 Welcome everybody to the second session. This is JD Bothma from Code for South Africa talking about flossing for a healthier society. Welcome. Cool. So, yeah, I'm JD. I'm a software developer. For the first few years of my career, I was working mainly in Urla. So I'm really into my distributed systems. I haven't really done any telecoms, but, like, people say among developers are a certain way. So I started off working on a massive Erlang monolith at Swedish payment provider called Klarna. And there's nothing like sort of doing something the wrong way to learn how not to do it. And then I worked on sort of small, service-oriented architecture kind of stuff, and that was fun. So I was in Europe for the last 15 years, but I moved back to South Africa a few months ago because partly Cape Town is awesome, as you can see, and also there's so many really fascinating problems to address here. So it's a really exciting place to live, and I'll tell you about some of the cool things we work on. I work at Code for South Africa. We use technology to promote informed decision-making for positive social change, so that's a mouthful. But basically we try to work with journalists and with society in general, NGOs, civic society organizations and so on, and use technology so that they can help people take charge of their nation and solve their own problems. So we're not really a technology band-aid over problems. We try to find some people working on a problem and see where technology can make them more effective, and that's kind of where we stand. So we don't really use it for healthy gums, it's mainly for society. We're based at Codebridge, so Codebridge is a community, it's a workspace and an incubator. It's literally office space under a bridge, but it's also more than that. We try to build a sort of technology community, especially civic technology in South Africa, and anywhere where people want to come together and work on this stuff. There are other organizations similar to this, so we code for South Africa, there's Code for Africa, there's also hacks slash hackers that are organizations doing this. So wherever you are, there's probably already a civic technology organization doing something. I'm going to show you some of the things we do and how we use open source software and free software to effectively make change and have impact on our community. This website is called Living Wage. The idea was really to inspire people to consider the cost of living for their domestic workers and think about how much they pay them and how much they ought to be paying them. So in South Africa it's very common to have domestic workers that come in, they clean your house, they might cook or take care of your kids, and they often have to travel very far to get to their job and have extortionate transport costs. And it's a really contentious topic in South Africa, and this had a massive impact, we believe. At least it got 15,000 views on the first day. In South Africa that's a lot. The newspaper that sort of ran the story and sort of linked people to this microsite got 12,000 survey form responses. There was a lot of comments online and a lot of them were quite nasty because it's difficult. I mean this comes from our apartheid heritage. But all this is a client side web app. It's a bit of JavaScript that combines some data from, I think it looks at sort of the minimum wage. It looks at basic analysis of sort of what a shopping basket can look like. And it tries to allow you to tweak some of the variables of, for example, how many dependents do someone have? And what impact does that have on sort of their standard of life based on what you're paying them? This site is called NPR, we really need a better name. It stands for Medicine Price Registry. It takes some data that the national government releases on medicines and prices. And it allows you to look up generics and alternative brands and all the pricing for those medicines. So if you're not to prescribe something you can immediately see, well actually there's a free version. So instead of buying Panadol or Panado you can buy the one that costs a fifth of the price. And again it's extremely simple technology. It's a Django app with a SQLite database. I mean you don't need anything more to get this done. We didn't understand the impact that this was making until someone broke it in an update. And then a doctor reached out to us and said, sorry this website's gone. I use this every day for my patients to sort of, she works in an impoverished community and she tries to prescribe the cheapest medicine that she can that will actually solve people's problems. So again a lot of impact for very, very simple tech. This website is called Wasimap. It allows you to show localized census data. It's based on a tool called Census Reporter that was made in the U.S. also to show their census data. And as they were building it, they had a massive budget. But as they were building it they just stuck it online so they released the source code but they didn't really think anyone was using it. They weren't sort of trying to make it reusable and when we initially got in touch they said well it's probably too difficult and we don't have time to help you. But my colleagues sort of picked it up and started building our version using our latest census data. And they were really surprised that sort of actually there was quite a lot of activity on this fork. So this lets you show household, for example, the number of households headed by children which is a massive problem in South Africa in your area or the different levels of education in an area. So I was at a hackathon in a township in Kimberley where all the diamonds come from a little while ago and I compared, it also allows you to compare two places. So I compared the place where I grew up and the place where these people are growing up. And it's shocking. It really allows you to understand the difference in living standards that people around you have or people far from you. And you can start understanding kind of people's backgrounds a little bit more. This is a simple Django app with Postgres database. It reuses the MapIt API and the Google API so you can just throw in any kind of place name and you'll try and figure out what you mean or just use your