IGF 2019 – Day 2 – Raum I – DC on DNS Issues

>> SUSAN CHALMERS:  We'll just wait a few.  Good morning and welcome to the 2019 Session for the Coalition on DNS Issues.  My name is Susan Chalmers, and I work in the Office of International Affairs for the National Telecommunications Administration.  We do a great many things, but most relevant to this discussion here, we strongly support the multi‑stakeholder approach for IGF.  Just a little bit of background and why we formed the Dynamic Coalition.

Importantly, though, the work of the coalition is a way to compliment but not to duplicate or conflict with work undertaken at ICANN or at the IETF.  So we have launched this coalition last year with Verisign and affiliates, and throughout the year we've had the extreme pleasure working with a group of knowledgeable and passionate people during our inaugural year as a Dynamic Coalition.  The community showed great interest in universal acceptance, and so that is what this session is about.  What can be done to advance universal acceptance, which itself serves to advance digital inclusion, and what in particular can be done to advance UA readiness within the public sector.

It is my great pleasure to welcome our key speakers today who will help us break new ground on this policy issue.  And we have Chris Disspain; Mark Svancarek of Microsoft; Emily Taylor, the leading author of the World Record on IDNs; Manal Ismail, Director of the International Technical Coordination Department at the National Telecommunications Regulatory Authority of Egypt; Leanna, External Relations Manager for ISOC Armenia; Constanze Burger, German Federal Ministry at the Interior, Edmon Chung of the Dot‑Asia Coalition; CTO of Avilias.  And last but not least, Dr. Ajay Data, Data Group of Industries and Chair of the Universal Acceptance Steering Group; and Dr. Data will be joining us remotely later on.  We have about 80 minutes.

This is four basic parts.  First, we'll hear about the venues where universal acceptance is being discussed.  Second, we'll hear about why universal acceptance is important.  Third, we'll discuss how universal acceptance is fairing within the public sector.  And, fourth, we'll wrap things up and discuss whether the Dynamic Coalition should continue work on this topic next year or choose something else.

So I would like to suggest that we begin by inviting Chris to explain how UA readiness may figure into ICANN's work in the coming year, and, Chris, if you wouldn't mind explaining what universal acceptance is.

>> CHRIS DISSPAIN: Thank you.  Good morning.  I'm here to talk about universal acceptance of ICANN and why we think it's important and why it's an integral part of our new five‑year strategic plan.  The majority of internet‑enabled applications were created more than 20 years ago, so many of those applications and systems don't recognize or appropriately deal with a string that is either is longer than two or three characters.  In solving that issue and getting it sorted out is incredibly important because universal acceptance and domain names and email address is the key to unlocking the internet for the billions of users who don't use languages that have ASCII scripts or for people who want to access GTLDs in ASCIIs that have longer strings.

ICANN's unique strategic goals are to identify systems, foster competition, consumer choice in the internet space by increasing awareness and encouraging universal acceptance, IN implementation, and IPV6.  It's a combined goal.  Universal acceptance and IDN implementation are featured very strongly in there.

We've expanded the oversight role, and we've done that by forming the universal acceptance working group, which is a group on the board that obviously concentrate on that issue.  And UA progress being set up in the ICANN org to engage relevance.  I know that Mark is going to talk about what the steering group does.  And, Mark, can you give us the details?

>> MARK SVANCAREK:  Hi, everyone.  I'm Mark from Microsoft.  I'm part of the steering group.  The universal acceptance steering group was created four years ago, and we've been working to insure, as Chris said, that all domain names and email addresses work equivalently on all software and servers.

We're trying something somewhat new.  We are more integrated now with the ICANN organization than we were before.  We've reorganized ourselves and do a collection of working groups.  We are still ‑‑ our mission is still the same.  Our stakeholders are still generally the same.  That is, technology enablers, technology developers, email providers for individuals and organizations and government policy makers.  But the way we define and focus on them is somewhat more refined.

We've split ourselves into five working groups.  There's a technology working group, insuring that code is written correctly and developers know what to do; an email working group, which is focusing on correct adoption of internationalized email addresses; a measurement group, so we can track what our current baseline is and what sort of progress we're making; a communications group, we've always had a communications group, but now it's more tightly integrated with the ICANN resources, which gets us more resources; and then, finally, local initiatives.  We have a local initiatives team that now also works more closely with the local ICANN resources, but we have our own ambassador program.

We currently have seven ambassadors in various geographies who are creating content and reaching out to the universities, governments, and businesses.  You can see what you are on action plan is, and the plans of various working groups.

>> SUSAN CHALMERS:  Great.  Thank you.  Does anybody have any questions at this point?  No?  Okay.  Well, I think we should move ahead and explore why UA readiness and universal readiness is so important, and to do this, I would like to turn to Emily Taylor.  We have some slides here.  Emily, please, the floor is yours.

>> EMILY TAYLOR: Thank you for inviting us here to participate in this very important session.  It's giving a wider context to the uptake of internationalized domain names.  The IDN report is a partnership effort.  It is by Verisign, and the organization to all who participate help to fund it and share data through the years, key facts and figures.  There are nine million IDNs measured at December 2018.  That's actually a 20 percent increase, which seems really exciting, but a lot of the difference between the 2017 and 2018 figures is the availability of data from China.  The Chinese Registry no longer publish their IDN data, so we've been ‑‑ but we did get data for the adoption TLD.

Of the nine million IDNs, nearly four million are located in China, which really highlights different registration patterns to what we see in traditional domain names, but still, less than 3 percent of the world's domains are IDNs, and when we are talking about universal acceptance, you can see that there is a huge potential for IDNs, but at the moment that potential is not being fulfilled in registration numbers.

