 Right, this is what we're talking about. It's digitally generated, kind of random. Let's take a chance, try another one. If it says error, then I still have a job because I'm not a digital artist or that's never how I thought of myself. I'm an IT guy. But Sydney runs a very, very small digital humanities operation so the IT guy does the crawling under the desk to make things work and everything else, including turning myself into a digital artist, which you can probably tell by my liking for grey. Even my hair has finally got with the program. I'm rather an unlikely digital artist. But here we go. And to be honest, I actually find quite a lot of the stuff that this strange program that I've been working on really, really beautiful. But I know something. I didn't put the beauty in there. So the question for me has always been, well, where did it come from? But let's just cover off where we started. So we started getting ready for a very small exhibition in quite a small little gallery in the Alexander Turnbull Library. Thanks very much. I am a lot more comfortable. So this, what you're seeing here, is just one of the things that Charlotte Darling, who was the research assistant on the project, did by way of getting ready for the presentation. So she was rummaging around through the Alexander Turnbull Library's collections primarily but also to Papas to put together an exhibition. The idea was that we were going to actually try and cram this. I shouldn't use my language. But what we were looking for was a winda kama. I'll use Sydney's language. So winda kama, it's this kind of idea. This is Neil Partington. This is the attic. This is Wellington Museum, just around the waterfront, up a few stairs. They pop their top floor full of stuff. And if you've never been there, I can't put in a plug for an exhibition at the Alexander Turnbull Library because it's finished already. But you could go up there and see their exhibition. It's a winda kama. It's this idea that you're a collector. You collect a whole lot of stuff. And then, of course, you want to put it all on display so you just shove it all in there. And these collectors, these people were, in many respects, the start of museums. Here we go. This is one of Sydney's favourite museums. I've never been here. This is Pitt River Museum. Look at the organisation. Look at how it's organised. I've just picked this one because I saw Captain Cook. But this is a kind of a thematic exhibition and actually led to a bit of a change. This is one of the museums that's cited in anthropology in different ways of studying cross-cultural artefacts. So we were going to build a winda kama. But just before I leave, let's dive in there. I've lost Captain Cook. Where are you there? So there we go. That's a winda kama. That's a whole building full of cabinets, winda kama cabinets. So the theory then is that the way the theory works is that we kind of had this ant type of person that would go out and collect, collect, collect, collect. And the ants would then organise their collections. But as you can see in this dialogue here, there's another way of looking at it, which is that it's a spider. This is a fictional dialogue between an ant and a spider. And the spider has this sort of relations and lays things out too. But the critical thing to understand here is that the web is just the way that enables the spider to move. That with the web, if you capture the web, you don't actually capture the person. So coming back to what we were doing, we were trying to do an exhibition about William Colenso. And this, I think, is the closest thing that I could find to Sydney's conception of what we were trying to do. That we were trying to, this is called a mesh work, that we were trying to do a thematic organisation of the life of William Colenso. And I need to just ask you, if any of you know what William Colenso was doing between 1852 and 1856, that is one of these big white spaces here. You see he got tossed out of the church in 1852, and he came back as a politician in 1856. And it's very little known. So if you think of this as all of the different ways of tracking William Colenso's progress, then there's a big vacuum. And Sydney's interested in that, theoretically, and so on. So what she did, this is where it gets really weird, is she went to Mitchell Whitelaw, who's a real digital artist, and he built this thing, a tool called succession. And it generates things by pulling down images about Newcastle on Tyne. So that's not, for me, a particularly evocative one, but some of them are really beginning of the industrial revolution sort of types of creations. That's more like it. And you can drill into that from the pics there. So she said, Rhys, localise that. Bring it, repurpose it to be the digital component of the exhibition on Colenso. So we started. Now I'm going to start showing you what I was working on and talking about the process. So the computer side of it starts with these numbers here. Those numbers happen to correspond to a bunch of sets. Most of them mine, but Charlotte's one, and also one that's been around for a while that was done by MTG and Hawkes Bay. So I started with those two, and at some stage I got a chance to talk to Mitchell, and he said that he'd spent three months, I think it was, where he spent half his time writing code and developing it and changing it, and half of his time curating the images that his work drew on in order to achieve the aesthetic that he wanted. Now there are some computer scientists out there, presumably, who are quite happy with the word aesthetic, but for most of us that's a little bit scary, and the idea that no-one's going to give you a specification for what you're producing. They're going to expect that you go backwards and forwards between your data, what you mean you haven't got clean data. No, you select the data, you write the program, and out of that going round and round on what's in your head, you get to your goal point. Bit scary, bit scary, but I pressed on. We pressed on. So next step is this. I had some tools that I needed for that horrid angular that I was working with and lots of XML, working with Digital New Zealand API, and some bits and pieces there that you will recognise if you're very long-sighted. So we won't dwell on that. This is half of it. This is the back room. This is what happens in what passes for a server environment. This is a shell script. You definitely need grey hair for this. I wanted to point that thing out up there. There's a strange string. That's the Digital New Zealand API. That's my personal API. One of the things that we bumped into was that there were some things that we could do online in an