Captions provided by @chaselfrazier and @whitecoatcapxg.
      >> Hi, everyone.  Thank you for staying.  We have a pretty amazing talk coming up.  But there's been a bunch of conversation after the opening remarks, I encourage you to continue your conversation, hopefully there are still muffins out there.  But otherwise we will start pretty on time.  Okay.  So with that, I'm going to pass it over to Robin.  Joel?  Yes.  I'm really bad with names.  So.
      And this is OpenStreetMap U.S. able to be involved in.  So it's really exciting.
      "Law Clinics and the OpenStreetMap Community."
      By: Rauvin Johl.
      >> Give me approximately 60 seconds to get the PowerPoint up and going.
      All right.  Thank you for your patience.  My name is Rauvin Johl, I'm a third year law student at Harvard law school, and I'm here today to talk a little bit about a project that I worked on as part of Harvard's psychological clinic.  And I'm applying for that project through OpenStreetMap, you may have heard of them.  And I'm here to talk about our project, our goals, our processes, our outcomes, but I also wanted to talk about the experience of working OpenStreetMap from the perspective of a noncommunity member.  And ultimately my hope is that this discussion will help three different sort of sub sets of the community.  The first is future law clinics or students working at law schools who are interested in engaging with OpenStreetMap.  The second is other outside organizations who want to engage with the OpenStreetMap community.  And the third is OpenStreetMap U.S. community members who are looking to form relationships and collaborate with outside organizations.
      So before I get too far, I'm going to introduce myself a little.  As I mentioned, I'm a third year law student at Harvard, and I participated in the cyber clinic.  It's to use Internet intellectual property matters.  The clinic is way to get students hands on experience while we're still students.  So, though, we're supervised by law school faculty members, I was supervised by professor Crawford, we really take the lead in engaging with clients and making client decisions.  I became involved in the clinic because I'm interested in intellectual law, so I thought it would be a good stage to get involved in my career.  And I was assigned OpenStreetMap as my client.  I don't have any background in geography, GIS programming.  My bachelors degree is international affairs, which meant that I actually heard of OpenStreetMap in the context of international development class I took where we discussed humanitarian aide work in Haiti.  But until I became involved with the clinic, I hadn't so much as logged onto OpenStreetMap.  So I say that to stress the fact that I am very much an OpenStreetMap novice.  Just as I'm a novice in the law with only two-thirds of a law degree under my belt.
      I say that because I want to make it clear that I am not here to divide universal insights in have law open database in day-to-day license.  Or to get ultimate guidelines on how to partner with the OpenStreetMap community.  I'm here to talk about my very specific experience engaging with OpenStreetMap, ODbL.  With the host of these next types of collaborations can build on steps.  I'm also an amateur baker, which I'll allude back to.  I promise it's somewhat relevant.
      So the project, the cyber law clinic project.  Our task is to explore around license, open map.  We selected geocoding as our use case because it was often mentioned as a point of tension between the community and the license.  Our goal wasn't to give the interpretation of the license or of how to tie geocoding.  Was to identify the points of tension and to adjust the steps moving forward that the community could take towards easing that tension.
      Our process was to talk to people.  We conducted interviews with ten different individuals from various parts of the OpenStreetMap U.S. community.  So that included large organizations like Mapbox, map zen, we spoke with a group called finder and members of the academia, and state and government employees.  We also talked to them about having OpenStreetMaps, how they contributed, and also what they didn't use OpenStreetMap data for and why?
      The answers we received actually had a similar set of concerns.  Our interview ease explained that for the most part they didn't use OpenStreetMap data for geocoding.  And they gave two primary reasons for that.  The first hinges on accuracy completeness.  OpenStreetMap data is primarily contributed by individual users.  As a result, the database has more some types of data than others.  For instance, street information is very thorough whereas address information and point of view interest information tends to be a little bit more spottier.  There's more of it just in the area where the community tends to be larger.
      This makes it challenging to rely exclusively on OpenStreetMap data, especially when bruising geocoding software that's meant on consumption by the general public.  For instance, some of the navigation software.
      The second concern we heard reiterated was a concern about share abrasions.  Interview ease didn't take a firm stance against share like.  What interviewees did say is that the scope were at times unclear and OpenStreetMap with things.  Risky.  We heard that they were going to use OpenStreetMap with something like a geocoder, but turned off when the provision would be triggered, especially when they're looking at supplementing OpenStreetMap data with other potentially proprietary data sources.
      In response to what we heard, we drafted a memo addressed some of the concerns the interviewees we spoke with and suggested first steps and recommendations.  Our recommendations fall in three main umbrella categories.  The first is specificity.  We recommended that OpenStreetMap considered the level of specificity at which it provides guidance.  There are great many resources available for community members or individuals who are looking to use OpenStreetMap for project.  But there currently aren't any geocoding specific resources.  So that's a thing that the community could consider developing.