There were lots and lots of really good things about IDNs when they are being used.  The world map that you can see on this slide, looks at the script of IDNs and what we see, you know, whether it's a script in Russia or script in China or Arabic script across the Middle East and north Africa, the way that you have a script that is strongly associated with a geographical region, that's where you find IDNs in that script.

Also, the script of IDNs, very strongly signals the language that you are going to find on web content associated with that domain name.  These are not random patterns, so when you see, say, a script domain name, you will tend to see Russian, Bulgarian, Arabic script associated with Arabic and Persian, so on.  Where IDNs are in use, they are really helping internet users to know what sort of language they're going to find on web pages.  That's great.

We look this year, and we've been honing our methodologies on identifying parking pages.  That's of interest to the industry as a whole, and, unfortunately, when we're looking at the GTLDs, the generic top level domains, the IDNs, we're finding that 81 percent of them are parked.

Again, bringing us back to the topic of this session.  Why are so few of them in active use or in really strong use is probably because using them is still problematic.  It's still difficult to send and receive emails.  It's still, you know ‑‑ the web ‑‑ the web situation is better, but it's still unpredictable.

The two blocks on the other side, just very briefly, what we've done with the E.U. internationalized domain names is we've looked at the language of web content before and after eliminating parking pages, and the big dark blue blob on the left‑hand side almost disappears, and that is English language.  What we say is parking pages are more likely to be in English language than ‑‑ and you get a much more even spread of linguistically diverse content right across the board once you get rid of parking.

We're going to be talking much more, and we have really very, very knowledgeable partners here in the room talking about progress on universal acceptance.  There's good news and bans, right?  Browsers, generally the experience is improving each year.  This year a big change we saw is that the support in mobile browsers is much more predictable and much better, and that's usually because the mobiles are using more traditional browsers, more up‑to‑date browsers than just putting them across.  When we're thinking about internet of things and embedded devices, where they're doing thinking other thick and creating their own browsers, usually they don't support internationalized domain names at all.  Where they embed traditional browsers and up‑to‑date browsers is a good news story.  This speaks to the value of shared software libraries that supports IDMs.

Sending and receiving emails, we've heard from Mark, Google supporting IDNs.  We will hear from Dr. Data.  There's really good news and good progress of internationalized email, sending and receiving.  These are like islands that are not yet joined up.  If you are not on one of those services or if you encounter a piece of infrastructure that does not support international internationalized domain names, sorry, it doesn't really work.  Where we see no progress year year‑on‑year is using these unique email addresses, which have guaranteed to be unique as user identifiers.  That's how most people log in to popular applications.  Still, it's flatlining, and one of the interesting aspects of research that was found by a piece of ICANN research this year is that HDML5 is not supporting internationalized email addresses where, that's an inexplicable oversight.

It's not actually validating that input as a proper email address.  This year it's UNESCO's year of international languages, and to help them celebrate that, we have ‑‑ thanks to the support of URID, we have done a report on the experience of European indigenous language populations in using their languages online, and we've built that up through interviews with the Somi community and the Greco community and Catalan.  You have a large, medium, and small size language.  I do recommend that report to you.  We also do find in our IDNs when we look at the quantitative side that we do see Irish, Welsh, and we see other indigenous languages represented in the web content.

I'm going to hand over to my colleague Giovanni here for the last slides looking at the experience of dot‑EU and IDNs.

>> GIOVANNI:  Very quickly and very briefly, 13 days ago we launched Greek, and we are happy to report that it exceeded our expectations, number of registrations so far because it could be a small number because we are 500 brand new Greek registrations, but 513 days compared to, you know, the experience of other IDNs, especially in the ‑‑ that's quite a high volume of registrations, and the number is growing, and we also heard that our Greek colleagues, they also are experiencing quite high volume of registrations, which is their IDN correspondent for dot‑GR.  We have been supporting all the languages since 2009, and in 2016 we have launched the ‑‑ 13 days ago the dot‑Greek, and we have been investigating with all the pioneers of the UNES on ‑‑ the correlation between the language and the language of website content as related to domain name is Greek or in Cyrillic, and we have noticed that, as Emily was saying, that there is a correlation, so if there is a website linked a domain name in Cyrillic or in Greek, the chances are the content in that language are quite high.  That you can see in this slide.

The main point I'll leave to Emily, but the registrations are going well, and there is an interest in IDNs, and that IDNs are really enabling certain communities to be present on the internet and do have their language on the internet.  Thank you, Emily.

>> EMILY  TAYLOR:  Thank you very much.  We are asked by you, Susan, to describe why universal acceptance is important, and I hope that these very brief overviews of our research highlights the fact that there is a clear potential that IDNs enable a more multi‑lingual internet, and yet, if they're not able to be used in a predictable way, then the uptake will be ‑‑ it will be stifled, and it will never reach that potential, and you are forcing other languages.

Social media support their languages beautifully.  The domain name system really has to be on catch‑up.  You can see the summaries on the slides, but this is the key point that I would want us to be discussing and reflecting upon in this environment.  It is how do we make it more of a priority?  Universal acceptance progress is great, but it is slow, and when you compare it to the sort of finance that's being put in say, AI or block chain, this is very much the correlation, and it should not be.  Thank you.

>> SUSAN CHALMERS:  Thank you so much, Emily and Giovanni, and for your insights, your research insights, for sharing those with us.  And does anybody have any questions or comments?  Yes.  Melinda.

