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


















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





>> SUSAN CHALMERS:  Good morning, everyone.  Welcome.  We will just see if the transcript is running.  There we go.  Welcome to this morning's session, Best Practice Forum, Creating an Enabling Environment For The Development of Local Consent.  My name is Susan Chalmers, I am one of the lead experts that has facilitated this discussion over the past two months in conjunction with my colleague, Stuart Hamilton who is with the International Federation of Libraries Associations Institutions.  Stuart will be joining us in a few minutes.  He is making an intervention in another session for a moment and he will have to duck out and come right back.

Thank you for joining us today.  I wanted to begin by explaining the discussion that has been unfolding over the past two months.  This best practice Forum is one of five that was established its initiative of the Internet Society.  So the point of this session is to be able to come to agreement and kind of tease out of discussion best practices on various Internet policy topics.

The one we have been following is the development of local content.  And so this discussion has unfolded over the past two months on a list serve, and I would just like to give a brief description of how we framed the issue, and how this discussion has unfolded, and that the ultimate outcome of this exercise is going to be a document on the best practices for local content creation.

There is a draft document.  John Laprise who is the drafter of this document, to my left, will explain the paper as it stands right now.  But for the moment, the way that we organize the discussion and the way that we will organize the discussion today is that local content is a multifaceted and complex issue.  What the list agreed upon is there is a tripartite or three part structure to the issue and that will be the format we will follow today during this discussion.

So the first area that is relevant when discussing local content creation is Internet infrastructure.  And my colleague, Michael Kende, Chief Economist at the Internet Society to my left will be covering that area.  The second area that we will be exploring are the regulatory frameworks or policy frameworks, legislative landscapes that are relevant to the creation of local content.

And so we look at areas such as telecommunications regulation, copyright law, perhaps tax, tax schemes for devices which can enable or assist people in developing their own content and putting it on line.  And my colleague Martha Giraldo to my right, she shall be speaking on area two.  And finally, the third area we identified during this session was the extent to which the human capabilities and capacities for creating mobile content.  So that really focuses on the personal capacities, digital literacy, for example, is a sub issue in area three or a degree of web accessibility and the locality is also an issue.

And our colleague, Sylvain Baya, who is joining us remotely from Cameroon, he shall be addressing that area.  And so without further ado, I would really like to introduce or ask our panelists to introduce themselves.  We have these three areas that each expert will speak upon.  At the end, we are lucky to have some insights shared on the technical policy and technical capabilities that relate to local content.  To that extent we have Glenn Deen down at the end from NBCUniversal so I would like to invite the colleagues to quick introduce themselves and then we will turn to John Laprise to present the paper.  And after that we will have the presentations on the three areas and my colleague, Stuart, will return by then, and we are really keen to have discussion after each area with the audience on proposed policy initiatives that can facilitate the development of local content under each of these areas.  And at the end what Stuart will do is he will conclude the session by explaining how this conversation will continue and then will share with you suggested next steps.

So if we could just go down the line, that would be great.  Please, briefly introduce yourself, where you are from.

>> MARTHA GIRALDO:  Good morning.  My name is Martha Giraldo.  I come from Colombia and after working different ministries and private institutions and ICT appropriation, I'm right now from the Civil Society ICT specialist for development.

>> MICHAEL KENDE:  I'm Michael Kende, the chief economist for the Internet Society.  Before that I was a consultant with Analysis Mason working in fact quite a bit with Internet Society and with other Governments around the world to help promote deployment and usage of ICT and Internet access.

>> JOHN LAPRISE:  My name is John Laprise.  I'm the consulting scholar on this project.  My background is in academia, Government and business.  I have consulted for the U.S. Telecom industry.  I was a consulting scholar to the Government of Qatar and was part of their broadband steering committee and I have also been a professor at Northwestern University.

>> GLENN DEEN:  Glenn Deen, I work in NBC to engage with the Internet.  So while I spend a lot of time in places like the IGF, I also go to other countries where we really try to engage on the problem of how do you get content functioning in ways of the ‑‑ how do you help people create content more efficiently on the Internet.

>> SUSAN CHALMERS:  Thank you.  Might Sylvain Baya be on the line and be able to introduce themselves?

>> SYLVAIN BAYA:  (?).

>> SUSAN CHALMERS:  John, would you please share with us your thoughts on the paper and where we are headed from here?

>> JOHN LAPRISE:  So this is a new initiative for the IGF and for the Secretariat.  This paper was crafted largely upon the contributions of members of the list serve prior to the drafting with some additional content based on my own research as I was asked by the Secretariat to do.  When the draft went up, and this is indeed a draft, and it's a working draft, I fully expected that there would be discussion and input from a broader community.  This was intentional. 

There are gaps in the draft intentionally left there because I knew that based on the list serve it was not perhaps as broad a discussion as we might get with a broader distribution and in fact during the course of this week and in the week following IGF I will be working on this draft on an ongoing basis to incorporate those comments from the community into a final working or into a final draft.

That said, I have been following the commentary on line and I just like as a side note that my own work, I'm a historian and social scientist.  I believe in rigorous research.  I also taught statistics for four years, so I have a healthy knowledge of the use of numbers.  When I'm looking for evidence, when I'm writing, I am looking specifically for both anecdotal information which points me in useful directions, but to make broad generalizations, I'm looking for solid empirical research.

And in that sense, I'm looking for best of all would be peer reviewed research.  And I will be uploading a draft of material that I have been working off of for the past couple of months that have informed some of the discussions, some of the writing I have been ‑‑ that's in the draft now.  And that's an ongoing process.  That said, I look forward to additional comments from the community to improve the draft and come to some sort of, maybe not Resolution, but at least some sort of points of agreement.  I think that's a good summary.

