5 June 2024
Mark Anderson: While a client who. Who worked under a bookbindery, of which only two left in this country. It’s amazing. You go up there and the tools were hundreds of years old. Well, you know, there was nothing.
Dene Grigar: Those two books side by side.
Mark Anderson: Yeah.
Dene Grigar: And it’s just the most. And students are like, oh my gosh, I had no idea that, you know, when, when books were invented.
Mark Anderson: Oh, sorry. This is the one that Deanie mentioned that she taught. And I was just saying from my skim reading so far, this just arrived today. It reminds it, I think it pairs really nicely with the Houston book. The book, the book which we know of. So yes, I shall look forward to reading that.
Dene Grigar: And this is the other one I would teach with it after Babel by Steinberg. Okay. That’s another one that the aspects of language and translation. Yeah. So I taught I would teach that these three and a couple more and and then go into multimedia stuff. So I teach the traditional and then jump into where we are today, which was, you know, starting back in the 90s. With the writing space. I would teach the writing space book by bolter.
Mark Anderson: Yeah.
Dene Grigar: So all of that together? Landells hypertext 1.0. It made for a great graduate seminar.
Mark Anderson: Yeah. I mean, I as as working on sort of paper at the moment I’m, I’m sort of thinking about of course, there are all these notions that come across from sort of literary theory about things like sort of death of the author, but it’s rather pertinent to some of the stuff we’re discussing in a different way, in that I’m very conscious now in terms of thinking about digital authoring tools, that one of the things I’ve got to let go of is some of my, some of my presumed rights as the author. In other words, I can show, I can indicate to people what I would like them to see, but I would be foolish to think that’s necessarily how they want to consume it all the time. So in other words, what I’m giving them is a serving suggestion rather than the thing that says, here you go, this is what you get, and this is all you get, which is what a book gives us. And not that that’s done with any ill intent, but that was, you know, it was a limitation of the technology we just don’t have now.
Dene Grigar: Hey so, Mark, the other question is, are you talking about tinderbox today?
Mark Anderson: Yes. But I well, I’ll make a preamble. I’m, I’m conscious that I don’t wish to appear to be sort of shilling for it anyway. It’s really I’m just using it as the best thing I have to hand to explain a couple of things that I think are pertinent to some of the things we’re discussing. And of course, you and Andrew will both see some familiarity with storage space because tinderbox is essentially storage spaces. Younger sister.
Dene Grigar: Can I tell you what store was tinderbox is. Yeah. Michael Joyce licensed storage space to Mark Bernstein. This is a story from from. Yeah. Michael, he he licensed it to Bernstein. Bernstein was supposed to pay him money for the license over time. But what Mark did is he pulled out parts of story space he liked, I mean, and turned it into tinderbox. And he quit using storage space, essentially. Yes.
Mark Anderson: Yeah. And also, he wanted to at the time, we sort of forget that there was a time when blogging was considered something interesting and fascinating. So the other thing that fed into that and was never in storage space, for instance. So in the interim, will the weather come along? So one of the things that tinderbox also did was it was thinking about note taking not so much for creative writing or analysis because I know, you know, story space doesn’t have to be just for, for creative purposes. But he was basically at the time thinking about note taking and blogging personal, you know, sort of static page blogging. And indeed, to this day, its output is is static HTML. But you can do all sorts of clever things, and I’ve done that along the way. But that’s what’s there. What’s really interesting about it is it is also the linear descendant of there seem to have been interesting parallel evolution of spatial hypertext. There’s not much talked about, but tinderbox and story space. Well, tinderbox feeds off story space and story space goes back to the mid 80s. I think they first started working on it, things like that. Yeah. And and it just had this notion of fluid spaces in which to, to work and think. And I and at the same time there was the work of, well, there was, there was the browser view and note cards, which basically had some influence on the work that became Aqua Net. But Aqua Net was trying to do argumentation in a visual space, which led to the idea of Mark, we’re all here.
Frode Hegland: We have to stick to schedule. Today we just have one hour. So history of spatial hypertext is definitely cool. And it’s very, very relieving for me, so to speak. And comfortable for me to say that Fabian and Leon are finally back in their boxes. Sitting on the.
Mark Anderson: Sofa.
Frode Hegland: We have an agenda for today. Believe it. It should have been shared just now. It’s a very obvious, simple one. And hang on, I’m just going to do the public link thing. Get a link, get a link copy. The boom. Here we go. So I was going to briefly do under announcements, a little bit of author reader, but there wasn’t actually anything I thought there would be. So the time we’ve used an introduction can be written up to that. Any other announcements? I have one small one. We’re going to go into mapping thought by Mr. Mark in a second. But Andrew, I’m trying very hard not to swear, but holy moly, what you have done with the latest version. Right. It’s nice to see the space and it allows us to think in a new way. I just want to briefly highlight a few things. Number one, it really shows that there is an up and down in XR in the sense that there is an up that is too far away, meaning it might be useful to store things there because I was thinking maybe spherical. And also the floor might have value. But what you’ve done with the pointer mechanism where the dot appears just when you want it, it feels not like a quote unquote horrible laser pointer.