>> Thanks.  I was wondering if you are doing anything different in terms of marketing and promotion with the Greek TLDs than with other TLDs, like dot‑EU, for example?

>> GIOVANNI:  That's an interesting question.  In terms of what we have been doing so far is approach registrars in the Greek and Bulgarian region to try to help and promote it in Greek, and we have been in contact with several, let's say, providers of email solutions to make sure that universal acceptance is available whenever there is dot‑Greek domain name registered.  I must say that the response from local registrars is quite poor, and the reason is there is a lack of demand from the customers, and, therefore ‑‑ and also, because they do have to invest quite a lot of resources, and because the demand is poor, it's not a priority in their list.  We are not giving up.  We have several meetings planned with those registrars again, and we count on having more promotion of the IDNs, especially at the top level now in the coming months.

>> SUSAN CHALMERS:  I think we have a question from this gentleman and then Edmon.

>> There's just one risk that I actually see.  It is the risk of isolation.  If I look at domains ‑‑ I have no chance to visit their website.  There should be some kind of a translation mechanism or whatever or maybe some kind of policy that when I register a domain with a specific character, that it's not an ASCII character, I register a sec domain that's somehow a translation of that.  If I just focus on very small businesses in a local country, that's fine because a bakery would like to get an accountant in the same city, and they are speaking the same language, but maybe this is a bakery that is making the best cake or pancakes, whatever, on the planet, and I would like to buy them, and they have an online shop.  I have no chance to buy that stuff.

This is just ‑‑ it's more like a comment that we are in a situation where we have to make a good choice if we don't isolate specific parts of the world from each other.  Thank you.

>> SUSAN CHALMERS:  That's an interesting comment.

>> EDMON CHUNG:  That's interesting.  I think we'll come back to that discussion.  I have one quick question.  You mentioned that the internationalized email address has kind of the progress has flatlined.  I wonder what methodology has been using.  One of the things at the UASG is the acceptance of email address nationalization, and that seems to be growing a little bit.  Not necessarily, you know, fully aware, but I wonder what the methodology is and how that leads to the results?

>> SUSAN CHALMERS:  Thank you very much, Edmon.  Just a quick clarification.  I'm not saying that email EAI has flat‑lined.  Its program on accepting emails as identified sign‑ins, log‑ins for popular applications.  You can read about it, and it highlights all the different things that we're looking at across the browsers.  In fact, this year we're reporting, you know, good progress, which will be familiar.  We've heard from Mark, Dr. Data.

There's good progress on the major email providers, but also, this very interesting progress that ExGen is doing in India.  It shows where it works and where it's available people really do use it.  If I may, can I make a quick comment on your very interesting observation about the risks of isolation?

I view it as a spectrum.  You know, where you have total linguistic diversity, yes, each of those languages will be a pocket, which is talking to each other.  That's at one end of the spectrum.  The other end of the spectrum is a mono‑culture, where there is one language, one script.  I think we're probably in the DNS we're a bit closer to that end of the spectrum, and so I think we could healthfully incorporate more diversity in the DNS without getting close to the risks that you are describing.  There are ways that, you know ‑‑ one of the wonderful things about the internet is that it brings people together across geographical and linguistic boundaries.  I don't expect that to stop, but just as Europe has many languages which coexist without any of them being degraded, that diversity is something that we are at risk of losing if we don't make more effort to support it in the DNS in my opinion.

>> SUSAN CHALMERS:  Did you have a question or a comment?

>> Thank you, Susan.  Actually, it's not a question.  It's more of a comment to confirm what Emily said earlier.  Actually, in practice we had our IDN launched, and a few days later we had, like, 3,500 registrations, but then the majority of those are ‑‑ as Emily mentioned, and, in fact, the number became static, so no further registrations because the emails are not working smoothly.  Afterwards this number is even declining now with people not renewing their registrations, so just to confirm Emily's findings.  Thank you.

>> SUSAN CHALMERS:  Thank you.  Ram, and then we'll circle back, Manal.  Ram.

>> RAM MOHAN:  Thank you, Susan.  This is Ram Mohan.  What we're seeing here is evidence that the mere presence or access to IDN domain names or to EAI is not sufficient.  If you look at success and the definition of success in the DNS space, success really ought to be defined not by the number of domain names that have been registered, but, rather, by the amount of content that is actually accessible, and whether such accessibility is possible, and as a judge there earlier intervened, even getting to these domain names on your browser or registration system is a challenge, and so I think it very neatly highlights the reason why universal acceptance is the real issue rather than IDNs or just domain names themselves.

>> SUSAN CHALMERS:  Thank you.  I would like to ‑‑ oh, yep.  One more, and we'll go back to our speakers, please.

>> I would like to bring a bit of the local experience from Serbia where we implemented Serbian.serb.  In Serbia we are using both in Serbian case it's been decided for the Cyrillic domain will be like one Euro cent, so it was very cheap.  The number took off because ‑‑ later, when a price was increased, but still it is cheaper than normal, but RS is CCPLD.  The number went down.  Also, because people didn't really see the value.  The perception of the market is the domain names are less regular.  In a way it's not only about how to set up technology.  It is incentives on the marketplace could be something that can create perceptions that are, in fact, not very well effective in the long run.  Thank you.

>> SUSAN CHALMERS:  First we'll go to Mark.