>> SUSAN CHALMERS:  Thank you, John.  It's good to have a very hardy academic approach to compliment the on‑the‑ground stories that have been shared over the list serve for the best practices.  Now I'd like to pause to ask the audience if they have any questions about the Best Practice Forum at this point, the process that we followed or the draft paper, please?  Any questions?  No?  Okay.  Well, in that case, then we will proceed to our first area, area 1, which focuses on Internet infrastructure and the locality; and, Michael, please, over to you.

>> MICHAEL KENDE:  Thank you.  So the first area was Internet infrastructure, and the relation to the development of local content.  We in our discussion focused a lot on content infrastructure, on the hosting centers, the data centers where the content can be located, and the importance of local hosting for promoting the creation of content, that is where the content is sitting is important for helping to stimulate more content for reasons that I will get to.

So starting with the observation that a significant amount of Internet traffic in Africa and Latin America and other developing regions is coming from abroad, so most of the African traffic would be, at least the Internet transit is all going to Europe for the most part, for Latin America is mostly going to the U.S.  That's kind of the starting point.

And a lot of that is content.  So historically this is a fairly common phenomenon.  At one time the majority of European traffic was going through the U.S.  In fact, when I was at the FCC in the late 90s, we went to see one of the original network access points, kind of a precursor to the IXPs in Virginia called May East, and we are told that at one point up to two‑thirds of European traffic was going through that switch which by the way was sitting in a parking garage in kind of suburban Virginia, very unprotected and very different than it would be today.

But that traffic consisted at the time of two types of traffic, a lot of tromboning traffic coming from traffic and going back either to the same country or the region and accessing content that was hosted and sitting in the U.S.  And the reason for that, I think, is important to take some perspective at the time, and this is not the late 90s, but earlier.  Europe was very expensive, it was very monopolized.  Everyone had to connect to the U.S. anyway for historical reasons to get access to the Internet, so they used that for all of their traffic for accessing content, for sending emails, all of the tromboning.  So what happened is the markets liberalized over time, introducing competition.

The European Union took hold and everything started to regionalize.  There was a lot more investment, and these neutral Internet exchange points emerged to take care of the tromboning, so, in other words, all of the traffic that was going from one, someone in the same city to someone else in that city that used to go through the U.S. now would be exchanged locally or nearby through an internet exchange point.

And eventually, those Internet exchange points became hubs for content.  And now you have these enormous hubs in London and Amsterdam serving the entire region with content.  So you are starting to see this take place now in other areas.  There is certainly more investment in cables into Africa, into Latin America, between Latin America and Africa.  There is more terrestrial fiber being put in place and through the efforts of many, including the Internet Society, there is a lot more Internet exchange points emerging in these countries.

And those are starting to take hold.  They are starting to prevent tromboning so they are being used to exchange local traffic locally.  The one part that hasn't taken place yet is the content hasn't started to migrate to the IXPs.  So IXPs are great for distribution and very efficient for distributing content because through one connection, you can access all of the ISPs or all of the ISPs connected to the IXP, but people haven't really started taking advantage of that yet.

So we have been doing some work on this, and, for instance, in Rwanda, the top 20 Rwandans, six are Government and they are hosted locally because they are required to be, and the other 14 top Rwandan websites are hosted in Europe or the United States.  The impact of that can be quite significant.  If content is hosted abroad, the reason for it, when you talk to a content provider, the reason they are hosting abroad, if they know they are hosting abroad, is because it's cheaper.  Right, I mean, if you go to one of these big data centers, Go Daddy or something in Europe or the U.S. it's extremely cheap to get almost unlimited hosting capacity, and a lot of the content providers don't even know they are hosting there.  They go to a developer (Lost Internet connection) to bring it back  not just the physical distance but because of the congestion.  So the international capacity is expensive, it might be under, under, they might have too little of it, and, therefore, it will congest and it takes longer and there is jitter and everything else.  And then the results ultimately for the end users are they are paying more because they are paying ISP to bring it in and they are paying for data by the bit because their ISP is paying for data by the bit and because of latency there is less usage.  We know you will give up if something runs slowly and most importantly you are not creating a content market, because others see this.  They don't see content market developing, there is no scale being developed in the content hosting, so it's inhibiting the market.

So the decisions of the content providers that exist to host abroad are inhibiting a development of a healthy ecosystem in the country.  And we have seen examples, for instance, Google has a very aggressive programme to put caches in countries around Africa and other places, the Google local cache programme.

And when they put a cache in, traffic will skyrocket, so they will put a cache next to the IXP mostly for You Tube videos that are very static content.  The traffic will explode.  That means that the ISPs are selling more data, so they get more revenues and they are facing lower costs because they don't have to import every one of those videos every time someone wants to see it from abroad.

So that would be the result then of more local content coming back home, as it were, as well.  So we talked on the list about a few solutions for this.  One thing that's very important is awareness, just an understanding.  Not everyone, again, knows where their website is even being hosted.  We talked to the CEO of an ISP, a major ISP in Rwanda, and he had no idea where his website was being hosted.  So we looked it up for him, and, of course, it was hosted in Europe.

So he is paying his own capacity to bring his own website back to his users.  Other cases similar to that, one thing people are doing is hosting what they call a local content Forum, where everyone can get together in the same room.  The content providers, the developers, the hosting, the data centers, everyone in the same room to talk about the issues.