Frode Hegland: It feels more like a magical Spider-Man power. It is really extraordinary. I was in the military, you know, so I know how to use weapons of different kinds. And when you aim a weapon, sometimes you lift it up and aim. Sometimes you hold it and you know where you’re going to shoot. That’s what this has become. Now, when I point with my hand just in space, it’s I know we’re not trying to be violent, of course, but it’s it is like a gun. I know where I’m going to hit. And if I don’t, it doesn’t matter, because the dot appears at the right moment to let me hit it. So I’m really grateful. It’s absolutely fantastic. So I look forward to Mark now and the rest of the time, because you only have an hour today, right? And then we’re going to to go through that, and then we’re just going to all flutter away in happiness. Right. Anyway, over to Mr. Anderson and Mark and Andrew again. Just wow. Thank you.
Mark Anderson: Yep. Just let me find wherever my screens go. Right to find the right one. Share? Share. That one. Move that there. Right. Okay. You should be seeing a pretty blank screen. Yeah. Yeah. Yes. Okay. Right. So I move that off the side for a second. So a couple of a quick just a quick recap for a couple of you who weren’t here. So I’m probably showing this to try and move us forward. Some of our thinking that that sort of got slightly tied up when we were talking about maps and what a map is and what it, what should be on it and things. I’m using a tool called tinderbox. Don’t don’t get caught up in that. It happens to be a desktop 2D tool. Mac only. But that’s by the by, it’s quite venerable. It’s 20 years old, but it goes back to it’s got it’s got roots in the sort of hypertext knowledge world going back to 19 mid 1980s. So this isn’t some new toy that’s just pulled off the shelf. And I’m partly using it because this has been part of my daily work practice for the last 20 years. It’s my primary work tool as a sort of thinking and writing space and sort of just working space. So I’m going to jump in straight away with a provocation by saying, this is a map, because and I say this because, sorry, let me move, let me move the people to whom I’m speaking in front of me, then I’m not looking at you sideways.
Mark Anderson: The reason I say that is we had this sort of this conversation about what must be on the map, and I was thinking, I’m not sure that’s the right question to ask because this is the map. So we put onto it what we want. This is this is a map. This is also a map. This is a map with something on it. This is a map that has. You know. This is a map and something that has text on it. My point being that I think we need to to stop being too literal in our thinking, in terms of think just describing what we think we might see and think about what we see represents. So I do see, and it’s something I had a sort of DM chat back a few days with, with Andrew, because I was trying to think it through from his end of the problem is and one of them is, well, do I create a map and put things in it or do I? Well, this space we’re calling a map, do I create a sort of an area, a presentation called a map, or do I or do I do I take some things and say, make a map of this? From my perspective, I don’t think it really matters.
Mark Anderson: And I think it speaks to how you enter that mode of use. In other words, it also says that it’s probably not useful to think of it in terms of a map having to have things on it. That to me comes much, much later where you’re showing mature work. And in a sense you’re using it more of as a, as a, as a pre set up space. That’s like a saved view. But I think to do that too early in the process pollutes the process. The point, the point of all this is of this, this, this lightness is that in fact, the design here is everything, in fact, is called a notice in tinderbox. But that’s by the by, that’s just its term. It could be a node, it could be a space, it could be whatever you want to call it. It’s just a point in the map. And each of these things you see in a tinderbox map is just a collection of attributes, field values. So you could think of this as a little glob of JSON if you want it to be. You know, today’s probably that’s what we do. We’ve got this little bit of this little bit of JSON. And what we see, importantly, what we see is merely what we choose it to be if we want it to be you know, if we want it to be a different color, it can be we might want to make it a different shape.
Mark Anderson: We might want to put you know, we might want to put some some, some sort of annotation on it or make it just so you can see that. But but the point is, these are incidental. So in other words. It serves you, not you serve it. And one of the problems with a lot of the systems around at the moment is they fall into the trap of thinking, I take some information, I stick it in a force directed graph. And this is knowledge. Well, I’d say that’s a lot of hooey. And, and the interesting transition I’ve seen in sort of 20 years of supporting people using this sort of space is that there is a tendency to people to go in text heavy first, because understandably, there’s there’s their supposition is that if they can’t see the text broadly, it isn’t there. And that’s I don’t mean to sound trite in saying that. It’s quite understandable. And this is the thing I’m interested in. I can’t see it. Therefore. Well, I can’t interact with it. Well, no, I mean, this could this could also contain a 10,000 word essay if it wanted to or be linked to it. You know, now we have the, the, the affordance to do so vice versa, you know, the web of things and, and that sort of networked thinking.