>> MARK  SVANCAREK:  When you describe AI, and people say, well, how can I use that?  I don't have that keyboard, et cetera, et cetera.  I think the point is, well, that feature is not for you.  If you think about the alternative, the isolation is not so bad.  I have Bengali and Hindi email addresses, and I have to admit I cannot tell them apart.  You can read one language, recognize one script.  If you ask them to use ASCII, it will be unintelligible gibberish to them.  If you send them confidence that this email is from the Department of Transportation for my tax authority, from my doctor they can't be confident.  When they receive an email from someone and it's in Latin characters, they can't be confident that it's the actual perfect until it gets to their address book, I suppose.

So in regard to isolation of email, just remember that we are trying to empower some people who really can't use the Latin script in any meaningful way.  Not in any way that is providing confidence of security.  It's fine.  We still have the marketing problem.  People don't realize they have this option at all, and that's where the experiment that Dr. Data could prove my point to be true.  Thanks.

>> Thank you, Susan.

>> I must admit I was skeptical about IDNs from the very outset.  I'm not alone.  Joe is equally skeptical if not to say ‑‑ if just to say at least.  Anyway, so I would certainly pick on what Ram said.  I believe that the greatest challenge is marketing.  I'm happy to let you know that as a real opportunistic, I jumped on the bandwagon by suggesting to our members and nonmembers to create ‑‑ to establish a very special task force, and that is to try to experiment as Laos is still awaiting the delegation of its IDN, it would be a very nice opportunity for the HPTL community to come together and create a special marketing blend, a road map, if you will, which over time might be some kind of universal blueprint for new and old IDNs on the root.

We are working on this.  We are hoping that we will be able to unveil the outcome.  This is a collaborative effort involving some people here at the table, so let's hope that we will be able to encompass all the challenges and all the opportunities and reflect them on that in our blueprint.  Thank you.

>> SUSAN CHALMERS:  Thank you.  Thank you for this dynamic exchange during our Dynamic Coalition session.  I would like to move on to the next part of our session that really kind of explores the value of UA.  Manal, would you like to start, and then, Leanna, can provide us some reflections from a session that was held on UA at Cedic.  Thank you.

>> MANAL ISMAIL:  Thank you, Susan.  Good morning, everyone.  So in terms of universal acceptance, as we have already said,ist not only IDNs, it's even new GTLDs, so it's basically important for everyone and not only IDNs, but speaking explicitly about IDNs and from a perspective as well, I would say there are several parts to this, so the use of the official language of the country is basic to governments, and also to reach out to citizens irrespective of language or not.

Also, this would increase internet penetration at a national level, and it definitely bridging also digital divide, preserving the cultural identity of the country, and promoting meaningful participation as well, and also, for governments, I think, it's sometimes interesting to have, for example, their equipment or their future proof.  Requesting this as a condition in their ‑‑ it would be also something I think they would be keen to have purchase something that is future proof.  I'm sure universal acceptance has its direct and obvious benefits to the end user, of course, but also I also believe it has its own benefits.  Even to the private sector.  It's more market reaching out to more market and more users.

Also, with the digital inclusion as a priority on most of the governments' agendas and when we talk about digital inclusion, we are saying that we need to include everyone and including indigenous communities.  We need them to have access to internet and ICT in general and to have the skills to use those technologies to benefit from the digital world and be part of the digital world, so it's only sensible that governments would be interested to have more citizens online, and part of this digital world and make sure we're not leaving anyone behind when we talk about digital inclusion.  He'll stop here.  Thank you.

>> SUSAN CHALMERS:  Thank you very much.  I know there are a lot of experts in the room who are steeped in this and especially from the technical side, but it's important that especially in the IGF context that we understand why UA is important too.  One of the main things to digital inclusion this year.  Thank you, Manal, for that.  Leanna, would you like to take the floor?

>> LEANNA:  Thank you very much.  I'm on behalf of the European IGF.  Given the diversity of scripts in the region we tend to have the session about IDNs since the existence, so for all five years we've discussed different topics and different angles about the IDNs.  For this year we recognize this session.

>> SUSAN CHALMERS:  I don't know if you have it off the top of your head, but how many different scripts are in the region?

>> LEANNA:  We mostly use Cyrillic script.  The region is different and indigenous in their scripts as well, like Georgian and Armenian.  This year we organized a session that was organized with DC and DNSI colleagues, and the effort was to bring the discussion governmental representatives of our countries and also the operators to see how we can advise the multi‑lingual on the internet, and see how the policies can change.

The situation in Bulgaria is not a good in that the operators, and the government wants to operate with them and bring them into the discussion and coordinate the work that they've been doing we know that maybe it was less popular scripting IDN registration after the Chinese and Russian and Cyrillic, and Russia is doing work, and I see colleagues from Russia sitting on the registry here, and they are sharing the experience, and we solve that increase and popularity maybe because of the support by the government itself.  They are highly promoting and collaborating with the operators to make it usable in the country.  Then we had a discussion how we can promote, and someone mentioned the Serbian registry.

I would say that Serbia is doing really good in terms of making it popular within the country, and especially lately the registry accepted a policy that they would support the script of minority nationalities that have been recognizing the country.  All the minority nationalities can use their own scripts within the registry of IDN, which I think is really great in the country.  For the Armenian perspective, I would say that it's not ‑‑ we do not also have many registrations as Manal, mentioned in the country.  From the beginning we had a it is bit more, but those were at the beginning.

Mostly the public sector registrations.  Now we collaborate with mostly the language, and we want to promote it in the schools and in educational system, and we even on the string we propose to keep for all the educational institutions, the domain names and the IDN domain names free of charge.  The key to face the challenge of not on the marketing, but also as we see the growth of social media.  The problem of those people do not have the domain name itself.