And in one case, there was a guy in Nigeria who runs a kind of a Bally, a Hollywood online streaming service and he was hosting in London.  Every time he would come home to Nigeria, he would ask people are they using his service and they said, no, no, it's way too slow we can't download the movies, but don't worry though we can watch your movies on You Tube.  He is thinking why can they watch it on You Tube but not from me.

So he went to this local content Forum and learned what an IXP was and what a cache was and the reason that You Tube was faster because they had a cache in the country that was serving people locally, and obviously then realized he had to do the same to grow his business so that people weren't watching his movies off of You Tube.

So a lot of it is just awareness, understanding what the implications are hopefully through these types of, obviously this Best Practice Forum is aimed at increasing awareness.  Efforts to increase regional scale are very important.  Amsterdam has a far, far bigger IXP and content hosting than the size of the country would warrant, and it's because there is no regional barriers so people from all over Europe and elsewhere are accessing content from Amsterdam.

And while everyone wants to be the local hub and the content hub in the region, it may be that, you know, every country is not going to have its own hub for all of the content.  And so it's important to reduce local barriers.  Governments hosting locally, like in Rwanda and other places can help to develop the infrastructure for everyone else acting as kind of an anchor tenant for putting in content and building up the hosting environment.

A lot of people talked about the use of CCTLDs that are hosted and the servers are local and that increases the speed for local content hosting.  So there is a few things that can take place, and ultimately, you know, through this Forum and these two other pieces we really want to put in, you know, identify the best practices to bring this content locally.  It's not just the infrastructure.  That's what I have been talking about.  It's the other pieces, but putting content locally will lower the cost, lower the latency and help to promote ecosystem for developing more content.

>> SUSAN CHALMERS:  Thank you, Michael.  I would just like to ask the audience a question here.  In terms of content and traffic, are there participants here who are aware of whether or not traffic trombones out of their country to Europe or the U.S. as Michael was describing and coming back?  If I could just get a show of hands.  Yes.  No.  Okay.  Well, one of the things we would like to do as part of this exercise is to encourage sharing of stories about the state of the Internet infrastructure in your locality.

So if anybody feels so inclined throughout the session, please, please do raise your hand because we will be inviting participation.

>> APARNA SRIDHAR:    I just had a more general thought in response.  I don't know if this is the right time.  Hi, everyone, for those of you who don't he know me I'm Aparna Sridhar, counsel with Google, and I appreciate you cited our example in terms of the cache.  I think one thing we have been mindful of is in general it's best to move content closest to the end user and that drives cost down.  That's obviously correct.  But I think there are some challenges of local hosting that you haven't sort of fully played out at least in the presentation.

And I would just name three principle challenges.  One is in country hosting especially in the Internet Ecosystem isn't very robust is often not as resilient or as secure as hosting services abroad, and so if you are a local content producer, you may very rationally want to contract not only for services that are cheaper, but that have more security, better resiliency.  The second is that up time can be a significant issue and in places that there isn't ‑‑ there are significant power outages.  So it's not an issue of the local hosting provider is not a good provider.  It's just simply that the overall infrastructure is not sufficient to support hosting.

And then the third point I would make is, you know, some,  for certain kinds of content, there are,  and really, for hosting providers in general, there are regulatory considerations that need to be taken into account.  So if you have a regulatory framework that is, in which intermediaries carry a lot of liability, for example, or if the Government is very intrusive and there could be significant censorship concerns, there are good reasons on the flip side that you might be weary of posting locally.

And so I just think that we ought ‑‑ obviously we agree in the general case that hosting as close to the end user is beneficial, but there are some countervailing considerations, and I think at least on our side, we would certainly say that a mandate to host locally would not make sense.

>> MICHAEL KENDE:  To take the last point first I don't think we are talking about a mandate.  I was presenting the optimistic point and certainly when you do one of these local content Forum, you know, people will raise that, but sometimes it's a perception that, and that's one other good reason to bring everyone in the room where people will say, you know, there is just a perception that there is less resilience or less quality locally, but they haven't really tested it.

And, for instance, the power issue is actually related to cost.  In most places they will have backup generators, but that is what makes it more expensive to host locally and abroad.  So that's definitely an issue as well, and we shouldn't gloss over that there are legitimate reasons why it's cheaper overseas and why people may put their content there.  They should just really understand, I think, the local market and make sure that it's not a post from three or four years ago that may not be true then.

On the subject of intermediate liability, we are obviously going to cover the regulatory topics.  Clearly that's an issue for Google.  The local newspaper is not an intermediate and they are not protected by the fact that their content is hosted abroad because they are all based in that country.  So no one, I mean, that's not an issue for the local content providers because they are already based there and where they put their content won't protect them from the results of what's in it.

>> AUDIENCE:  Right, but it's an issue for the hosting provider.

>> MICHAEL KENDE:  Oh, that's true.  We will raise that.  It didn't come up in a lot of the discussions we have had, but I guess that could be true though I haven't had that come up with the discussions over the cost and the other issues.

>> SUSAN CHALMERS:  Thank you.  One of the distinctions that Michael raised during the discussion is one thing that's so tricky about this issue is that there are a lot of stories about the policies that indirectly facilitate the development of local content, and so I think that a lot of what Michael was just going over could be attributed to that indirect policies that indirectly facilitate the creation of local content. 

And just while we will be moving into that second area dealing with regulatory issues, there were examples that came up during the discussion of Government policies that requires that Government information be hosted in country.  And so it was also discussed that sometimes policies like these indirectly encourage local content development because they feed into the infrastructure within that country.