Mark Anderson: So I think this makes it much more flexible in terms of what a map space can do. And really it’s an exploration space as well as As, as as well as as that aspect there is the thing I’ve mentioned before of you know less is more. So this, this is, this is actually me doing a deconstruction of a Pentagon brief that turned out to be really weird. It’s about it’s actually about the US anti-insurgency in Afghan. It turns out there are only two inputs, weather and money. And there are actually zero outputs. Let’s just show the insanity of modern life. But also it’s incredibly busy. And and, you know, there are forces built in here. I can do some animations and stuff, but but actually it it can be more useful if I sometimes. Tensorflow. Excuse me. I just the things I’m doing here are things that I would envisage in our modern space. I wouldn’t be doing through a menu like this. Suffice it to say we’d have some other interaction, but I’m just trying to. I’m just trying to make the point that now we’re already seeing a a much less busy. A much less busy space. Which goes to the point of the problem of lines creating unintentional salience when you’re trying to investigate things. So sometimes just a spatial relationship is good.
Mark Anderson: And stepping back from and making this point also about text. Again, if I show go to No, don’t try and read anything on here. That’s not the point. It’s intentionally small. The aim here is what I hope you can see is there is a yellow things and blue things and dark green things. So the yellow things are must be in some way different to say the dark red ones. Which was essentially what I was doing. And this was me, actually, during my master’s mapping out the sort of ecosystem that OpenStreetMap lives in in terms of technology and social, social contexts and things. So that’s why OpenStreetMap is is sort of highly linked, but there are quite a lot of things here that aren’t linked. And that’s fine, because essentially I’d make a note. Oh, here’s another thing. I put it in my exploration space. I put it close to the things that I thought it had an affinity to, and I was immediately sense making. And this map hasn’t changed apart from me tinkering around with the show. This hasn’t changed since 2014 when I made it. That’s all it had to do. And it makes the other point about these exploration spaces is this is not graphic design. I think this is the other mistake. And I watch people sort of, you know, lining everything up.
Mark Anderson: And that’s fine. There are tools in this if you if you want to play that sort of game and have all the boxes the same size, you can do that. But I think the value of these spaces is absolutely antithesis of this. It’s not about making it for graphics. You can do that. But I think that’s an entirely separate activity that can be done with a map like this, but it takes you to a completely different space. It’s also the case here, I mean, and the way that, for instance, all these all these different colors are done is, is quite literally I. Can I make a note? And by choosing their by. Something’s eating my notes. Let me do it. Let me do it on another. Let me do it in another. Another file doesn’t really matter. So if I Have another note which I will make into a prototype. Oops. And I’ll make it. Just to make the point. So in any note, I can pick a prototype. That note now inherits all all the custom properties of that object. And this this goes back to my earlier point about thinking that when we. So when we take our data, in our case, maybe it’s all the data about a particular paper we’re looking at and put it into a space like this. We don’t need to visualize all of it. We don’t need we don’t have to make it a white rectangle because it’s a paper.
Mark Anderson: It can be something very abstract, and indeed it doesn’t necessarily need to show anything textual. What it needs to show is what’s pertinent to our problem of understanding at the moment. But none of the other data is lost. It’s all there either in the either in the object itself or accessible via dereferencing. And one of the things I’m very conscious of, and one of the reasons I’m sort of talking and thinking about, about data is I’m mindful of those who are coding it. So some of you here today it’s easy for me in a glib way to say, well, it’s all just data, you know, it’s all it’s all in the metadata. That’s fine, but it’s not. It’s never quite that simple. So some stuff has to be close. Other stuff can be be pulled across. And part of the challenge is actually seeing the granularity of that. It’s one thing to posit that the data can be there. It’s quite another to make that available in a tractable sense, so that the thing you’re working with at the moment can do what you want, and so we can play the tunes. Another point of the another point of the prototyping is essentially inheritance, so that if I. I’ll just make another note and I’m going to make this one.
Mark Anderson: Black notes, and I will also set his prototype. But if I now go and change this note. Sorry, I’m just going off nearby. You’ll notice that although these both inherit from this prototype, only this one takes the color. Because, in fact, I’ve set a property at a more local scope. I’m sure those of you who do programing and those who don’t are probably mostly familiar with this. So forgive me for just going over this, but again, I think it’s very pertinent to our thinking about how we how we explore in space, because my observation of people using maps as this type of, of this type of observation is this sort of incremental formalization is a very strong part of it. So enabling this and thinking about doing things that don’t disable incremental formalization and this sort of inheritance are will will pay us well in the long term. And I think one of the strengths over the 20 years, I’ve used the sort of stuff I’m showing is this lightness of touch, of being able to essentially recast a whole lot of information in sort of slightly different clothing, almost in the sense that the touch of a button is. Is something I got so used to and in most other tools I use is actually a remarkable amount of effort. It’s not that it can’t be done, but it can’t be done easily.