current location. And it also allows some facilities so you can easily, if you want to build a localized app, you can use an API released by this or you can embed little snippets. So the next site is actually GQ's website and we didn't initially work with them. This is a project we did with sort of ourselves but with some other news agencies in South Africa. It's called the Tax Clock. So we wanted to make the national budget a little bit more personal for people because I mean the budgets are so boring especially at the national level. Like what does this mean? It means to me, the defense spending or whatever. And how this works is basically we provide the news organizations with a snippet of HTML that inserts some JavaScript that then puts in our part of the site. So they would write their article and just stick in our little tax clock. You can see a bit more of the tax clock here. So someone would enter their salary and based on that it figures out what tax band they're falling. And then tell you, okay, this is the current time of day. So if it's a work day, it would say, well, right now you're working for education. And you can immediately see, okay, sort of of my eight hour work day, I spend eight minutes on national debt or 12 minutes on basic education. Maybe you're willing to work a little bit more so that people get educated in the country. Or maybe you think that sort of the national debt isn't my problem. Why am I spending eight minutes doing that? I'd rather spend more time just working for myself. So I don't know how many of you do web development. It's quite removed from kernel hacking. But once you try to insert something in someone else's website, especially when it's responsive and especially when their site is responsive, all of these things need to resize and jiggle and stuff. And the sort of safest way to insert something into their website is to use an iframe. But iframes are quite static in their sizing. So we use a library called PIM, P-Y-M. That was released by NPR, the national public radio in the U.S. And all it does is it figures out how big is the parent window, how big is the child window, and it resizes. And there are some functions you can call so that if you're resizing things, they can resize again. So if you flip your phone, then our thing can adjust. And if you interact with our thing, then the parent window can adjust. It's a simple library that does one job really well, and it becomes static infrastructure for basically all of this work we do. So we provide a lot of embed codes for interactive ways of presenting data in news organizations in South Africa and abroad. And so just making this available is so fantastic. It's really helped us to do something. But what was especially fun with this one is we hadn't made the embed code and documentation available yet. I think you picked this up from one of the other news organizations, I think. Sorry? I think it's a lifestyle magazine. Gentleman's Quarterly. Gentleman's Quarterly, yeah. We were so happy that they found it and they picked it up. And basically everything we do is usually public on our Github. Even when it's with news and you might imagine news people are kind of scared they've got a lead and they don't want to release stuff. Our job is to release information to help people make informed decisions. We were so excited that they picked this up by themselves. They're helping us to help people be informed. But I mean, it's a lifestyle magazine, but it applies to people. Once you make something personal, you can sort of capture their attention and hopefully have a positive impact or at least start a discussion. So this isn't actually a website, this is a poster and I've got another example of this poster in the hallway if you want to chat afterwards or this afternoon. This project is basically people in local communities going to clinics and social grant offices and local government offices and asking people using those facilities what their experience was. So they're trying to understand are men having the same experience as women? Are people being asked for bribes? And the reason we make posters is a lot of these people are too poor to have access to internet. They even, a lot of them don't have high school so you have to be careful how you present the data. You can't show percentage because a lot of people just don't understand percentage. They haven't had sort of exposure to that. So you have to show the raw numbers instead of percentage and very graphical. But people can print this out and they take this to meetings with these facilities and they can say okay these are the problems we've found. Now in the first year the organization sort of coordinating this used a proprietary survey tool and it was kind of awkward to get the data out and analyze it. In the second year we joined the project and they started using Open Data Kit or ODK Collect which is an app that allows you to sort of collect survey data. You've downloaded a form, you collect the survey data. Once you have internet connection you can upload the survey results. You can upload it to any server following the protocol. So they were using Form Hub. Form Hub isn't available anymore so the next people using COVA Toolbox, another one. But it's so fantastic that this data collection tool is available so we were able to run our own when Form Hub was running slowly. So to gather this data and do this analysis and produce these we just downloaded the CSVs from this data, transformed it, put it into spreadsheets, used a bit of scripting to generate the SVGs. Only now are we making sort of in the next phase of the project are we making the general platform for gathering community-based surveys. We've been making this impact in these communities with very, very basic tech, just helping people download a spreadsheet, teaching them how to use pivot tables and presenting the data in a way that is useful for the people actually using the data. So now we're in the next phase and we're beginning to make a platform where anyone can go and run a survey in their community and try and have similar impact. And my work on this was basically to download the data and automatically visualize it immediately. So as soon as the survey is available you can start seeing results online which is