thanks to the power of content online, it will happen that two persons speaking different languages wearing Google Glass will be able to communicate because of voice recognition in Glass will recognize the language, automatically translate the language and the user interacting will have possibility to listen directly to translation through the device and therefore speak in realtime in answer to the conversation without any problem.
I think looking at these trends you can see two things, first is that absolutely the Internet is global, not saying anything new but not just a statement. It is a reality which has a lot of implication in both in terms of market opportunities and in terms of development of new technology and knew application. Second argument this is not a prediction, this is something we are seeing here right now. In -- not just in how I can describe basic application actually functionality of Glass that I just described, it can be used today as smartphone with translation service realtime translation system or application it can see this is actually something, reality we can used to but also in perspective to like glooingle Glass.
So how to match the development of these application and services with the majority of user with the struggle of IDNs and struggle of the standards that we described so far.
In understanding and in analyzing the difference, gap between first development of translation services, allowing communication, both in the real world and in the offline world Google Translate works with a lot of other services offered by like Microsoft works in realtime entrance lateing content available online on website or social networks so how to analyze how this is going fast and yet at the same time IDN is struggling probably we can learn a lot and try to see how we can make the two processes more aligned.
>> GIOVANNI SEPPIA: I think tool in Firefox works now if you would like to.
>> MARCO PANCINI: Yes. As I was saying translation -- the connection is not very strong. Okay. So as I said, I put as parameter 2007 today and you see a shift between the most -- search, shift from search that were coming from English-speaking countries to countries from all over the world in relation to the keyword Facebook and which shows us today clear availability of this content in different languages and in which is in different alphabets as well.
>> GIOVANNI SEPPIA: Thank you, Marco, good visualization of expanding of multilingualism in a certain sense.
Is there any question from the attendees? Yes, please.
>> I'm Satish from India. I have a question the panel on the fact the next several million people coming on to the Internet from India are going to have mobile devices as their first devices, first computer. Now we learned that the mobiles have a problem in IDNs. What efforts are being taken by the international community and the industry to ensure that this can be addressed? Thank you.
>> GIOVANNI SEPPIA: Thank you. Who likes to --
>> As I said, the Director General of UNESCO drew attention to the IATF, that that is an issue. Also this report is a good reference material which gives very clear indication what are the main things and in 2012 when we presented previous years' report -- who was on the panel and wrote the foreword for that edition acknowledged that is an issue and from his perspective he also committed himself to talk to those who are responsible to fix that problem. Of course that takes time and but this campaign is ongoing and more we can spread the word about these gaps which need to be still crossed, faster that issue will be resolved.
>> GIOVANNI SEPPIA: Emily.
>> EMILY TAYLOR: Just supporting that, completely agree with the premise of your question that most people whether in developing countries and developed countries are increasingly accessing the Internet through mobiles and expecting much more of that, completely agree. Yes, there's efforts on the policy side. I think often issues are solved by accident as much as design and I think something that may proceed an impetus and incentive, something I heard on a panel yesterday, is the advent of the new GTLDs and particularly first of being approved are in non-Latin scripts and that large manufacturers even of mobile devices and large vendors who are in some cases applicants of these non-Latin script GTLDs are finding to their embarrassment that their own products don't work on their own products. That might well provide an incentive and motivation to do better.
>> GIOVANNI SEPPIA: Thank you. Anybody else?
>> MARCO PANCINI: What you can take for granted is there is huge attention from the ecosystem in these trends so there is not an intention to exclude from availability, from the access to content, big part of the world especially taking into consideration that as -- wrote in is book very recently the goal is the next billion of Internet users, not only the maintenance of the already existing Internet user but really looking into the next billion Internet users and for empowering them through technology which is most likely mobile.
These issues are making mobile trends more interesting for developing countries, huge challenge for us, so understanding why this is not taking up as it should, not only from a standard and technology point of view I can't understand. A lot of progress but also from cap of these standards from the ecosystem and this could be an interesting exercise.
>> GIOVANNI SEPPIA: Another question. Please.
>> Hello. I'm Irina from Russia. I'm actually keeping asking every Google representative I meet the same question so if it's not a question exactly to you probably could address it to the right person. But if I missed, then just please repeat and explain. When does Google plan to provide -- Google Gmail plans to provide full IDN e-mail support?
>> MARCO PANCINI: Very good question. As you can imagine, this is question that requires announcement from someone working on this project but since as you say I know from most of my colleagues that is a coding question and I think we need to provide an answer so I will be active, proactive, in asking the Gmail team to publish on their blog, post something to clarify this point.
>> I also appreciate if you do that and if you also probably can point out the right people to address because like all our letters just to Google representatives in Russia were left unanswered and probably just tried to reach our own people.
>> MARCO PANCINI: We will.
>> Thanks.
>> GIOVANNI SEPPIA: She's very tough on questions! I still remember the question she asked me at the end of my presentation about our environmental responsibility in the last General Assembly. So I shouldn't -- that's by the way an environmental thing. Just in case. Thank you. I think it was a good question. Good question.
Is there any other question from the floor?
A good question to wrap up this workshop in the sense there is a great call for action at different levels to make sure IDNs become successful and become a good part of the bridge to build more and more online multilingualism to allow people to get connected in their own languages and communicate in their own languages in the Internet
I don't know if there's any panelist who would like to say a final word.
>> JANIS KARKLINS: Thank you. Then let's continue working on these issues.
>> GIOVANNI SEPPIA: Thanks a lot.
Let's say -- I'd like to thank all panelists, also thank the two panelists who could not be with us today, Pat Kane of Verisign and Minjung Park. Unfortunately they were not able to join us today but we will continue to investigate the beautiful complex IDN environment with this report and we believe that this report is a small element to generate further awareness around IDNs.
Again, the report is available in the memory stick here on this desk. Please feel free to pick up some memory stick. Also it is going to be uploaded in the European morning on the urid.eu website.
Thanks a lot, everybody. Thanks to the HelpDesk.
(Applause)
(Session concluded)
* * *
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.
* * *