Mark Anderson: And, and the only reason is because the original design of the tool being used was slightly different. So and so, so there’s, there’s no kind of, there’s no critique in that. It’s more or less a, it’s a reminder to self that be careful what you do because it has consequences in the design terms. And that’s why I probably may come across as overly cautious in terms of saying to those of you, the coding the stuff is well, is is the idea that I’m expanding to you? Does it make sense? It’s not. Can it be built but more? Is it sensible to build it in the way I’ve expressed it? Because my my experience along the way has been more often than not, the answer is yes, it’s possible, but that’s not the clever way to build it. And then someone will explain. There’s a much better way that that is, is that works better with the sort of the technology and, and all sorts of things that I didn’t take into account that still gets us to the same happy place. I think I can show you one more thing while I’m in here, because it’s a slight tangent, but then it might be interesting to people this this hot off the press. This is literally in late beta. But this is using something a bit like a Voronoi space to look at affinity between nodes.
Mark Anderson: Now, I can see this probably wouldn’t necessarily work in a three dimensional context, but possibly essentially what you’re trying to do is you’re trying to cluster birds of a feather. And it basically just works. It just actually works of a query. So what this view is saying is I want to be closer to all the people who have all the information similar that I hold and tag information that I hold, and then it does clever maths that I don’t begin to understand and it goes together. But I could ask a different question. And of course. If we take the metadata sort of concept, that could be anything under the sun. I could have put all sorts of information in here behind these same notes and these same notes. Suffice it to say, are just tinderbox notes. So what you’re looking at here also are view spaces, view specs. As I’m as I wish we would start. I mean, I think it’s a really useful term to bring back, because one of the things that I’d forgotten after using a tool like this for so long is that I’m always surprised when I have to explain to people who are beginning to use it that it’s the same information, it’s all the same information just shown in a different thing, and that I’m surprised that how surprised people find that hand up. Ill. Yeah.
Frode Hegland: We’re running low on time. Do you have time for your questions?
Mark Anderson: No, absolutely. I’m done. Essentially. So far away? Yeah.
Dene Grigar: The more I, you know, I’ve been working with storage space since 91. Right. So a long time. And I also know Storage Space three. Right. I’ve been using that two since Mark has been selling it and I’ve been teaching it. So what is the difference? I’m looking at this and I’m not seeing, with the exception of being able to manipulate the shapes. And do some interesting things with the way the. The boxes looked in the past. There’s not much difference in the two different software programs. Am I wrong about that? They look familiar.
Mark Anderson: So here the shape is incidental. I mean, I can I can sort of change. So it’s all experimental and this, this as I and I think this is I think what you call a Voronoi view. So the point here is you make polygons for the most efficient packing. So the it so that’s one of the derivations. So this is this is slightly different to a map. But but all it’s trying to do.
Dene Grigar: Intellectually is the same right.
Mark Anderson: Yes. But it’s here’s the really interesting thing about it is that, I mean, I see this from watching the sort of development of people who start using tools like this as you go. It’s actually learning that, that, that the quality of the output is the quality of the question you ask. And the quality of the question you ask is slightly constrained by the, the, the effort you put into the metadata that you put into your notes. So people start by thinking they just they’re trying to mine text, but then they realize the text is ambiguous. And you get you get, you know you get semantic collision between word meanings and things. Once you begin, once you begin to realize, okay, it isn’t just a matter of making the computer guess the meaning of the text, which is partly the I route. Then you start to okay, you start, you start to extract and then, you know, I see people do really, really interesting things. And you begin to find connections and collisions. Another thing, just flipping back to maps for a second is, is the is the concept of negative space, which we, we often forget. Now it’s not deliberate here, but but you might, for instance, have been working on a map, and there’s this area here under the cursor. And you might be saying to yourself, well, wait a minute, why is there nothing there? Because it might be in the way that you laid this out that you would expect from your knowledge of the information you’re working on, whether it’s gene therapy an 18th century novel or whatever. It’s all just information in a sense. But you might you might think, okay, there’s something there. And I think this is that’s part of the power, again, of these, these information spaces through.
Frode Hegland: Just really, really briefly. And I’m asking respectfully, not trying to be funny in terms of using the map here. How do you feel different. This is different from the view of the map in author. Of course they are different. But in terms of applying the term map.
Mark Anderson: No, I mean, in a sense, they’re they’re both maps in that they’re just a space full of objects. So in that sense, they’re they’re different. And indeed, I could see that, you know, if I were able to do it, the, the the notion you have of, in a sense visualizing a link as a rollover, that’s perfectly possible. It’s not implemented here, but could be done. I think I just the point I wanted to get across is that I think it’s a thing of treating the map as something you do. Not that something done unto you. I think we’ve been wrong footed by a generation of static infographics where you take it a bucket of stuff, you pour it, you pour it into a sort of big graphic design hopper. And at the end comes something beautiful. And the assumption has to mean that be that it has meaning. And I’m coming to see that in skilled hands that is true. But more often than not, it doesn’t necessarily do something. And the thing of the maps that I’ve discovered through through the practice of just using them is that less is more. So if you don’t try and make them look pretty, you don’t try and sweat them too hard. The, the the understanding just, just floats to the surface, if that helps.