exciting but that isn't where the core thing is happening. It might sound like a joke but my boss always says the most pragmatic thing we're doing is providing Excel export. A lot of our clients don't really know what analyses they want to do with the data that they collect or with the data that they have. So if we build them a platform where they can do this analysis from the start it's not really going to help them. There's a nice quote from Joel Spolsky where he says at least once every three months there's a startup that re-invents pivot tables. Like you don't want to be writing this thing again. People have done it. It's very powerful. We even have it in LibreOffice or whatever your favorite fork is of that. So you really have to figure out what technology you actually have to write. This isn't the site we've made. This is Government UK. It's the British government sort of making a website where that's pretty much the face of government for anyone. Whenever they want to interact with government, if they want to understand how to do something, if they want to pay certain things, they go via this website. They have a design principle. Build services, not websites. The actual thing that you want to get done. Like it's easy to say we need a website. Sure, we probably do. But what service does that need to provide? Focusing on that helps you get a lot done, a lot more done. And what's cool about them is a lot of what they're making is also available on GitHub. Why? I mean maybe no one's going to do something. Maybe another government can replicate this. Or maybe someone can just provide pull requests to improve their government's website, which is very powerful. And this is maybe beginning to happen in South Africa as well. So we were part of a pilot program in South Africa to develop a national open data portal. And right now we're working on a project with National Treasury to release municipal financial data. So local government financial data. Publicly. It's already been released for a long time, but it wasn't spreadsheets and on a rubbish website. We're making an API so that anyone can build an app that works on it. And we're making a sort of an example app that shows the financial health of any municipality, any local government in the country. So that people who don't understand finances can actually understand the stuff. It's extremely exciting. It's also a bit scary, because maybe this is going to make some people look bad. Maybe it's the truth coming out. It's very difficult to interpret these things without bias. A large part of this is providing education alongside releasing this information, but releasing data and then providing the tools to analyze it. But the right tools is very important. So I'd like you to get involved. And I think a lot of you already are. If you're writing software and releasing it, that is fantastic. Open source software helps us to make social change. I've spent a lot of this talk saying how little software we're using. And that's kind of the point. The building blocks are out there. You don't need to reinvent the wheel. Figure out how to get big things done with basic tools. It's the unique way. Or build those tools if they're not available. We save a lot of time, so that when we actually do write software, when we're gluing these things together, it's actually making real impact. And it's worth it. So, yeah. And just get started. Even small things can help. We're two full-time developers, generally, and we have impact on a lot of sites. Find local data and play with it. There's a really exciting thing that happens. I don't know if you've done that, but this finance data that we're working on right now is it sounds so boring, and when we opened it at first, it was so boring. And then you start associating it with other things, and you start realising, wait a minute, maybe I can find this there. Maybe I can understand this question there. Questions start popping up of what you can maybe do with this data. It's really exciting and it's really strange because it sort of develops as you're looking at the data. Take us to Battical. We're also hiring. You don't have to leave your job completely. You can come to us for six months, wherever you are, and I was in Sweden for the last five years, and I didn't find a hacksackers organisation. So, I started thinking maybe they don't need it. I mean, they're doing okay. And then the whole sort of refugee crisis happened, and they started realising, wow, we're not so open to people coming to our country. They didn't realise that before. So, maybe there's some work to be done there to help people making informed decisions. Yeah. That's it. Thanks a lot. We have time for a few quick questions, I think. So, yeah. Hello. Probably the first question is some of the services that you are mentioning are quite powerful and very, like, necessary services, right? One quick question is that as you stated, like, the one with the medicine, clearly probably you didn't foresee probably the need to have an impact analysis. Do you have, like, that in your horizons? Like, do you have some impact analysis or some of the services? Because I'm thinking if you can have, you know, that, then you can obviously motivate some of the services that you actually want to take. I think that's a very good question. So, especially, and that comes in especially when we're funded. So, we're generally funded 50-50 between contract work and from grants. And obviously for the grants that we have we have to explain and show the impact we're making. Usually that doesn't fall under us because we partner with organisations already doing these things. So, they do that impact analysis. So, we tend not to sort of do this kind of thing but a lot of the time associated with the project there's a researcher sort of analysing the actual impact. That's actually one project I did with an organisation. They wanted to do impact analysis of the stories that they write for, they're called HealthE, they write medical news and they try to, all they really need to do is gather data about what they're doing and what impact that's having. But we don't do that ourselves generally. Any other questions? Any questions from IRC? Did anybody check? Maybe not. Okay, then, well, let's thank the speaker again.