There is economic development of the infrastructure in that country to meet that policy.  So I think that since we have now we are kind of on the topic of area 2 and regulatory landscapes I would like to turn to our second panelist, Martha Giraldo, will be discussing this area too.

>> MARTHA GIRALDO:  Thank you.  I was asked to bring up some examples on content development and maybe perhaps as to this regulatory, which they are not very good news about it in individual, but let's start giving examples of content development policies.  Not as a national basis, but maybe at some ministries in my case, public regulations.  For having kind of context of my country, Colombia has a population of 47 million people and the rural population is about 1 million.  It's like a fourth of the country, population of the country.  And we are just starting a new period of national precedents.  We had a reelection that is continuity of the programs which is good in some ways, but in some other ways it's not.

And about some picking up that I was asked for picking up examples, good examples of content development in my country.  I just wanted to bring up the attention to Ministry of Agriculture initiative which is called Agrinet.  It is the agricultural sectoral network which provides timely and reliable information to actors in the production chain to support decision making to profitability and for delivering market opportunities.

They have services, the main one is access to a digital library with about 20,000 full text specialized documents.  They also have a statistics analysis and reports on productive chains or sectoral information out of 44 databases and they have online courses and Forums, product offers and demands, national and wholesale prices, supply prices, foreign trade prices, information about credits, information about productivity.

They also have Agriclimate reports with weather reports by region with vital impact on agriculture and products.  They have multimedia applications for delivering this information to small producers, making it easier for them to access that information, and they also have the possibility of accessing the information by producers via cell phones.  It can be downloaded at domestic prices, the supply prices and with the weather alerts for the specific region.

With these services they are then supporting the user capabilities for planning, production, investment and all related processes.  But I bring this experience which started in 2005 and initially it was a good idea, but it didn't really work too much because it was a huge work for a ministry to really coordinate that, and we can say it passed from a good idea to a really good project due to a second project, which is the reader product, which is documentary agricultural information network for Colombia, so a network of public and private agricultural institutions. 

They get together to generate knowledge and provide services for the agricultural sector.  They are really giving their information to Agrinet and that's how Agrinet is now collecting this information.  They are just having the environment, but the information is coming really from the institutions.

So this is a sector initiative to strengthen all of the agricultural sector, and as an example, we have as the members for this initiative or the Ministry of agriculture itself, ten universities, one mixed institute, four research centers which we can find agriculture general centre, and also coffee research center, the main one in the country, the sugar cane and the CIAT, which is a big international research institute for associations of producers, the palm producers, the palm and the banana association and also fowl is collaborating with this function.  So these institutions are the contributors of the Agrinet at the basis and this big digital library.  That's how it came.

That's the first example I wanted to address.  The second one is Ministry of Culture, the National Library for Conservation and Digitalization of the National Library Habitat.  the objective is to give meaning and make accessible the national library.  (Lost Internet connection).  Strengthening higher education.  This is a joint action among the Ministry of National Education, and the higher education institutions with the purpose of conceptualization planning, production, management, use and research of open digital sources.  There are actually 84 institutions involved, and but they have done a search and harvesting algorithm for protecting information from all of these universities and really helping the people to make just one stop shopping  having an agreement on the specification and also having some kind of evaluation on the resources according to a set of parameters. 

So this is in terms of examples, but although they are very good initiatives, they are not really the result of a national policy in terms of agreeing on how this information should be delivered and there are more good ideas, and I know all of the Latin America, if I can talk about on a regional basis, there are lots of good examples on local initiatives for content development, but I think that really we need to think about the policies that the countries need to work about, for example, that of the local hostings and some more to really make this information be accessed by people in the municipalities.

In my country, for example, the Ministry of Communications, which is actually here at the Forum, it is connecting all of the municipalities of the country.  We had a donation of the, for the library of our countries have,  I say one library in the municipality and will have really a country with a nice, very nice important infrastructure as a basis for delivering the information for the people of the municipalities up. 

But my concern is that phishing on the Internet, we know that we are growing and growing and the Internet is growing in contents, but I know you have had this opportunity of putting content on the Web and then trying to retrieve it.  It's very difficult.

So delivering policies on content are really, really, very important that the countries support, give all of the support for the development of content, but not just at the moment.  We are doing in my country, we have a lot of information, and that's very good news, but at the moment the law just, they are working on the law, and it's making as an obligation for the ministries to publish the information.

But we are lacking this opportunity for having a policy on some agreement on how to publish that information, so it has sense for the local people and the local people can find it and not putting them to go to one, and the other and the other, and it's a real, very difficult, very, very difficult.  And we are really losing that big opportunity for having this infrastructure, agreeing on this aspects.  It's a long term and it's a long work, it's difficult work, and we need to start working on that.

So that's what we are really looking for in my country having some agreements and starting talking about that because it's not an initiative coming from the Government, and so from the other institutions we need to really put a word on that and start moving these necessary policy regulations, policy declarations in that term.  So that's what I really wanted to talk about.

>> SUSAN CHALMERS:  Thank you so much, Martha.  Thank you very much.  I'm going to change out with Stuart here who is going to take the helm, but I would like to invite Constantis to say a few words on this area too.

>> AUDIENCE:  Hello, everybody, my name is Constantis.  And I'm part of the policy team.  Going back on what we heard, I participate in this group, and one of the things that was identified in the context of this regulatory framework was the need for robustness, and what we mean regulation that can support local content.  Now, there were various areas identified as part of this robust regulatory framework.