Frode Hegland: That’s great. Any other questions? Today we’re a little bit on a sprint, and I’m really, really grateful that we got this perspective. This is real stuff, real mapping of knowledge no matter what we will call the space. Andrew is up. However, Fabian, did you want to show something and if so, can you do it relatively briefly? Mr.. Please stop.
Mark Anderson: Sharing.
Frode Hegland: Yeah. Thank you. And thank you for sharing. That was really good.
Fabien Benetou: I can do it in two minutes.
Frode Hegland: Two minutes. Okay, I’m going to share something super brief. Then while we’re waiting, just one single slide. While we’re waiting for Fabian, just speak up, Fabian, when you’re ready, okay. I’ve been looking a little bit on what Andrew’s done for today. You said two minutes. Okay, give me 30s just 30s so I don’t do it at the end. But looking at what Andrew’s done, which is phenomenal, now we need to look at how we’re going to interact with it. One of the options, by talking about a hand, I’m going to show you this within 30s we drop it, maybe you all have ideas independently for next time. The notion is that we have to decide what things are we going to do. Often they should obviously be quickly accessible and what think later. So for instance, layout. We’re not going to do that all the time. So that could be in a submenu workspace view saving can be maybe slightly more onerous pulling out from your hand, but choosing and hiding entities same level however, will probably need a way to focus. So when you select something and you want everything else to fade a bit, maybe we dedicate a finger to that just to focus.
Frode Hegland: There may be other things. And then finally, last thing I promise. Mark and I had a very fruitful discussion last week in our session on different types of tags like what’s the difference between names and a view like this in our academic proceedings, view author names and names of people who may have been mentioned are really different things. One of them is extracted from the metadata of the document. As a official thing. The other ones are what the user, through manual work or LMS, may have chosen to extract maybe names of people, products or whatever. That’s why I think we should maybe have it as a separate category. That’s it. Any questions before I close this slide? Wow, such clear communication. I’m sure that’s what it is. Okay. Thank you. If you have further thoughts on how to do this stuff. Yes. Also, Adam texted saying he really wanted to be here today, but after having been away for so many days. Not possible at all. So. Hi, Adam. I know you’ll be watching this. Fabian, please.
Fabien Benetou: So can you. Can you see my screen?
Frode Hegland: Yes, we can say infinite loop.
Fabien Benetou: And you can see this on the right. Yeah. So the first the meta demo part is I can now stream the Vision Pro to my desktop, and then I hope to show you demos much faster now. I activated the microphone on the demo. So that’s actually for Adam to pay even more attention, even if he’s not there. And then it adds fear.
Speaker5: Up.
Fabien Benetou: Add a sphere. Add sphere. Okay, I’m trying again. I said I need two minutes for that part.
Frode Hegland: That is what demos do. Trust me, I live and die in this world. So. Yes.
Fabien Benetou: Add sphere.
Fabien Benetou: Add the sphere.
Dene Grigar: That’s like me trying to talk to my streaming media system, like.
Frode Hegland: Claude. Please note that Fabian’s screen rendering beautifully.
Fabien Benetou: Okay. Add sphere. Well, anyway, you get the idea. That worked, of course. Right before I connected. I guess my English is not clear enough, but basically it’s about voice commands, so adding content and then applying modification to that content. Most of it is done with so it’s done on the device itself, and most of it is just using web API. So it’s like couple of dozens of lines of code. It’s relatively short. It does applying primitives or adding primitives from a frame. So it’s about a dozen of them, like sphere box plane like couple of geometric primitives and then the components. So for example, R I know why the I should have said add color pink and that would have applied the color pink. Add scale. Ten 1010 would have made it ten times bigger. So yeah, it’s relatively short because it’s not super simple, but one can imagine easier syntax. Yeah. Let me know.
Dene Grigar: Fabian. That’s great. So what I’m seeing then is the. Is the possibility of even integrating this with AI, so that AI is developing these things as we’re speaking inside the headset, which is what Apple is supposed to be announcing next week, right? Ai enhancements in the headset.
Fabien Benetou: So I don’t know what Apple will announce next week here. It’s a much simpler than I in the sense that it’s like comments that are interpreted there and comments that are like just to now add and apply. But yes, it’s true that as soon as you you get some text in as input, one can imagine generating more based on this. The trade off is here. It works fast and on the device. And it’s predictable, like, except I mean, the accent, I guess but it’s going to do what you say. It’s not going to delete your documents if there is some kind of crazy typo or something. So, yeah, we can plug all the more complex system, but then the predictability is going to be. Of the window.
Dene Grigar: Thank you.
Frode Hegland: Bark, bark, bark. Okay.