When we have proposed domain names, the IDNs to do the schools by giving you this for free of charge, and we're talking, but do you provide the hosting, do you give us ‑‑ will you give us help to construct the website and, et cetera, et cetera?  If you're not doing that, we do not have stuff.  It's easier for us to take the social media platform for marketing, for what we're doing about our activities. Ist another topic, actually, of bringing the problems of domain registration itself, so it brings with it its own challenges.  Taking it a part of the IDN issue.  Still the IDN is better advancing.  Now we're in Armenia.  Because of this collaboration with the language community, and we try to do that to bring it to the government itself.

There was a discussion, actually, that if the government ‑‑ the public sector, they themselves use the domain names on the official website and official domains will be in the IDN and the international script, then the public ‑‑ the community themselves will use it as well.  All the public sector publicity and promotion.

There is an interesting case in Georgia that we have been talking about the reason of translating all the products we're inputting or exporting, it should have the scripts in Georgian.  All the things that is related to the products, and in this, they want to bring this policy that the IDNs, the domain names, should be by mandatory given for all the organizations.  Those who are registering their organizations, they should have domain names and the IDN domain names.  IDN is practical, Chicago think is a good initiative, and that could make it popular within the country as well.  This is basically the discussions that we have.  That is a good attempt of bringing the governmental representatives and the technical community and to see how we can collaborate and bring the value to this multi‑lingual internet.  Thank you, everyone.  Thank you.

>> SUSAN CHALMERS:  Thank you so much, Leanna, and I think that is a perfect segue into the next portion of our discussion, which is going to look at the role of the government in particular and the different ways that it can promote UA readiness.

For our first speaker in the section, Ram, I would like to ask you to take the floor.  Please, thank you.

>> RAM MOHAN: Thank you for organizing this really important session.  Governments are clearly a key group in driving international acceptance.  There are a couple of things to think about.  First, if you look at the mission of many governments, their missions include advancing digital inclusion as well as integrating linguistic integrating.

There is another thing.  Governments also have an ongoing responsibility to allow and to enable their citizens to access the internet through web sites and URLs regardless of the language or the length of these web sites and these URL ‑‑ these domain names, right?  There is an access responsibility as well.  It just reminds me of the Greek mathematician who said "give me a place to stand, and I will move the earth."

I think governments have a tremendous opportunity particularly in the area of procurement for various contracts and various things that they do to specify universal acceptance and UA readiness as an important criteria that comes to the table.  That is not regulation.  What that is in the area of incentives of ‑‑ to industry to get to the table, and it aligns with the mission that governments have of inclusion, et cetera.

Now, there are a couple of other things.  We know that ‑‑ or we should know that the problem that is being looked at here on universal acceptance is not a technical problem.  The problem really on universal acceptance is one of coordination and collaboration between many diverse players in the ICT spectrum.  These are far flung institutions, and having just one specific message or having a group like DNSI or UASG, they are not sufficient to get these messages across.  If you really look at how do you ‑‑ how should we view universal acceptance, I think there are really two lenses to look at it.

From the point of view of those of us that are involved with the domain name system, with the operation with the management or with the coordination of the domain name system, we ought to look at it in terms of utility, relevance, and legitimacy for cyberspace itself.  That is one lens to look at.  From the government's perspective, however, the lens to look at might be the lens of inclusion integration trust of the citizenry who they are required and there to serve, and if you look at those two lenses, I think you will find that universal acceptance is at the intersection of these two lenses.  Those I think form the basis for how governments and why governments ought to look at universal acceptance as the lever to move inclusion and to bring people online in a much easier way.

Just a couple of other things that I wanted to also mention in my time.  2001 in ‑‑ in October of 2001 I launched dot‑info, the company that I work at, and that was the first time really that this universal acceptance issue reared its head.  From there on, you know, in 2002 or 2003 I helped create kind of three rules.  Ground rules, if you will of clearly acceptance.  At that time it was an old TLD will be accepted more often than a new TLD.  The second rule was an ASCII only TLD will be accepted more often than an IDN TLD.  The next rule was a two or three letter TLD will be accepted more often than a longer CCTLD or GLD.  Those were rules I crafted in 2002.  It's a pity that here we are in 2019, and those rules still stay the course.  I would love for those rules to be for longer true.  There is a role that governments have in changing those rules.  There was a question that came online about my estimate or what we think might be the estimate for full integration of EAI, an acceptance of internationalized email addresses in registration forms on the web and on the internet, and how long do I estimate it might take, and I think that's really set up as an overnight success, and I think it's ten years.

>> SUSAN CHALMERS:  Thank you, Ram.  Now we are actually going to turn to somebody who works on these issues within our host country government, the German government.  Here we are Constanze from the Ministry of the Interior to provide us with some presentations, and she does have some slides.  If I could ask that Constanze's slides are shown.

>> CONSTANZE BURGER:  First of all, I want to welcome all of you in this room and to the session.  Also to Berlin.  We are really happy to have you here and to host this IGF, and we hope you enjoy the time and we have many good discussions, and we are here, and we will support you.  Thank you for the invitation to this real interesting session.  I want to give you a short view in our IT infrastructure, and DNS is an important brick and also the universal acceptance, and IPN agreed to Ram, public administration has an important role to bring this theme or topic forward.  First of all, digitization of the public administration in Germany is an important topic of this political legislation period.  All layers and much more involved in this topic because we have not only to handle the universal acceptance, but we have to have a look at the human resources to contracts and procurement to IT services applications, business applications, and the core infrastructure, and the layer down three.  It's very important to have this view also.

