 Awesome. Hi, I'm Rob Schechter from Oregon. And here to talk about OregonLaws.org, a site I made pretty much by myself in my part time. I'm a law student. I also have a job or two. So again, the same kind of thing showing what can happen with just a little bit of time, but with the right data coming into it. And so basically, I'm going to run through quickly the status quo and how this came to be, but then really focus on the benefits and implications of this new way of making legal research sites. Let's see. So anyhow, the status quo. You can see here in the big screen or the big picture here, this is what the state of Oregon puts online from their statutes. Now, it looks pretty hard to read. And I really find it really hard to read. So just focus on the negative for a little bit. The words run all the way across the screen. It's as big as your window is. But if you ever actually open up a paper book and look at a book, there's certain conventions that we've used in typography and publishing for a couple hundred years of how do you make something so you can read it legibly. But then also looking at this document that Oregon wrote, it's actually really, really well structured what the legislature does. And it's like this throughout the statutes. There's actually a really nice outline form in here. But it's not being reflected purely in the individual representation. The legislature did their work, but just this 1% of the final presentation doesn't really reflect what's going on behind the scenes. In the smaller views here, I have a couple of the four pay legal research sites. I took the names off to protect them. And this is just showing, as far as usability in the top right, you can see that's a main menu of your options. If you want to begin doing legal research, I'm sure everyone has seen this. And in the bottom right, this is a search result. And yeah, these are small pictures, but just to get the idea. And in the bottom right, that's searching for the word, oh, that's murder, is what that is. And it's all the little places where murder appears somewhere in the text of something. So it's not a lot of understanding what is murder. And now maybe a computer can't really understand that kind of thing too well. But somewhere in between, it could maybe help us out a little more than just showing us plain text hits on that. So let's see, moving ahead. OK, so the way that, just to show you how what I did came together by eating my own dog food, which is a computer programming sort of mantra, learning from what works and using best practices. So the ways came together. And here's an anonymous law student, first year. Maybe this looks familiar. Some of you, you've got the lecture hall. You have the deer in the headlights expression. And so I'm here in class, and I have professors calling out to me, OK, everybody, turn to ORS 163.095. And I'm used to the actual real web way it works now. But I'm back in two screens ago with the tools that we were given for legal research. And so I just realized, well, I'm just going to make something that works for me. And it's a great mantra, because if it works for me, it should work for other people. And especially if I use it on a daily basis, I'm going to want to make it better. Also, as a computer scientist and software engineer, I'm always looking to see people who have really been successful, what and why are they successful? So these organizations are successful for lots of reasons. But these were the takeaway things that I got when I was making OregonLaws.org. The biggest mantra for me was Craigslist, that there's no instructions when you go to Craigslist. You don't have to go to a class to learn how to use it. And every time I found myself wanting to write instructions on a web page, I realized I had done something wrong, and I needed to back up. So I just always had that in my head. Google, they do amazing things. And I hadn't really realized, but a big mantra for them on the back end was, everything should be done in a fifth of a second or less. So I've made my back end architecture similar. Let's see, Apple, the New York Times, and then 37 Signals, they have this idea called opinionated software. They're not trying to solve all the problems. They're trying to do less than the competition. And I also was trying to do less. I'm really just making it work for me. And then finally, these best practices have been mentioned before by the others, focusing on open source tools, using the modern software that we now have in 2010 that lets us really combine the things we've learned over 20 years of information technology. Let's see. So now moving on, where are we? OK, so readability, now into benefits and features of the things that I really focused on. So this was the key, key thing for me. So you can see here in the big picture, this is what you get if you print out a web page from my site. And this is actually being done in an open, standards compliant way, just with CSS. There's no special print thing you have to click and then go through a bunch of menus. So actually, ironically, I get people running the email saying, they need to be able to print. Where's the button to click to print? I'm like, well, it's already there in your browser. And so I'm actually thinking of adding a button just because people are used to the old bad ways, how I see it. But you can see in this version, I have the statutes that the visual representation actually reflects the structure that the legislature wrote. You have your outline form. And then I had a little bit of fun with it. You can actually see, well, at one point, the word and is bolded and underlined. Because I'm learning law and I've learned, OK, there's these elements and you've got to know your elements. And is it like any of the elements to be true, or do they all have to be true? And so and or or is just hugely important. So it's like, oh, so for me as a legal researcher, that just makes this a lot more usable, even if I'm not really noticing that and has been highlighted. It sinks in a little quicker. Then moving on, let's see. In the upper right, one of the things I focused on, and I took a lot of this from looking at the New York Times, how do they embed? How do you embed a link in text so that you can actually still read the text, but you know it's a link? And so I spent a lot of time focusing on, OK, well, these are links. How do we decorate that, as we call it in the programming? In the middle right picture, you can see a little bit of deeper readability added in there. There I did some parsing out. So the legislature turns out they create names for statutes, and then they have a little mini outline of the statute embedded in the name. And they're so nicely structured. So I wasn't working from XML like you all were, but Oregon did a really nice job. Luckily, somehow I ended up in that state, and it just kind of, yeah. And so anyhow, I was able to parse this and break it out like this, which is very nice. And then finally, in the bottom right, even going a little bit further, you can see the benefits of, I mentioned it really quickly