Mark Anderson: Thanks. I’ve really enjoyed that. I’m just going to say something which I hope now is going to be blindingly obvious, which is that, I mean, to draw the link from what I was just discussing to this to, to our experiments is so the, the, the blocks, the blue blocks and the, the white spheres are no different than the abstract, abstract things I was showing you in the map sense. So, you know, when, when, when I look at that, I was imagining, yes. You know, the blue blocks might be short papers and the red blocks might be long papers, or they might be ones I’ve rejected versus liked or something. So again, if if we allow ourselves to think in that gentle, abstract we have this really powerful, malleable space. And what, you know, Fabian’s showing is the programable side, which is really super. So thank you. Thank you for that.
Fabien Benetou: Actually. Thank you for for doing indeed that’s parallel because I apologize. I tend to like just jump in just so the technical part without necessarily connecting with what’s important to us all. So indeed, I don’t really care for the cubes of the pink sphere itself. It will represent. I mean, it is simplification for other things, including documents, citations and all that. Well, so thanks a lot, Mark.
Mark Anderson: Well very quickly. Sorry. I know, it’s just to say that I wouldn’t worry about. I don’t think we have to apologize for the shapes, because it just helps make the point that we don’t have to overthink it, outset what they are. It’s nice. It’s nice to know that we can get to the pretty stage, but at this point, it’s probably wasted effort because it’s actually making the things all too late. Sorry.
Frode Hegland: But don’t. No, don’t be sorry, Mark. Not at all. Dina, just briefly. You guys have gone next Wednesday or not, I couldn’t remember, sorry.
Dene Grigar: We may be able to come for an hour, is what I said.
Frode Hegland: Okay. Thank you.
Dene Grigar: Good. I’ll open at ten. So we need to be at the at this place at night at 915 at the latest. We can walk in 15 minutes. So I’m guessing from 8 to 9.
Frode Hegland: Okay. Thank you. I’ll try to keep it in my head. For the record, it does not seem like we can do this for native apps and visionOS. We can’t figure out a way we can speak commands. It’s absolutely crazy. That’s the way it is now. It may change next week. Right? So moving across to Andrew, and I just have to mention, I was sitting in the very same cafe spot today where I tried your demo, Andrew, where I was sitting six months to a year ago doing the animation, where I close the laptop lid and the the space comes up. Today I did it for real. Holy smokes. Please introduce. And if you haven’t tried it already I have rejigged our website just slightly to make it even quicker to get to the current. Current. Under work.
Andrew Thompson: Alrighty. I guess for starters, there is a now like a dedicated link that I’ll be updating with every new version. So we’ll of course keep the archives like I currently have. So we can revisit old versions. But if we want just one link, that’s always the newest stuff. I have that now. I believe that’s what Freud has set up on the website. Yeah. If you’re the only downside. Sorry. Oh, yeah. Go ahead.
Frode Hegland: No, no. No, please. Sorry. No.
Andrew Thompson: Sometimes if your browser saves the cache, you may end up with, like, outdated styles on a modern build. So if you’re ever seeing something that just doesn’t look right, clear the cache. But hopefully that doesn’t happen. And nothing I can do about that on my end anyways. The main bulk of stuff.
Frode Hegland: I just wanted to say that link is up. I’ll just unmute now.
Andrew Thompson: Cool. Yes. Yeah, that that link shouldn’t have a problem this week anyways. The main bulk of this update is, well, a lot of back end stuff. But from your perspective, it’s got a view of some sort, whether or not we’re calling it a map or a view space or whatever we want. It doesn’t really matter. But it exists and it’s populated by a whole bunch of text. The text are the titles and the names mentioned in the documents that are currently in the library. I also have support for the keyboards, but there were so many keywords it was lacking, so I just turned them off for the demo. That was an issue we knew was coming. But yeah, it’s here now. In the map view, you surprisingly cannot do much of anything because I didn’t get it all working. But the main point is you can move things around pretty easily. And the selection box, which took me, like, the entire time to develop. You can select a big chunk of text and then move all of it together as one, and then deselect it by tapping elsewhere. Hopefully that’s really intuitive. The reason why I say it took forever is because I developed it like 4 or 5 times. And there was always a big issue when I’d get to like, the last stage every time. Either it was like massive performance issues or it just wouldn’t detect stuff properly. I tried so many different solutions until I got to where we’re at, but it is pretty light on performance from what I’ve tested on my end.
Frode Hegland: Quick question. Have you thought about the issue of selecting to do something like selecting to open rather than selecting to move? Because now it seems very easy to move. That’s all I was wondering.
Andrew Thompson: Yeah. So this text you would never be able to open because they’re just tags anyways. How we want to orient things like left and right, align and snap things into position. I have no idea. That’s what I was hoping we would cover today. And it looks like briefly, Fred, from your example slide you showed us. Maybe you’re thinking tapping the other hand. There’s the bright side there where it’s easy to do. The downside is it’s not intuitive. So we could figure out how we want to go with that. And I can absolutely implement it for testing. So that might be a direction we’re going.