We are set up in the public administration a digitization program.  We want to have online to support every public administration to be online with services and offices until 2022.  The joint portal network bundles to different portals into a single platform, and we can sign out the first step to grant universal acceptance.

Our IT service provider of the federal government supports domains and UGF8 for unique characters.  It provides the IDM, and we discuss all these topics around universal acceptance.  This program based on the public administration information network because networks ‑‑ network infrastructures are essential for the public administration.  We need strong security and high availability for all these levels.  In the moment we are going to ‑‑ networks to respond to the threat level, to prevent future attacks and to create transparency and reduce complexity.  We have to insure that these infrastructures all support also this issue of universal acceptance.  Further IT infrastructure supports the topic.  In the DNS, in web servers, in proxy servers, and in our server systems firewalls.  Special character domains are hardly used.  We have to handle threats.  Domain names of the public administration usually are based.  Special character domains are, however, fully supported by the existing infrastructure, but I have to be honest.  They're not already granted these infrastructures.  It's a way, and we are going step by step.

Let me add some additional facts.  The DNS issues have ‑‑ they have some more strategic issues with DNS.  We have to handle that in Germany.  We have to discuss the digital severity in the case of upcoming DOH.

It's a huge discussion are involving cross‑border information leakage and the kind of business case with UH data coming up.  The next relevant topic is to strengthen the online reputation of public administration, IT services.  For instance, we do this with DNS.  The administration has to take over the responsibility for their DNS infrastructure, and in my opinion we have to take over the responsibility for the whole DNS system from our perspective.  Thank you for your attention.

>> SUSAN CHALMERS:  Thank you so much for giving us the state of affairs of Germany's federal IT infrastructure, including how it supports universal acceptance, and I'm sure there could be a lot of questions for you, but we might just for the time being and the interest of time power on and go to a different part of the world.  We'll have Edmon talk about promoting universal acceptance.  Please, Edmon.

>> EDMON CHUNG: Thank you for having me.  Building on what others have said, the government part is quite important in terms of pushing out universal acceptance.  I'll come back to that a little bit.  First, on the situation in Asia.  Many countries with actually use characters and use scripts that are not the U.S. ASCII or Latin‑based, and, therefore, it is very relevant for the country and for the citizens within the country to be able to reflect their own name.  If that cannot be used in an email address, that presents an interesting situation.  I guess some of the governments, for example, in China, in Thailand, in India, has been pioneers in the field, especially in terms of Asia and embracing the IDN, the internationalized domain names, and internationalized email addresses.

I really want to spend my time here a little bit, maybe sharing some of the experience about reaching out to governments and what have worked or, you know, what hasn't worked so well and maybe that could help us build further.  One interesting thing is that in the past we've always ‑‑ the community and always ‑‑ many who have been working on UA have focused a little bit more on talk down trying to get to the decision make and trying to get to those who can make and influence decisions.  Over the years I've now come to realize that top down and bottom up is actually quite important.  Decision makers have jam this down the process.  When developers haven't even heard of internationalized email addresses, then it presents a challenge even getting through to the actual implementation, so I think reaching out to government internal developers' web masters government IT departments.

We have been overly cautious in the past, I think, in terms of outreach worried about protocols and stepping over others, and I think from our experience, actually reaching the developer within an IT department in government, actually people feel more appreciative of knowing it even if they then bring it up to their boss, and their boss didn't know about, it or their boss heard about it, but wasn't sure whether their team was ready for it.  They actually appreciate much more than being dismissive.  Another thing that I think is important to note is that this is ‑‑ I think Ram and others have mentioned that it's not really about selling more domain names, and we need to make sure this isn't really a position that way.

It is really about making the internet's core system embrace multi‑cultural, multi‑lingual community for which this global internet now serves.  Various people have talked about the importance of marketing.  I see it slightly differently.  I think right now we are looking at a classic case of market failure.  Because of the need of diverse actors across the supply side for readiness to actually contribute to the ability to fully reveal the lateen demand, we have a market failure here.  The supply side is saying there is no demand, and, therefore, we're not preparing our systems to be ready, and the demand side doesn't even see the system because they can't use it, and it doesn't create demand.  That calls for market intervention.  I think this is what's happening.

More marketing is not going to work if there is a market failure.  That's why we need governments.  A couple of things I will end with.  In terms of the experience, one thing that is quite important as Ram actually mentioned also is fitting it into existing programs that governments have.  Whether it's about diversity or inclusion.  It could also be about local content, development.  It could be about the sustainable development goals where it actually talks about cultural heritage and development.  It also could be many governments are developing upgrade incentives, getting corporations, SMEs to upgrade their infrastructure.

If UA becomes part of that program, then it makes a lot of sense and makes it easier.  The other thing that I found about governments is that they tend to ‑‑ it's important for them to have a long‑term view.  It needs to be into road maps, strategic plans that are maybe a couple years out, a few years out.  Maybe especially in Asia.  Many years, and that comes into play also with the procurement and tendering process.  Having these terms in the government procurement process is very important.  Not only for the awareness within the government departments, but also to system integrators, to developers in that locality.

I guess I've been calling upon the community for a long time to drink our own Kool‑aid.  Whether it's the civil society or ICANN, we need to use more of these domains and use these IDNs.  ICANN.org should not be what ICANN uses.  It should use ‑‑ for their email addresses, and we should do that too.  We're working on it.  We are ‑‑ at dot‑Asia, we are not fully ready for it yet.  We know that, but we ourselves need to do it.  This is another part when we talk to government that would be very interesting, and this came out of a discussion in China a few months ago.