There were participants that talked about information, about intermediate liability, about copyright, but we didn't really, the group didn't really delve into the specifics, but one of the things that was particularly interesting here is that there are two streams essentially, the incentives that these regulatory frameworks can provide for the creation of local content for inviting companies to go to the countries and host the local content.

We have the challenges in terms of infrastructure.  So I would just like to say that when we think about this regulatory framework, we think of it in a much broader sense, not as an intent of the incentives that can provide content to Governments. Thanks.

>> STUART HAMILTON:  Thanks.  My name is Stuart Hamilton and I here for Susan Chalmers to talk about this session in the setting of the session.  I have been, on behalf of the international formation of library associations kind of co‑moderating the practice discussions as they have gone on and I'm extremely pleased that we are here now talking about the document.  I'm looking forward to picking up discussions now.  I will bring in at this point Glenn Deen and after Glenn we will bring in Sylvain Baya who is a remote participant.

>> GLENN DEEN:    The way we found that content is (Inaudible).  I was trying not to boom too much.  The moment of trying to find what you watch was relatively simple in the past.  And it was a one‑way flow from professional creator down to a consumer.  One of the changes that's come about with the Internet is we are all now readers.  And local content isn't just content as coming from maybe the local television or radio station. 

It's content that people in a community have created for sharing amongst other members of the community.  And in fact, when our group got together at the very beginning, we had to sort of find what is local content, and it's because of this change that a lot of people have to incorporate this new idea that content isn't just that stuff I get from TV or radio or the library.  It's stuff that I create too.  Now, we have a problem.  The problem is this.  How do we get other people to find my content?

It's hard enough to find dealing in something very simple like English language which, of course, everybody in the world speaks English, right?  No?  If I want to share my content, I have to get the word out to everybody and how I get the word out is through meta data.  Now, in the past, meta data would be only the concern of a bunch of engineers, but I just down at the end of the table here heard meta data being discussed.  How meta data standards were being required by the Government so that the information being generated could be shared and found and discovered.  So discovery becomes a big problem.  One of the problems I would like to talk about a few minutes is in fact local meta data and how do we find content around local meta data and how do we do it in different languages.

If I create a video today, I don't speak Portuguese.  I don't know anything about Portuguese to save my life, but if I want to share with people who speak Portuguese, I might want my meta data to be in Portuguese as well as English.  Today I don't have a way of doing that.  If I publish content on the Internet, if I go to Facebook and say here I am on the beach, that's in English.  One of the problems I would like to highlight is in fact how do we start translating meta data, very simple descriptive things like this is me on the beach, how do I translate that into other languages so it can be found and accessed around the world because I may have friends that are Portuguese speakers and they want to see when I tag a thing that I'm on the beach. 

They may not need to know that it's me on the beach, maybe somebody else, but, and when we have countries that are having content they are creating in their local context, whether it's the various bits of data we have just described or maybe it's the coffee prices or maybe it's the new research around distribution of coffee into a particular marketplace as generated in that local language, there may be people outside of the local community that also want to find that content and learn about it.

In this way, we then start building a bigger global Internet community because we have content that's shared amongst all of us between our local communities.  And the heart of that problem is how do we discover it.  And so we start our own best practices and I think one of these that emerged in the discussion was in fact we may not have a best practice around how to translate meta data, but we discovered we have a problem in translating meta data, and one that maybe we need some work done on.

So one of the policies in the future that we can see that would actually help local content would be the creation of concepts around meta data, how do we publish it, translate it, how do we make it available for search so we can connect people watching this content with the content they want to consume and do it in the languages that they naturally, normally work in?

>> STUART HAMILTON:  Thanks, Glen, smart discussions we are having.  Here is we have Sylvain Baya in French and I understand when I was out of the room that it was difficult to hear what he was saying but thank you for the input there.  The references to meta data are great to librarians on the end because that's a buffer and has been for many, many years and it clearly is becoming more and more important as we move into this area of discovery.  It's a term which I think it's gone from being relatively niche to very well known now.

>> GLENN DEEN:  We had the revolutions about meta data and trying to hide it.  In this case we have lots of meta data we want to make it available, easily available for people to find and consume.

>> STUART HAMILTON:  Absolutely.  Absolutely.  Okay.  So I think the plan was for us that we are going to wait until Susan comes back to see if she is able to help us a little bit with the incoming feed from Sylvain Baya so I want to open the floor at this point to see if there are any questions or comments.  I can see one immediately there.

>> ELLEN BLACKLER:  Hi, I guess volume isn't a problem at this end.  I'm Ellen Blackler with Disney.  I wanted to link a couple ideas that we raised earlier to think about the role of the kind of Government content that I think you were talking about in Colombia.  One of the things that might be a way to kind of link up the market opportunities and the regulatory issues is to think about the Government content and generated content as kind of the anchor tenant idea.  To the extent some of these issues are about, some of the hosting issues are about having the scale and scope regionally and locally to make the economics work out well, the Government could think of itself as providing that kind of scale, and so being an open anchor tenant, you know, starting to build the local hosting facility so that the economics play out in a way, you know, differently than they are playing out today.  So that might be kind of a best practice to think about.

>> STUART HAMILTON:  Definitely I think that that's an issue that we talked about, and thanks for highlighting it.  Obviously that helps and I think that story like that and the Google one also help in that it shows that it works.  Hopefully, that with Google, you can see that when they put a cache in the country, that it's faster, and people can see from their own experience, and that the Government websites hopefully are up and running and that gives more confidence in the infrastructure.  And one hopes that they would make sure that there is the power of reliability and everything else.  So I think it's also good just in demonstrating that the concept works as well as creating the scale and fabric for putting the content.