      The next umbrella area of recommendations that we had was clarity.  As I mentioned, there are many resources available for those who are interested in using OpenStreetMap data for a project and for individuals who have questions.  But those resources aren't usually the most user friendly.  They overlap, they sometimes critic each other.  So really looking to streamline that is something that the community could do to ease confusion and concerns about the scope of the license.
      A strategy related to that would be perhaps -- having someone on the board on the U.S. board whose job it is to be the relationship point of contact.  To connecting individuals who have questions with those who might have answers.
      The third umbrella area of recommendations is just reflecting on OpenStreetMap's mission.  The interviewees that we spoke with and who reflected concerns about the scope of the license and how it distracted with geocoding represent a small slice of the overall community.  And so if making the data more useful for that -- those types of projects isn't a priority, is a part of what OpenStreetMap wants to have their mission, then it isn't necessarily -- shouldn't necessarily be a priority to make big changes.
      We also suggested some future research.  Both for law clinic projects and also just for community members.  For law clinic students, we recommended projects such as analyzing the implication of different interpretations of the open database license.  Or providing options and recommendations for clarifying the content of existing community guidelines, based on perspectives expressed by the community.
      And the third will be drafting perhaps clear geocoding cases in permissible uses of the data.  So in a community I understand reading our memo.  It's been distributed to the OpenStreetMap board and also to license working group, and it should be made available to the general public some time in the near future.  So that is what we did.  But the memo was actually a very small piece of the overall experience of working OpenStreetMap U.S.  Much of our time is spent adapting to narcing open source community.  So in the lead up to this conference, I was trying to think of an antidote to working with OpenStreetMap U.S., and I wanted to sound like I was having a Ted talk and have some a really great psychological study that I could talk about, but what I kept thinking about was cake.  So I bake a lot in my free time.  Mostly cake and cookies, occasionally bread, and for the Fourth of July, I made a cake with a flag design, relates to when you sliced each slice, it looked like a American flag.  It's not as complicated as it seems, so I made some diagrams how to do it.  You make four cake layers.  Two reds, one white, one blue.  And you cut all the layers in half, and you cut a circle out of the middle of the layers, and you just stack them so that you have a flag.  I had a friend helping me cut the cake layers in half.  And also double-checking my arithmetic because I'm detail oriented, and I wanted the proportions of my cake to match the you'll proportions of the flag.
      I had sprinkles.  I had sprinkles for stars, and they weren't 50, and I wasn't going to make the cake tall enough for 13 stripes.  But it turns out after some Googling that the field, that's, like, that blue rectangle on the flag, is supposed to be 40% of the overall width of the flag.  So the other 60% length-wise is supposed to be red and white.  So I had an eight-inch cake layer, so eight inch diameter cake, four inches in radius, and so I thought given that information, that I needed to be cutting a five-inch circle out of the middle of the cake.  But my friend, who's insisting that it should be a three inch circle.  We're both pretty stubborn, so things escalated and escalated and escalated.
      [Laughter]
      Until we had spent an truly embarrassing amount of time and energy arguing about the diameter of cake layers.
      At which point we realized there's actually more than one way to bake the cake.  See, the way you built the cake depends entirely on whether you want the flag to face inwards -- we were both right.
      [Laughter]
      But because we were talking past each other, it took us a long time to realize it.  So the cake that I baked is the one on the yellow plate and then the blue plate is the example of the alternative way to do it.  Ultimately both methods get you to the same place.  Now, it seems like a pretty silly story, but it's actually a tidy summary of some of the take aways I had from working with OpenStreetMap.  My clinic partner and I struggled to adapt working in a open source community.  Law school is about problem solving.  So we began clinic experience confident that we were going to take a look at the license, maybe talk to some people, read a few cases, and then solve whatever problem we were presented with.
      That was in retrospect a little ridiculous.  A lot of the conversations with Alyssa and Cal who were point of contacts for the project, where it wasn't our job to provide tidy answers because OpenStreetMap U.S.?  Just a top down hire key.  Interpretation of licenses or suggestions about next steps for the community aren't taken seriously if they don't involve some level of buying an input.  That was difficult for us because it flew in the face of the standard legal approach to problem solving.  And it wasn't until we really pushed ourselves to acknowledge that there's more than one way to bake the cake.  More than one way to build a strong OpenStreetMap that we were willing to engage meaningfully with the project.
      With that in mind, I wanted to share what I believe are the three ingredients for success between collaboration, OpenStreetMap, U.S., and outside organizations.
      The first is having difficult conversations early.  If you take the time to brainstorm early on, whether about the way that you want your cake to look, or the type of project that you want your law school clinics working on, you save yourself stress down the road.