Closed environments may actually be the best environment to get email address internationalization up and running.  That is an environment that might get things started.  Ourselves, closed environments, fitting into government, fitting into government programs themselves and also broad bases and not just top down.  I think those are some of the summary of our experience for pushing UA.  Thank you.

>> SUSAN CHALMERS:  Thank you so much, Edmon.  In terms of the bottom‑up approach, sometimes it's helpful just to simply submit a bug report to your IT office, and that's a very good way to get people to pay attention to the issue.

We have about 17 minutes left, and we want to make sure to save time for questions and comments and discussion.  We'll have our last key speaker present, and this is Dr. Data.  We have five minutes, and there he is.  Hello, Ajay.

>> AJAY DATA:  Great to participate.  I could not join you.  Sorry for that.  Let me tell you about what is happening.  In the I participate and businesses, they have taken a step to provide an email address to every region of the state.  There are 70 million people.  Almost 6 million citizens have taken an email address.  It's the largest step.  All departments all over can learn, and ‑‑ can learn and implement the language barrier who do not speak English.  Those people who have an issue District of Columbia everybody else who do not speak English can enjoy the power of internet.

The procurement policies they are looking for.  If in the future any software or maybe as it connects on ‑‑ can amend their government policy and follow the guidelines by UASG so that these people can acquire a software or service so that the service allows people to ‑‑ and government do not ‑‑ we are creating and making a demand.  The government starts adopting easy services and provide them through created demand, and this demand allows vendors who modify their software to the UAE because they can now it becomes a push by governments to support the UEA.  Even though they have the pushing it forward, so that we know now what governments are thinking and how governments can work together in this area.  We all should know ICANN ‑‑ we are now having a working group.  That means now in the future we are going to have more coordination and we will have many, many governments accepting in their working and offering those services which are UA and EA ready.

We all need to do a separate UA and EA in our own.  We need all the people who are now UA and passionate about being online.  Yes, we all need to become ‑‑ but what we are doing here, and ‑‑ so that we are ready to demonstrate how easy it is to become UA‑ready, and there are some difficulties.  We can solve them together through the UA.

We have now ‑‑ we are now having communications.  If you are interested to have a solution with a problem you are facing, please pass it on to the working group.  Send it to us.  We'll be happy to look at it and come back with a solution to you definitely.  If you want to document ‑‑ or resolve a particular issue, can you look at it and provide a document to you and it's a document that can really help.  Easy to understand and take up solutions.  Another thing, there are enough documents provided by the UA working group.  There are documents that are available on UASG.net website.

Please visit the website, participate, and the discussions we have.  There is a link available to join our working group, and there is a link available to all the documents.  These documents are a great resource.  Millions have been spent.  These documents are very helpful to start the initiative for UA, and they have a basic group for you.  Thank you very much.

>> SUSAN CHALMERS:  Thank you so much, Dr. Data, for joining us and for providing that information so people know how to join UASG.

Now let's take some time for questions and comments.  Mark, please.

>> MARK  SVANCAREK: This is Mark speaking from government primer, and I'm a UA ambassador.  A study has been mentioned here a few times.  I urge you all to look up on your favorite search engine for UASG025.  The website is being redesigned, and I swear that by next year, we'll have to do it this way, but just be able to point you to please have a look.  This is a study that was commissioned, by the way, and my team led its development.  It was an interesting work we did with some very young people from the youth programs.  One of them this year.  Thank you for your support.

We have some interesting results that I would like to share with the table.  I did promise last year that we would do this.  On the web ‑‑ we're not talking server site here.  We're talking web.  Stands up 97 percent, which I would say is pretty decent, but the moment you look at new long domains, and by this we mean anything that's five characters or anything beyond that, that's 84 percent.  So it starts to look not that good, and the moment you go down to IDNs, that's 50%.

It's a very steep decrease, and it only looks worse the farther we go down in terms of complexity all the way to right to left in which we have 7 percent acceptance.  As you can see, we are really far ahead in terms of accepting new short domains, so ‑‑ from there it's a bit of a race to the bottom, and this shows just how much we really need to get working.  If this is ‑‑ this level of acceptance on the web level for sure on the server side it's even lower, and something that was brought up previously is in relation to HML5, and this is something to consider considering the number of influencers that we have here.  It's a conversation we need to get started with the developers.  Not only the receipt and the maintainers.

As of right now in the email, which says the default way to render HML to receive the emails, it's not compliant with anything but ‑‑ or ‑‑ which is ‑‑ that of IDNs.  The home you start unicode, it doesn't accept those emails, and the reason for that, according to the developers, is the server might not expect we can support it in this way because if you put it on the front end, then the back end doesn't accepts and creates a problem.  We could have another feud.  We could have input type EAI email.  Anybody who is ready has a very simple HML way to do it.  It's not complicated.  It's talking to them and convincing them of the point than exactly being a very difficult situation.

The moment we get on the web level, we already take one box completely and get that off our heads pretty much forever, and when we have 100% compliance, maybe that can be marched back and become just in the email, and we are done with the website and the equation.  I urge anybody who has contacts within that side of the fence to get this conversation started next year, and that we can move this ahead and perhaps this is the right venue to do it.  I don't know.  I'm just ‑‑ this is one of our key conclusions of things that we can improve in the short‑term.

>> SUSAN CHALMERS:  Thank you, Mark.  I'm sorry to cut you off.  We have very little time remaining, but I appreciate those comments.  We'll just do ‑‑ please, it you could just be very brief, maybe a minute or less.  We'll go to the gentleman next to Mark, and then to our youth participant, and then we will wrap up comments with Chris, and then I'm going to ask my colleague Melinda to share briefly some results of our survey, and I may actually have to run because I have to go to a session.  You might have to wrap up for us.