Frode Hegland: Yeah. On that. This is okay. One thing that that’s really stood out for me demoing this, the whole headset to different people over the last few months is it is a computing platform. It is not an iPhone. So there will be there will be some teaching involved, obviously as little as possible. I know we all agree on that. So that how you select something in this space and how you move it, as long as we can get to a stage where we feel it’s, you know, after a touch of training, I think that’s absolutely fine. And I’m wondering if any. I know you’re all in in VR space now. Oh my goodness. This is the first time I’ve seen this. Leon, are you a headset as well? Now. Andrew and I are the only ones in the real world at the moment. No. Okay. You’re in the real world. I don’t know. Do you have any feelings for how it should be under the. Touch to to do something rather than touch to move.
Andrew Thompson: I think the default action should be the one that’s used the most. And it seems to me like in the map space it’s about organizing. So the default action should be move. There’s not really anything to open by just looking at the text, but there is things to interact with for sure. So the way I’m picturing it you can select multiple things and then apply actions to it. So you select a bunch of the text and then you’re talking having the off hand be a menu of sorts. So I’m just going to go with that. Perhaps to keep it. Intuitive, because just having to memorize what each finger does is not going to be something a new user can do for easily. It would make sense for a experienced user. So maybe we could have like actual text, like when you open your hand and look at it. Yeah, a bunch of yeah. Different menu items kind of appear and then it would just be the result of tapping each finger. I think it needs to be visual as well. So having little spheres on the fingertips that you can touch would be nice. Maybe they sort of grow when you get close to it. So it can tell you kind of which one you’re about to touch. Things like that would be nice. I think we’re going to very quickly run out of fingers. So I don’t know what you want to do about that.
Frode Hegland: Well, current thinking on that, and I just share the image that I’ve been working with, with everyone is the notion of hierarchy and shortcuts. So for most things like show and hide entities, you would do a thing and a thing, so you would do a thing to open up. Either it could be a list in front or it could be another finger ideally. So you know, you click here for instance show hide entities. And now all the fingers are entities. You know, that kind of hierarchy. I think over time a user will be able to do that without looking. But some of the.
Andrew Thompson: Fingers become submenus, you’re saying. So this would be a line. And then I tap it and then now it becomes left or right up down.
Frode Hegland: Something like that. Yes.
Andrew Thompson: Okay. I wonder how we would intuitively do that, because there’s not a click, and we know that the finger tracking is a little bit off sometimes. So you may tap to go to a line and it may double trigger and then do left. And you’re just like what just happened? It’s possible we could put a force delay between it, but then that would get frustrating if you’re trying to work fast.
Frode Hegland: Yes, I’m just.
Andrew Thompson: Voicing concerns now.
Frode Hegland: It’s there’s a yeah, I’ve been walking through Soho today looking like a complete maniac. Maniac playing with my fingers in different situations. Yes, I agree with all those. So the level of item, the level of submenus is of course a huge issue. I think two should be maximum unless it’s something really, really you shouldn’t do often, right? And we should have some that are instant. For instance, maybe the pointing finger on the other hand is to focus. So we selected something. You want to hide everything else a little bit if you touch that no submenu. But through usage and thinking, we have to decide what are the few things that should be one action rather than. Reveal layout, reveal workspaces, or reveal entities.
Andrew Thompson: I would argue that we keep the the dominant and non-dominant hand and have no menus on the dominant hand. Dominant hand always does the pressing to keep things intuitive that way. Yeah.
Frode Hegland: No, I agree with that. All I meant was, for example, we’ve gone through doing the selecting and now we look at my non-dominant hand and I tap my pointing finger. And that does the focus mode, so that everything that isn’t selected goes to a snap in the background. So I can work with the selected bits, even if I deselect them without losing everything else. Not saying we need that thing, but it’s it’s an example of a should be quite quick to do rather than something to open, what to hide with an entity, etc..
Dene Grigar: Can I say what I like about this? I like the space. I like having 360 degrees to work in. I mean, I’ve got things all around me now. Right? I’ve highlighted and had objects everywhere, so I’m totally surrounded by text, which I think is very much what I’ve been imagining. I’m immersed inside the text and I have levels now. So let me just finish one more thing. Having the the some text on top of text there gives the sense of three dimensionality in the space now, which we didn’t have. You know what it means. So now we’re in a 3D space for the first time.
Frode Hegland: I was just going to say. Danny, I’d like to see the size of your smile.
Speaker5: Can I see? There we go.
Dene Grigar: I mean, it’s just really. It’s really exactly what I was imagining. Andrew.
Frode Hegland: Yeah, I think you can buy yourself any alcoholic or non alcoholic beverage on us today. Yeah, it’s that kind of a day.