      We rushed into the project, settling on OpenStreetMap as our client for the clinic project because we needed to select a clinic project client early in the semester.  And OpenStreetMap U.S. was willing to work with us.  Rushing into this decision, meant the project attracted some negative attention early on, which for someone new to this community and this type of community was disheartening.
      Having spent more time in the outset of the semester discussing the goals of the project and the way we were going to engage with the community, there's a chance some of those road bumps could have been avoided.
      The next ingredient is just to listen.  That's pretty straight forward.  But it's also critical.  There's been a lot of discussion recent years about have corporate members of OpenStreetMap, the OpenStreetMap community feel about share like and why and so it was important that my partner and I look beyond that history and actually engage with and listen to our interviewees.  That's part of what makes working with a law clinic or any other outside organization so valuable.  The fact that we're a fresh set of eyes unbias participants who aren't familiar with the scripted responses of the community to certain issues.
      It was remarkable how much common ground my partner and I were able to find when we asked questions of our interviewees and then really listened.  Not just what they were saying, but how they were saying it and what they weren't saying.
      To connect it all back to cake baking, moving you past desperately trying to convince someone why you need to cut a five inch circle out of the center of your cake and closer to actually creating something.
      The final ingredient is being open to differences and flexible in the face of them.  Being up front and listening helps get past some roadblocks.  But there are some differences between working with an open source community and other organizations that can't be solved quite as easily.
      For example, we had several discussions at the outset of the project about issues of privilege and what would or would not be shared the rest of the community.  Our instinct, given our legal background and training was for less transparency.  Narcing close setting is more efficient and has more honesty between the client and counsel.  But when your client is an organization and that organization requires openness and collaboration, you have to think critically about how you're going to design a process of both the client and the counsel.
      Questions that we considered are whether or not make our interview notes public.  Whether to make our interview memos public.  Whether to solicit community feedback on early drafts of our project or just to share the final version.
      Ultimately be compromised.  The OpenStreetMap U.S. board held a town hall to discuss the project and create the Slack page for community members to get feedback, ask questions, give input.  We also shared memos with our project coordinators but didn't necessarily share those beyond that.  And although we didn't put input on drafts, we do plan to make our final work project public with the board and with the licensing working group.  Those compromises were inevitable, but if I was redoing the project, I actually might change the calibration of them.  Tilting more towards transparency as much as possible.
      Ultimately the biggest take away from my metaphor, which I have now completely beaten to death is that even when you're standing 85-degree kitchen covered in flower arguing about the diameter of cake rings, it's important to remember that you're all there to make a metaphorical cake or in this case a very not metaphorical map.  So if I were to give advice to future law students involved in OpenStreetMap, I would advise them to be patient, listen, and embrace the open data around them, even if it seems a model that seems almost directly opposed to what you learn in a classroom.  And OpenStreetMap and its local chapters and foundation consider if and when and how to work with law clinics, I encourage them to remember that we like to solve problems and to give tidy solutions.  But that we can be encouraged to do other things.  Though, we are used to operating in a realm focused on compartmentalization attorney client privilege, we can adapt to a more stakeholder approach.  And I say all of this to the fact that there are certainly different ways to bake a cake or build a map.  But at the end of the day what really matters is baking it the best that it could possibly be.  Thank you.
      [Applause]
      I think technically I have time for questions if anyone has any.
      >> I'm really interested in building memorandum of understanding, in this case Mexico.  There are big organizations, and they have so my -- the other idea is how the hell are there going to agree.  I know they might be flexible, but there's a limit there how much would you push it or which would be the argument you think we could build together, so they feel that they are not being scammed or something like that.  Because I mean I can tell that they are interested, but they are not so up to open anything.  So how would you see that?  Mostly because thinking about national statistic, that is to mean statistics, we each in turn cannot be segregated to lower levels, which is the key of any mapping.  Getting lower, assuming in and in and in.
      So what would you say about it?
      >> All right.  So that was a good question.  I don't know if everyone heard.  But I think the heart of it was asking how to encourage partners like government organizations to be willing to share national data with something like the OpenStreetMap community.  I think to answer very small part of that question, simplicity helps.  The instinct is often to give all of the information to give people links to pages big documents discussing, you know, OpenStreetMap user data for and what the project are going to be.  But I think that that can be scary for organizations who aren't as familiar with how the community structures and with open data.
      And so trying to really reflect on a document that says here's what needs to be open and here's what we could do with this data and here's the impact it will have on you and making it seem less scary is a really important first step.
      I think we're a little bit behind, so I'm going to get offstage so someone else can talk.  But I will be around.  Please feel free to find me and ask me more questions.  Thank you.
      [Applause]