>> AUDIENCE:  I would say the influencing factor there is the scope of eGovernment in the country because obviously if the Government is not too adroit with computers yet, then the amount of content they are going to be hosting is lower than in a country that's already well advanced in that.  So that's a complicating factor.

>> STUART HAMILTON:  Thanks, John.  We have more from the remote.

>> AARON VAN KLYTON:  I haven't presented yet.  I'm Dr. Aaron van Klyton from University of Greenwich.  One of the research I conducted on Internet Governance in Africa dealt with how to generate local content.  And that was a real struggle that came through in all seven of the countries that we studied.  And I think that the discussion of local content from what I have heard so far and done some reading around it, the more nuanced the country is ethically speaking in terms of different languages being spoken in the country, the greater the difficulty in terms of policy and Government structure and so on.  So is there any kind of discussion around what to do outside of French, Spanish, and whatever other main languages there are moving into the Wolof and other small languages how do countries deal with that?  Is there any kind of roadmap to deal with that?

>> STUART HAMILTON:  Just a second, maybe John will comment on a couple of things that came through in the paper, as we have an audience that is people from all over.  I wonder if there is anyone that may want to comment if you are from a non‑major language group about how that might be dealt with.  So is there anyone from the floor that might be able to address the question before we go back to the panel?  You all come to share? 

Okay.  Well, we did actually, we did actually have some input from the discussions from different sorts of countries and maybe, John, you might want to pick up on that.

>> JOHN LAPRISE:  I think, well, as the story of technology, my gut instinct is to always trust the users.  So the best way to go about this in a practical sense is to enable users to create content, and in this day and age, that largely means putting a Smart Phone in their hands and giving them or enabling them to have a relatively inexpensive broadband package and access in connectivity.  Because people will figure out how to use Smart Phones.  They will figure out how to make content and put your reliance in individuals to make that content.

Speaking more broadly in terms of favoring particular communities or bringing that out,  we found some countries that have initiatives to favor particular content based on particular language groups.  It's unclear how successful that is, I mean, it's Government, not quite subsidy, but Government influence to do that.  That's one suggested or that's one suggestion, I should say.

It varies very much with the country and the Government and the people and the situation.  It is one of the challenges for this particular thread of the best practices is that it's very unique, and it's culturally specific, it's culturally relevant.  And what technologies that people employ in a particular locale vary depending on product, depending on socioeconomic status.  It's a big problem and I don't think there is a one size fits all hosted as the access to content.

I suppose that as the international community of Internet Governance has moved away, or embraced it as sort of initiative because it's a complex web and it deals with the politics of the country, who is in charge, from which tribe is he or she because that often plays out in terms of who gets priority or what languages get priority, so anyway, thank you.

>> STUART HAMILTON:  I wanted to comment on something and go back to the discoverability element of this as well, because what libraries can do in this sort of situation is to bring together collections of content in minority languages, for example, and to make people aware that it exists and to actively get people to contribute more.

So the question recalled to mind a project I knew in Mexico, sort of a digital archive of Mayan Indians in the Yucatan province where libraries have been active in creating digital archive of oral history contributions and non‑written content, but that's been so successful, they had sort of 10,000 people coming to their archive in the first six months.

Once the awareness was out there that this content sort of existed and could be put together and people could come and add to it, then it really took off, and I think making the small collections of material in Wolof or other languages discoverable and raising the profile of them is a job for organisations like libraries.  It doesn't have to be just libraries.  Any other sort of group could do it, but when we are trying to bring the Government agency in as well, if you think about coming in from the other angle then placing emphasis on discoverability could yield results for small collections and get the ball rolling in these areas.

>> GLENN DEEN:  One of the things I like to put a plug for here is when we talk about discoverability in meta data it is a fragmented space of use and standards.  There is no standard for meta data.  In fact, the problem is when someone decides to go or work on a meta data project they create a new standard and they create a new standard which goes to another country and creates another new standard.  Some of the standards are proprietary.  So one of the things I would like to put out as a statement I believe in very strongly is that we do need standardized meta data that is very broad across different types of medium, but also we need it to be an open standard.

And in fact one of the new frontiers I think that the different standards organisations really need to start looking at is in fact working with meta data, standardizing it and making it open so we can build protocols at the IGF level, access and discovery in other places like WTC, for instance, build into the browser and let it work together with infrastructure elements and it's sort of a new frontier we haven't done before, because if you look at the way we have our existing organisations tiered around the information, we have policy, we have getting the data there and we have getting it presented in your browser.

We don't have a group that you would go to that would work on meta data and make it work with all of these other various groups in a consistent way.  And I think that maybe sort of in bits and pieces we evolve this next generation Internet.  We have had, you know, the semantic web has been around a long time and it's very popular in technical and research communities.  Maybe we are at the cusp now where it's becoming an irrelevant question for everybody because everybody is now creating content and trying to share it around and finding it is at the heart of the semantic web.

>> STUART HAMILTON:  That's kind of interesting as well because I think from the library community's perspective and obviously this is what I know best, so I will keep coming back to it but meta data and meta data standards is something we have worked on for many, many years and we have embraced how hard a shift to open standards.  So you have open archiving projects, et cetera, et cetera.  Now, what the library community is not very good at is actually letting people know what they are doing.

And we are getting a little bit better at it, but what I suspect quite strongly is a lot of the fundamentals for an agreed standard exist in our community and others as well.  So I do like the idea that a recommendation is to move towards something that everybody will use, but this is kind of intriguing to see how that take up would kind of work as well, because we have other projects, creative commons brings to mind in that shift to get people to try and take on something new across the board when it comes to licenses and contracts, but I think it could be definitely explored within the context of this Best Practice Forum.  But it's making me make a note to see what my colleagues are working on because I think it could definitely help.