Please, sir.

>> Thank you.  First, thank you to those people for longer domain names.  That was much appreciated for me personally.  Just briefly, I with find it tremendously helpful if maybe ICANN created a site where I could register any sites which didn't accept maybe longer domain names or those with non‑ASCII characters so that information set was sent to the registered owner of the site to give them information on why it would be a good thing for them to amend their site to accept different domain names.  I think coming from ICANN that would be seen as quite helpful for the site owners.  Just basically going to a page inputer domain name and then leave that to ICANN to reach out to them and that would be a useful tool.

>> Hello.  Sorry.  I'm from the Brazilian Steering Group ‑‑ Internet Steering Group.  I would like to talk a little more about the opinion about acceptance.  I've talked with some people that ‑‑ our system administration, and the thing of this community is that actually they are not really convinced about universal acceptance, and in some cases of the HTML5, about what it says ‑‑ if you read the solutions, they are really with the possibility of the websites like stopped working, crash, with the HTML5 stop accepting EAI.  The idea is maybe take some approach with the community to convince the ‑‑ do not do the opposite work and after the importance of universal acceptance.

>> SUSAN CHALMERS:  Thank you so much.  We're going to go to Chris, and then I'm going to go to Melinda and then perhaps if there is some more discussion afterwards, you like to moderate that.  Chris, please.

>> CHRIS DISSPAIN:  It was passion that set this off, and it was passion that drove that through as quickly as it did.  There's a 1989 movie called "Field of Dreams."  It's about a guy who built a baseball field in the middle of a field because if they build it, he will come.  If we built the baseball diamond but the gates are still locked and not only can the players not get in, but neither can the audience.  We talk about marketing and say that it's all about the demand from the audience and once they're there.  That's not right.  It's actually about us having built, it opening the gates, and then they will couple.  It foals like we've built it, but the gates are still locked.  We are on the infrastructure.

We are governments and businesses who make an awful lot of money off of this internet thing, and in between us we cannot influence everybody who needs to fix this, to fix this, it would be an extraordinary thing.  We have ‑‑ we built it.  We made IDNs and surely we can ‑‑ I don't know how, but I do know that one thing that isn't going to work is coming back here year after year and talking about it.  We actually need to do something.  Thank you.

>> SUSAN CHALMERS:  Thank you, Chris.  I couldn't have asked for a more perfect segue for closing comments.  One of our goals this year was to conduct a survey that was meant to give us a little more insight into the challenges that the public sector faces.  Both in terms of awareness and how we need to position things.  Unfortunately, we were not able to get funding.  We might potentially get that next year.  I'm not giving up hope.  What we did was launch a pilot survey.

This is the rest of the slides we had today will send out to our mailing list.  If you're not on it, let one of us know.  I won't take you through all of this today.  I'm going to hit a couple of highlights.  One of the reasons we do this pilot is first and foremost to define what our target odd Yen is, attempt to define the hypothesis that we are testing, and then understand who this network is.  How do we actually reach that target audience of respondents?  All of this then will inform the survey that we intend to get funded next year.  The data set was too small to be relevant, but it did highlight a couple of things that brings us back to a couple of comments.  Whether it's staff or men /STERZ or folks in a CIO office.

They could be the ‑‑ from the private sector.  They could be consultants.  They could be contracts.  What we found was this group of people is the most knowledgeable.  Knee got the most confidence that they can impact change.  I think it's it is a good intersection of both Ram and Edmon and Chris brought up.  We need to get a good hook for the public sector.

One of the ‑‑ in addition to what we've talked about here today and tying this to SDGs, something that came up in our session in at Euro Dig was why don't we look outside of SGDs and think of something more purely commercial?  What are the really grand tactical programs that are a little more commercial in interest that we are going to tie this to?  Whether it's an IOT project, smart cities, some other sort of project that might have a little more wow factor and be more heavily funded, that that's someplace we could also consider influencing.  To say that inclusion is not important.  It is.  They could be parallel paths.  That is something that we want to consider next year in focusing our efforts.  We've spent this first year of DC‑DNSI and thank you all for participating.  It's great to have you here, and I know the majority of you are on our mailing list, and we appreciate the that.

As we transition to next year, we want to get more strategic and more tactical.  Take all the learning that we had and push those out and fulfill this responsibility of evangelizing and piloting what the proposition is, and we know as was mentioned, we have a body of data from the UASG that can then help people with the implementation percent peck.  That's what we look to focus on this next year.  Can you expect a report from us as we get the transcript.  We'll define what those next steps are that we are proceeding to send it out to the group and how we can really get a lot more tactical in our efforts to evangelize and somehow work UA into public sector day‑to‑day life.

With that I'm going to let everybody go somewhat close to time, but were there any final questions?

>> Very briefly.  If you find a page that has problems with universal acceptance, go to UASG.  On the top part you see log issue.  There you can guile a complaint.  It will go straight to ICANN and from there they can process it.

>> MARK SVANCAREK:  This is Mark.  It's staffed and works because Microsoft receives issues from there.  I can verify that.

>> SUSAN CHALMERS:  Good to know.  Thank you for your ‑‑ one more.  Really quick.  I forgot to mention.  I have some leaflets that highlight the summary of the IDN world report.  If you like one, come and get one.  Thanks.

>> I thank you so much for your time, everyone.  Take care.

(Session concluded)