George Buckenham, as it says in his Twitter bio, “always liked broken games the best”. He makes games himself, but also works on non-game projects, like his “free and radically accessible platform for writing Twitterbots” over at cheapbotsdonequick.com, gives talks and blogs.
George found time in his busy schedule while traveling to speak with me for the second issue of Botmaker Profiles.
Stefan: Hey, how’s it going?.
George: I’m good! I’m currently at work, prepping things to show our game at GameCity festival later this week.
Stefan: Right, is that the reason for your upcoming travel?
George: Yeah! And then next week off to Zoo Machines, a french festival for digital/physical videogames
Stefan: Nice, well, I hope you’ll have a great time! What game are you going to be showing?
George: Fabulous Beasts.
Stefan: Oh yeah, that one’s very interesting, I will have some questions about the games you make in a little bit.
George: Haha, sure.
Stefan: So I was wondering: what does your username on Twitter mean, @v21?
George: Ah! So… nothing? I’ve had it for a long time though, I use it everywhere, and probably identify with it more than I do my real name? I think I was doodling in school and ended up writing it a bunch? Spent years in school rewriting it and designing it like a logo.
George: And started using it online, and have kinda stuck with it since. The nice thing about online is you get to choose your name, but actually the name you end up with is often kinda accidental or embarassing.
Stefan: Really. But the “v” does stand for “version”, right? I was going to make a joke along the lines it took you 21 attempts to come up with a good username, hah.
George: Nope, not “version”… just v21.
Stefan: You really seem to be interested in video games, or games in general. You made games like Cubes or Hell Is Other People. Is this also how you got involved with the whole botmaking “scene”? Maybe seeing Twitter bots as little games?
George: Ah, yeah, I guess? I mean, I’m “from games” – I make a living making games, and before that I was making games as hobby (and trying to make a living making games). but that doesn’t really feel right in terms of interest. The way I like to think of it is “I’m interested in interactive digital experiences, and videogames are the easiest way to make a living from that”. Not hugely easy, but easier. There’s no real dividing line, too.
Stefan: Yeah, I definitely see where you’re coming from here. I mean I am a web developer, but I am not all that into “coding”, to be honest. Yeah, sometimes I write a nice closure that handles some async calls in a very elegant way. But I much more enjoy creating, well, as cheesy as it sounds, experience. That’s what I like about making bots, for example. You don’t have to worry about how your site looks like on mobile. You just focus on the interactions your “work” has with other people.
George: Hah – but it’s kinda easier to write nice code than worry about the experience, and actually think hard about how it’ll be experienced, and do the whole artistic thing.
George: And — you still have to worry about how it looks on mobile, sometimes. Like, the magical world of extended unicode characters, and where they render, and where they don’t, which I learned the hard way from @unicode_garden. Or, to scope out a bit – there’s a nice thing about making an experience, and knowing it’ll be rendered differently by different things. but you still have to design for that second level experience. Did you see Robert Yang’s recent post about designing the context for your work?
Stefan: Hmm, I don’t think I did.
George: It’s really good, packages up a load of thoughts I’ve had in a more disconnected fashion
Stefan: Thanks, I will definitely check this out!
George: Like, bots are more accessible to experience directly, but even then – if you write a good blog post, or make your own screenshots, that’s how it’ll be used, that’s how it’ll be seen. Or with bots, there’s the thought of designing the experience for looking at the twitter.com/whatever page, and for the experience of having it mixed in with your feed.
George: Or Tumblr especially, (this comes to mind because I made a tumblr bot this weekend), where there’s many different formats the blog can be presented in.
George: Or there was some chat the other day on the #botALLY Slack about “depth” – building an output procedurally that has an interesting texture – some common stuff, some rare stuff, etc – so the first experience with a bot is good, and then living with it provides new things, or a sense of anticipation for a thing it will produce, or excitement when it does the rare exciting thing – again, with @unicode_garden – there’s a really small chance your garden will have a snail in it.
Stefan: Yeah, you have some good points there. And I’m also thinking, interestingly, you can easily lose control. I mean, maybe your bot relies on the inner workings of Twitter, like @pngbot.
Stefan: Or they may change the visual presentation. There is quite a lot to consider when making a “simple bot”.
George: Hah, yeah – and also you can choose to relinquish control over parts of it. but that’s also a choice.
Stefan: Going back to games you made, it’s interesting to see you don’t only stick with the same medium and make games like Punch The Custard or Fabulous Beasts. What are some of the reasons you might prefer a video game over a “live” game and vice versa?