>> GLENN DEEN:  In my business of TVs and movies, one of the things we have to grapple with is that there hasn't been in the past a universally applied even identifier for what a piece of content is.  And, you know, even before you can do anything else, you got to find it, and before you can find it, you have to be able to say what it is you are looking for.  And we lack the identifier.

Now, the entertainment industry has adopted a thing called EIDR, acronym for entertainment industry digital something, I forget.  I always forget the R, but EIDR is a universally applied piece of content and actually track it through edits and conversions it allows you to find it.  One of the things we are lacking on the Internet is that same universal identifier for non‑studio content, non‑professional content, something that you could actually use to name a piece of content on Facebook and have it named the same thing on You Tube and the same thing on ZK and have it named the same thing everywhere. 

So when you say I want to find this thing, that would be a great place to have an open standard and then places like the IGF could innovate around it in their standards so we should talk afterward and maybe one of the recommendations we have is get an open standard.

>> STUART HAMILTON:  This is getting us a little bit into the sort of discussion that we wanted to have at the end of this session and I'm conscious that we still have Sylvain to speak, but I'm conscious that Susan might be the only person to translate for us so I will hold off just a second, but we wanted to move into a discussion at the end of the session about what policies are missing.  And I think Glenn has got us not a policy but that initiative that we might be able to look at, but I wondered from the people in the room if there was any thoughts on policies that need to be created.  This was one of the things that really came out from our online discussions was that there were plenty of stuff for us to discuss, but there was clearly some blanks in terms of policy that needed to come up.

So I would like to take sort of any observations or thoughts from the audience in the room about what you thought might be missing in your own country, what you have heard from our panelists which you don't have in your own country that might be quite useful, or conversely things that you have heard from the panelists that work that you think would never work in your country.  So I am looking for sort of input here.  

This is not just an opportunity for us to go through the paper we have written online.  It would be great to get some thinking from the people in the room.  So I had a professor who would literally at this point turn the microphone off and sit silently until someone put their hand up.  It made me squirm quite a lot.  I could try that technique or ask if anyone has any comments, so opening the floor to you.

>> AUDIENCE:  Thank you.  While everyone else is thinking about what they are going to say, one thing I thought of when we were talking about kind of all of the complexities that content producer looks at when they are deciding where to host their content and perhaps one of the things Michael identified is maybe people weren't really deciding they were not aware of all of the factors that could affect their users' experience. 

We might think about that as a tool kit issue and all of the work we have done, ISOC and other people have done around IXPs and making people understand what it could do and how it could work.  Maybe we need to think about just even kind of factually saying here are the factors that affect your user's experience and that you should be taking into account when you think about all of these different parts of kind of the value chain that affects what the, you know, I keep thinking of it commercially, that affects what your commercial experience would be when you try to reach our customers.

>> JOHN LAPRISE:  Thanks.  I think that is a great point and one thing we are trying to do is not make today the end of the discussion but a way point moving forward and keep the online Forum going and identify issues like this where we can keep the discussion going and keep identifying new themes as we move forward.  So I think that's a great one, just kind of as, not an aside, but a lot of, so we are working on a project with the Ministry of ICT in Rwanda, we being the Internet Society, on this very topic. 

We were trying to get data and just having trouble getting it piecemeal so we are working in one country to try and gather the data about what is the cost, the latency, the traffic, hosting, and so we will certainly share those results here and elsewhere to further this discussion, but I think that's a great point to, you know, as we move beyond the IXP to think about other areas where we can create tool kits and best practices.

>> AUDIENCE:  I think this would be valuable on small producers and small language issues because a lot of the producers that are going to produce that kind of content aren't going to have a lot of, you know, coming to the enterprise with background knowledge.  So that's something I think that could be valuable to someone looking to build a website or an app or anything in minority language in their country.  You know, some instructional material about how to think beyond the coding experience.

>> JOHN LAPRISE:  Something I would like to bring to this discussion here and this has struck me when I have been working on the written part of the project as well as the discussion here is that we should bear in mind that the vast majority of content, the mast majority of local content is not commercial content, it's content between people who just post on Facebook or send an email, or people who are not looking for any sort of financial gain out of a given piece of information that they are putting online.

So when we are thinking in local content, we should be thinking broadly, both from the commercial perspective but also from the non‑commercial perspective, people who want to share pictures with their family with extended relations.  So there is an aspect of non‑commercial content generation that's really important for local content and for the life of these cultural communities that we need to continue to fold back into our discussions.

>> AUDIENCE:  My name is Greg Wiomy.  My English is not good, so I'm a French speaker, but I will try to speak English.  So I think that we must have a lot of, service provider must have a data centre because in Africa, the thing people don't believe to put the content in data centre because they don't have Internet or Internet literacy.  So they want to go to U.S.A. or Europe because they want to have availability of their content. 

(Lost Internet connection) Determining what is a fake email or a scan email or spam or not spam, but a phishing attack, for instance, and I think this is something of a learning curve for people who are unfamiliar with the technology.  They are learning what to trust and not trust and a trial or error and I think over time that will resolve.  In short term it's problematic and what you can do to address it is talk to other people especially regionally and across in your case, across the Middle East about other countries that are doing programs on media literacy, because there is a lot going on in the region that you may be able to.