before, using these best practices. I created what we call an object model. When I took the statutes, I didn't just cut up the text. I actually created objects in the database. So one statute links to another. And what you see here is that this statute, it links to this other one, 164.395. And now normally in the text, all you see is a number, and you don't know what that is. It just says, if a person violates this, then such and such is the case. But the program knows, because it knows, or this statute, it's an entity. It can find out what's your name. I grab it, and I insert it here. So just again, me as a lost ear, and I'm like, oh, and here, robbering the third degree. I'm already understanding. And this here happened to be from robbering the first degree. Oh, it's basically a predicate. It builds on robbering the third, or the other way around, whichever this happened to be. So you can see, again, eating my own dog food, but really is it useful for me as a legal researcher and student? Another area is the authentication, verification, security, where I focus a lot. So I don't know if I can totally solve the authentication problem, but I'm trying to do the best in my little world that I have, which is giving nice citations and pointers back to the sources that I got my information from. So every single piece of information on my site, and none of it's for me. I don't write anything. Every single piece on there has a citation and links back to the original official source. And again, I'm able to do this, because I've created this object model. Objects, we can store information in them. And then here I'm using a blue book or an all-wood sort of standard citation here. When was this file last accessed? Or when was this URL last accessed? Where is it available at? This kind of thing, which is a lot of fun to do, and not that hard. And then finally, I thought it was really strange that most of the legal research sites I had seen, they weren't actually encrypted. Like when you go and you do online banking, you check your bank account, that's encrypted. No one else should be able to see what you're up to if you're at a coffee shop or even at work. And so it seems like we should have the same kind of thing for legal research. So the entire site is over SSL. Let's see. Oh, and then here. So now this is moving on to something I was totally surprised by, that I didn't expect a benefit to come from this. One thing I implemented with the site was really nice permalinks, we call them, permanent links so that it's a URL, it'll never go out of date. This is sort of a known best practice. But now that I've started creating that, I've seen now new political discourse and new discussions happening online and people able to cite back to my site. To basically support the arguments or just to figure out what is the law. So here in the big picture we have two people talking about bicycle law and traffic law. Sort of arguing about it, what is the law? What should it be? What does a full stop mean on a bicycle? Do you have to take your feet off of it or not? These are big issues in Oregon. I mean, I'm being a little funny, but it's true. I mean, we take it seriously. And then the top right, we have people talking about handgun licensing. So it's wonderful, it's actually across a political spectrum, people understand this is important. And the bottom right, someone just blogging about their personal experiences, nearly getting hit by a car and then they're wondering who is in violation of what statute. So this, I find particularly great and compelling, just that opening up the political process in some interesting way. Let's see. One, another area of benefits that I focused on, which is the whole search, oh no, I'm sorry. I just jumped ahead of myself, is research tools and secondary sources, basically bringing things that, these features that usually you have in the expensive systems, making them available to everyone. So here just basically seeing, here's where I've been doing research before and I want to see that again. In the upper right we have what's called a secondary source in legal research. So I have a mostly automated system of collecting, abstracting and linking to these. And then finally in the lower right, you can see annotations basically, which again the Oregon legislature has collected and written these, annotating the statutes for important cases. And so just putting them in the same area makes it all very useful. The final area I would say that I worked a lot on was searching. And so now just a little bit of semantic searching is about getting some meaning out of it. So again going back to this object model and doing things the right way, the modern way, just because every statute, it knows what chapter it's in and every chapter knows what volume or what kind of topic it's about. When you do a search, you can actually see what sort of topics do they fall under. And as far as your search, how many statutes did you get back under each particular topic? If you click on one of these topics it will filter the search and it will restrict your results all into that area. In the upper right you can see that a little bit I have. I'm able to make links more like this. And so more like this link under a search result is a big deal. You need that kind of semantic information. And again it's all flows from that nice structured data that I got as inputs. And then finally in the lower right something, I'm focusing on there just to show that it's a spelling you type in a, I'm sorry, you make a spelling error as you do your search. And the system knows oh, did you mean this? And then redo your search. This is one of these things that we take for granted these days with Google and Yahoo. So it's nice to add in these, bring the legal research up to people's current expectations. Okay, and that pretty much concludes it. Thanks a lot. Fantastic. Thanks so much Rob and fantastic work on orgunlaws.org. We probably have time for one or two questions. I'm going to open it up to the floor if there are any questions from our panelists. Just in the corner back there. Rob, that's great work. Very cool. I was just wondering were there any technical or legal hurdles along the way? Can you expand on that at all? Yeah, well coinciding right with the time when I got started with this is when Carl was really involved with the legislature. And yeah, at first Oregon didn't want something like this to happen or maybe they wanted it to happen but they didn't want to let go of control over the website. And so luckily yeah, so my work is all predicated on the work that Carl and a whole bunch of other people did. And Oregon now, they believe they have a copyright but they're not going to pursue it. And so people can make what I did. And as far as technical hurdles, now I think the technical parts are the easy parts. I've been programming for a while and no, it was all kind of fun and not too hard actually. If you've got experience doing this kind of thing. Any other questions? Fantastic. Well, please join me in thanking our panelists for this first panel.