George: Hm hm. Well, generally – normal videogames can be sold much more easily, they can be duplicated endlessly, they can be seen by thousands without you having to spend days running the game. We’re trying to break out of that with Fabulous Beasts, of course. Making a physical game that can be sold, can actually reach people in their homes… it’s an exciting thing to be doing. But the live experience – you do get to control more of the experience. You get to see people actually having fun. You get to do more stuff! You don’t just have to design for the hardware people already have. But that hardware is pretty capable, and pretty flexible (thanks, general purpose computing) – there’s still loads to explore.
George: I mean, the other point of comparison is the work I’ve done putting on events. Which is a different experience again – even less control, showing other people’s games, but similar satisfactions to live games. Like, it’s fun to run around stressing about real world problems, rather than just type into a computer all day.
Stefan: Ha, this reminds me of that one web comic comparing the way of live in the past and now.
George: I mean, sure. The other viewpoint is the sense of wonder I felt as a kid as computers got more capable. It was like… everything I need, I can do with this one thing. And all the things that fit between, it can do those too. Now, it’s like that, except…it’s your phone? It’s a thing you can carry with you everywhere you go? The moment of wonder when I could download a song off Napster in less time than it took to play… or when I got maps working on my nokia, and could see where I was in realtime as I took the bus (and so saw I was at my stop without having to look out the window)
Stefan: Yeah, it kind of creeps up on you. You like all the cool stuff you can do and then you end up looking like the guy on the right, in that comic.
George: Yeah – I did a theatre training thing, and we were sitting in public, watching people go by – and it was like 1 in 5 people interacting with their phones as they went past in at least some way.
Stefan: So, is there a “game maker” you might call your inspiration, someone whose work you really enjoy or perhaps even emulate?
George: I guess lots of people fit into that. like, for me a big driver behind my creative work is getting to work, exist on a level with people I admire. people I try to emulate…Hm. With bot work I have been incredibly inspired by Katie Rose Pipkin’s work
Stefan: Yeah, she has done some really interesting stuff.
George: Like, her attitude is great, but I think the main thing I take is this really defined aesthetic. it folds in so much of retro, “authentic” computing – this honesty, this sense of how the machine really is. But it’s also clean. And it’s also so clearly about a real joy in beauty. Somehow finding no contradiction between a willingness to let go, and a careful precision.
George: Ugh, it’s so great.
Stefan: What is the best advice you ever received on how to make a good game?
George: Hm! I’m not sure. Definitely not sure as a useful lump of advice I’ve received. Mainly it comes down to paying attention and trying things. I once wrote out a whole load of predigested advice for game making : most of it still holds up, and most of it I learnt from other people.
Stefan: So what’s the story behind Cheap Bots, Done Quick? How did you come up with the idea? Are you still actively working on the site?
George: So. I’d seen Kate Compton‘s Tracery a while ago, and kept returning to it as a great way of making kinda messy procedural logic simple to edit and user friendly. But I found most people I showed it to weren’t impressed – they were either programmers and had their own ways of doing things, or they weren’t and weren’t interested in procedural stuff. I was into bots, generally – I’d made some of my own, and generally was/am very enthusiastic and excited about the scene.
George: So then I ended up pitching doing a bot workshop to the games festival Feral Vector. I tried to show friends how to setup their own ebooks bots, and that worked, but it took a day, and it was just this impenetrable thicket of programming bullshit to wade through. The programming bit was easy, but dealing with OAuth, setting up long-running processes, etc – ugh. So I still wanted to give this workshop, so I started working on making the process accessible to people who didn’t come from a background of being used to dealing with that bullshit, and who didn’t have half of this stuff setup already. So, first I made “I Will Dance The OAuth Dance For You” – which is a useful tool itself. I wanted to give people something more interesting that an _ebooks account at the end – it’s fine to just set that up, but where do you go from there? So I hit on the idea of teaching people to make a Tracery-based bot. I think the first version I made was @cant_miss_it, running off github.com/v21/tracerybot. But even that still needs a server, and there’s no good free hosting any more since Heroku tightened up the free tier. And it’s pretty computationally inexpensive, and I know how to make a PHP based thing that talks to Twitter, so… That’s where CBDQ came from. When it came time to give the workshop, I basically just opened up CBDQ and said “go at it”, and that basically worked as a workshop.