>> STUART HAMILTON:  That's from our perspective a good seed, the idea of sharing information across the region is to actually find out what's already being done and with the setting of up of this discussion over the last two months with the online mailing list and this panel here, we don't see this as being anywhere near done.  There is a lot more to discuss here.  So I mean, it's worth talking a little bit I think about where we might go next with this.  I think Nicholas' ideas about this being picked up in the regional IGFs is a good one and there is more and more national IGFs each year, and it's clear that there are some national conversations happening which it would be great to get some feedback from.  As you will all be aware we have the draft of the paper online.  The commenting period for that is most definitely open so we want to have more people taking a look at that.  If you haven't seen it, its extremely easy to comment.

I'm wondering if one of my colleagues has an easy URL for people to get ahold of it.  I don't have it in front of me.  Maybe you can ping that to me.  The plan is that all of the comments we are going to get here at the IGF and during the commenting period will be worked into the, sort of the next round of the draft, and then we will work towards finalizing it.

And then I think, I mean, we can, we had some pretty decent discussion there and I think it can go further on the mailing list so we are going to keep that moving and we are going to invite more of you to pick this up.  So this really isn't a done deal.  John, what's the sort of immediate next steps for the next revision of the draft?

>> JOHN LAPRISE:  Well, the next step is I'm going to be working on the draft over the course of the week incorporating comments, reacting to comments as they come.  I have just forwarded a bunch of the primary research, the second research I have been using to supplement the work of the lists, and that will be made available as well as a librarian archive of background material.

So the next step is I'm encouraging everyone to please continue submitting comments, go to the website which is ‑‑ I have got it up on my computer, and I have been Tweeting it.  So get on Twitter and take a look at me and you can find the link among other things.  But please do comment.  Your comments are what drive this process.  And without your comments the outcome is not as strong as it might be, so please comment and your comments will be incorporated into the final draft of this document.

>> STUART HAMILTON:  We do ‑‑ I'm sorry.  Yes, please, go ahead.

>> AUDIENCE:  Just quickly in continuing this work, this was one of this group were the discussions that were so very interesting so I think that it is very important that we all even after the compilation of comments has been done and we have the text that we continue and delve more into issues that have been identified.  One of the things that was particularly interesting was that we had a lot of case studies.  A lot of people will bring in personal experiences from Africa and from Colombia and from other parts of the world, identifying the issues in relation to local content.

Technology issues and then we have the experts like Glenn who also do this work in other fora so continuing with this work and bringing all of these things together and tying them possibly either to regional IGFs or the consultation rounds that will be happening around the MAG; and, of course, having as a vision Brazil 2015 in order to actually have a much more concrete and more robust contribution when it comes to local content I think is particularly important.  So I would like to invite everybody to continue the discussions.

Those who have not been on the list subscribe on the list and continue to feed into with more information because I think that this is an area that we can all work to actually get something out of it.  Cheers!

>> STUART HAMILTON:  Just before we close, and I'm extremely pleased to say we have input from Sylvain Baya to close.  The URL that will take you to these drafts is Review.INTGOVforum.org  so that's review.INTgovforum.org .  And, Susan, I understand ‑‑ is this connected to the ‑‑ no, we have another remote contribution.

>> REMOTE MODERATOR:  He is from University of Helsinki, Finland.  He has a general comment or he has a general comment.  Three key words seem to be cache, libraries, and meta data; cache on the level of infra, I guess infrastructure, removing unnecessary traffic, libraries collect local content.  Multilingual meta data enables relocalization.  So these are the three thoughts she wanted to share.

>> STUART HAMILTON:  Very good.  That's great.  Susan, over to you for conveyed closing remarks.

>> SUSAN CHALMERS:  We were finally able to have just a few phrases from Sylvain I wanted to share because they were quite nice, and the third area deals with the human capabilities and capacities and Sylvain Baya wanted to say that the group has worked for ‑‑ I'm translating here so please bear with me ‑‑ the Working Group has worked for over two months to collect and identify best practices which create an enabling environment for the development of local content ‑‑ sorry.  But perhaps one of the most important aspects that, is the fact that the man, the person is at the centre of the ecosystem of the promotion of local content, and it's for that that factors tied to infrastructure, to applications, or even to regulatory environment should all draw back to the person and what we are capable of.

And that's ‑‑ I'm afraid, since we are out of time, he is sharing some further phrases, but we will have to incorporate these into the paper within that clever platform.  So, again, thank you, everybody, for being here.

>> STUART HAMILTON:  We do have a closing word from our sponsor.

>> AUDIENCE:  Thank you very much, and thank you for the work that went into this process.  I want to emphasize this year is the first year where the IGF is launching these Best Practices Forums.  The idea is to encourage the IGF community to work towards tangible useful outcomes.  We have been saying a lot that the IGF is a place where there is a lot of discussion, but actually nothing tangible for delegates to take home.  This said, and I'm very happy about the way Susan and Stuart led this work.  This said, this is not a negotiation.

This is consensus phased and it is very important for us from the technical community as this is the way that we work.  If we have no consensus, we put the issue aside or we just recognize that we don't have consensus, and that's perfectly fine.  The objective being to identify common ground, and if possible, build new common ground.  So with regards to the draft outcomes, this draft was put together over a very short period of seven weeks over the August month, and a lot of people were on vacation.

I just want to emphasize that this is ongoing.  This is in process of development.  Comments that were expressed today will be taken into account.  If there is a need to extend or refocus the draft, that will be done because we are working on a consensus phase.  So I just wanted to make that point extremely clear.  Thank you very much.

>> STUART HAMILTON:  Thank you, everyone.  That concludes this session.  Could we thank our panelists in the traditional way?




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