app, and anything that involved the API couldn't be done. We couldn't give that away in a wear-back. So that's up there. Here we go. This is what the back calculation produces. It produces a set of effectively Digital New Zealand APIs, but it's been filtered. Firstly, on whether we're allowed to remix it, as a consequence of that, the digital exhibition is not located in the Turnbull. It's probably closest to Papa, but there are only two or three things that come out of the Turnbull, and those are the things that they put into pilot releases to Flickr Commons and other places like that. Most of the Turnbull stuff, they haven't sorted out copyright, and getting entangled in trying to sort out copyright for 100, 200, whatever that number was was quite beyond the scope of what we were doing. So a bit of filtering, and some of these are a bit strange. Some of the resources that Digital New Zealand classed as images were wear-backs or actual HTML pages, and there was a bit of wear-scraping involved in getting a straight image out of what was supplied up by Digital New Zealand. That's the back end, the front end. Well, you've seen the front end. It's a wear-back application. It's delivered... I don't know if you can see it up there, but it's delivered by white-out-of-press.gitlab.io slash unexpected. So it comes off a pipeline out of Git, that wear-back, and the wear-back, of course, has all the angular in it and the code, and I'm not going to drag you through this, but I wanted you to trust me that word random in there, and that's the essential thing that's going on. So we're randomly selecting out of those resources that were kind of curated by me, mostly, and then we're performing random operations on them. The operations are happening in your browser and come from out of the operations that are provided by browsers through a set of operations called blending that happen on canvases, which are the viewing surfaces inside your browser. OK, off the end of it. Off the end of it, we get unexpected. While I just figure out a conclusion here, let's just try our luck. I want to mention just one little... well, the best thing about this, if you don't like it, you just push the button again. And then you have this marvellous thing. What's happening in my head? Why don't I like that? Why do I like that? How did that happen? But coming back to Sydney's purpose, the thing is really to provide a new way of looking at those collections. Now, to be honest, I think most of the people who came into the little gallery struggled with what to do with the iPad that had been kind of nailed on the wall, and we didn't get the interaction right. There were actually too many totally unexpected, for me, issues in dealing with how to get people to interact with a thing on the wall. I spent all my time trying to harden it so that it would actually survive to really manage to do much about how we interact with it. Really, really challenging issues. I shouldn't be loose on any type of computing appliance that pretends to be a kiosk because I can make them go wherever I want. There are a few individuals who visited the Alexander Turnwall Libraries Gallery during that time that it was on who also know how to make iPads go wherever they want, and sometimes they took them to some unsafe places, really. Scary. So, where are we going from now? Where are we headed? It was a tiny job that I never got around to which was coming back and digging out a piece of work that was done for NDF in 2012. A similar way of providing a kind of random, thought-provoking way of looking at a collection. Let's see if this still goes. I'll just hit the Make the Magic button. There we go, floods. Oh, I want images. OK. Maybe now we can do a little better with that out of what we've learned. But what Sydney wants to do is actually take some of those images that unexpected was generating and take them back to her letterpress and her other lithographs and her other technologies and kind of, how do you say it, sort of back to the future. So we've kind of gone to the future with this digital stuff. She wants to take it back and have the old machines turn out some of this new stuff. Again, just as totally different ways of interacting with the collections. Right. I got to the end. Thank you for listening, by the way. Thank you. That was a beautiful journey. Is this one back on? Thank you. Yeah, that was a beautiful journey that you took us on there. Just before I throw questions to the floor, I have one myself. You're talking about Sydney wanting to take some of those images and turn them into physical images again. When people are generating images through the web app, is there a way for you to save or export the things you create? Great. Okay. Thanks for that question. I completely forgot one of my tabs. Two months ago, I filed an issue in my GitLab. Allow a composition's URL to be saved and have it regenerated from that URL. I should have been able to do that in five minutes. It's a relatively straightforward refactoring of that Angular code. But in the meantime, Angular's moved on. It's a whole new version, a completely different tool chain. But the plan is that that's going to give it. So if you like, the random pushing the button is going to generate a URL, and you can bookmark that, and whenever you want, you can just bring that URL back to life in your browser, and the browser will recreate. So we'll put all the random bit into the generation of kind of like a code and then the browser can just rebuild that whenever you like. As you can see, I started... I've made the issue two months ago. I've created a branch. I've done lots of work. It's pretty pathetic at the moment, but it's coming. It was supposed to be ready today, but... Is there any plan to allow people to choose a few images to then generate the artwork? Is it always going to be random? With the caveats that... Well, maybe it's time that we actually talk to the digital New Zealand people. This would be a strange but wonderful way of doing something new with the stories that people generate on digital New Zealand. So that would be over to them, and of course, I think they'd be most welcome to do that. Halfway houses are a little bit tricky because the press doesn't really want to be in that particular thing, but there are certainly possibilities. If Mitchell's okay with releasing that code, I think it's probably time we had that conversation. Sydney had that conversation with Mitchell, and if he's okay with that, then we'll just be making the whole GitLab project minus my API key available. So I hope there's progress on that one way or another. Well, join me in thanking Rhys again for this presentation. Thank you.