George: TL;DR : I over-prepared for a workshop.
Stefan: Haha, well I think this is a really good story. Thanks!
Stefan: I started working on this series, that promises to take you from a non-coder bot enthusiast to a “pro botmaker”. I reached out to you a few days ago on Twitter, asking whether CBDQ is open source (which it is, which is great).
Stefan: You’ll remember me mentioning that I’d like to add third-party API, easy public dataset integration, and so on. Basically helping people who like making bots stay non-coders. Now, after that brief conversation, and seeing online tools like Dexter, I actually like this approach.
Stefan: I was always against the idea that literally everyone must learn to code, but I do think that people should understand how the various technology around them works and why it does the things it does.
Stefan: And that we need tools that make learning about and interacting with technology easier.
Stefan: What do you think about all this?
George: Yeah – so one thing I’ve been a bit disappointed about is that no-one to my knowledge has gone the route of CBDQ bot to self-hosted tracerybot to different functionality. I mean, just disappointed in that I put some effort into making that a viable upgrade path and it wasn’t worthwhile.
George: But yeah – I’m torn – on the one hand, I want to see people doing interesting stuff, and making weird and new and unique stuff, and that kinda requires coding. on the other hand, I totally don’t think coding is superior to other stuff, and am pretty sceptical at this “learn to code” rhetoric
Stefan: Yeah, I mean, the way I see it, we should be doing two things: Some people, who are into it, should learn to program. But they should also make tools that make it easy to create using computers for those who are not interested in writing code.
George: Yeah – but it’s more interesting than that. The tools we use shape what we make.
George: When writing CBDQ, I kinda deliberately added grain into the tool. I was focused on making your first bot the easiest it could be, so I took away options you otherwise could have. There’s a bias towards writing in the editor, rather than uploading a file – this means bots are shorter, more handwritten. The bots all post at the same time, and you don’t know when that is. If someone else had set out to make CBDQ, they would have made it very differently, and I’m sure that’d influence the bots that are produced. I mean – most people who use it weren’t looking to make the thing they made. They were looking to make a bot, and then they did the thing that made sense with the tool they were provided. Again, it comes into this thing of second-order design, of deciding where you have control and where you relinquish it. Making the tool open-ended, but also containing enough prompts to give people a grain to work with or against. For example, @nyxtaki generated some labyrinths and then just pasted them into CBDQ. With some randomization using the Tracery rules. So it’s not just allowing people who can’t code to still do a subset of the things coders can do.
George: Or (and this isn’t an attack, as I say similar things) the idea of “advanced” twitter bots. Like, who’s to say that the code lying behind a bot being larger and more intricate makes it worth more? If it’s devastatingly written, that can be a lot more effective than clever code. I’d like people to be able to explore more space than I’ve broken with CBDQ, though. I want to see all the corners of this possibility space, and I’ve just helped make one corner more accessible. hopefully people will start at CBDQ and end up making their own tools, their own techniques, their own styles and aesthetics, and i’ll get to see all this cool new stuff and boggle at it.
George: (I’ve written a bit more about this at v21.io/blog/tools-for-novices)
Stefan: Yeah, I mean personally, I can’t even really formulate what makes me like a bot. It goes way beyond how complex it is, whether we’re talking about the codebase or the idea.
Stefan: There’s a bot — and I wish I could find it — I think it used a picture of Dick Dastardly as its avatar.
Stefan: And all it did was it posted various versions of the “evil laugh”, like Mwahaha, and Muhuaha. I think the bot was called something like “Evil Laugh Bot”?
Stefan: And I don’t know, I think it’s the same as with laughter being infectious, but it just made me giggle.
George: Haha, that’s great.
George: Perfect concept.
Stefan: And I’ve seen other bots that are about as simple and it just wouldn’t do it for me. And I’ve seen bots that I know took work to create. But that’s not always enough either.
George: Yeah – like, I really like @metaphormagnet, but that’s just the CS grad in me. It doesn’t use all that cleverness to as good an effect as it could do, if it was written better. But that’s also fine – that’s basically the role of academia, right? To break new ground, to find new techniques, to make that possible. then other people come along and actually use them in anger, ignore the interesting theoretical bits and just focus on making them effective.
Stefan: Well, we’ve been at this for a while and I’m sure you have things attend to. It’s been a real pleasure speaking with you!
Stefan: Good luck at the GameCity festival!
This is an interview from the Botwiki Interview series. Read more.