Dene Grigar: He doesn’t drink, but we’ll get him something. I’ll get him chocolate. He likes chocolate.
Frode Hegland: Two, two virgin special mojitos or something. Yeah. No. It’s wonderful. Danny, you raised a really important point, and that is the three dimensionality, which, of course, can very quickly become messy. So, of course, we must use that dimension. And how are we going to use it? It’s going to be a very interesting challenge. That Mark was so polite. Go on.
Mark Anderson: Yeah. Just quickly I was reflecting on the fact that just how much of a trip up the dominant hand is because I being left handed, I just immediately tripped over that without thinking. And the interesting thing is, and this shows because I probably didn’t spend as much time in this space as most of you here is the sense that actually the hands have a different purpose. And so, you know, in as its run at the moment, I must use my right hand to do certain things. It doesn’t matter. And I know that we have plans to make that, that, that change. But but I was just reflecting on on and going back to your points about, you know, a new starter. So one of the probably one of the very important things we want to figure out from the get go is to make sure we’ve got the right dominant hand. Otherwise probably everything else will unintentionally just not quite work as smoothly as we think, because they’ll keep going to the wrong place. That’s a point of reference.
Andrew Thompson: I can comment on that really quick. This is horribly unintuitive. And it just kind of there because we needed it. But you can currently swap dominant and non-dominant hand in the software. When you’re first entering, you keep one hand behind your head and then you load it in like that. And that will swap dominant, non-dominant. In no way do I want to keep that long term. It’s just it’s there at the moment. Eventually it’ll be like a toggle or something.
Mark Anderson: And I just want to acknowledge that. I mean, notwithstanding my comments, I totally take that as given. So it’s not a sort of complaint. It’s just an observation about about the sort of in a sense where in the process, when we when we get to the more finished part, we need to introduce some of these things.
Speaker5: That’s.
Frode Hegland: I’m glad you say that, Mark, because it brings up the the sphere on the wrist again. What what what it’s there for. And I wrote half jokingly in slack today that maybe we should put a clock so you can see the time on the sphere. Because when you’re in this space, you can easily become a little lost. It’s nice to have access to it. And of course, in the Vision Pro there is no clock unless you go into the settings. It’s ridiculous. But now I’m realizing that the kind of setting you’re talking about is exactly the sort of thing that when you tap on your wrist, you should get that prism. Maybe the prism will interact in a different way, as we’ve talked about, but that should probably be a change. Dominant hand. Another one should be.
Speaker5: Yeah.
Frode Hegland: Maybe even show the time on there. And maybe it should actually there. Show the data you have in front of you, in case you’re wondering what it is, and maybe the reload the load button for all the data should be there. In other words, we need to look at the prism as well.
Mark Anderson: There’s some useful prompted by your observation, and I know it was a throwaway comment about the time. It made me think, well, actually, yeah. But it might also show you either the session time or your battery state because, you know, for a few years yet they’re not going to run for hours and hours. So there are two aspects there. Like, have I been doing this for too long? I should give everything a rest. And the other one is, is all my information about to disappear into a dot in the middle of the screen because the battery is flat? So actually there are some quite practical things that one can put there. So they, but that are either telling you about your overall outer environment and indeed time or a rundown clock to a next appointment or, you know, if you, if you’re working in a very diaried environment where you have to keep to a time schedule it might be that it might be a countdown to the amount of time you’ve got left in your session. For instance before you need to be somewhere else and just and just also, I think useful sort of human cues, you know, essentially the, the sort of nudge that maybe you should just, you know, take this off for five minutes to give your eyes the rest. So there’s lots of fun things that, that, you know, going back to the original idea, and I think it was probably suggested, just in a sense, having this in, in environment sort of access through to controls, I think is a really useful, a nice idea.
Dene Grigar: Can I mentioned also was I think is this the the environment is quite beautiful with the lines running through it. And you mentioned a flaw. I don’t see a flaw. I just see a I’m inside of a of a round space. I’m inside of the of the sphere itself. And there is no doubt. I mean, this is what Eduardo Katz talks about in his space poetry is that when we’re in these spaces. There really isn’t the kind of directional. Layouts that we have become accustomed to in gravity spaces, right? We’re free from the gravity of the Earth in this space.
Frode Hegland: I would say that the weight of the human head is important. For instance, to look a little bit down is easier than to look a little bit up. But also there is not a visual flaw, but there is an area beyond which you can’t really see, and it’s kind of down there. And similarly, when you put things up, it becomes really small. Obviously it has to. So I’m not saying that as a limitation. I’m saying it as something that I felt was real and something we can use as. And we have a sweet spot that is 360, of course, but it has a certain height and a certain lower ness, and I don’t think the floor is aligned with the physical floor. I think it’s probably quite a long way below that. But this is exactly the stuff for us to discuss how we how we perceive it and how we want to be in it. So that’s pretty cool.
Dene Grigar: Well, I’m going to say that in a sphere there is no floor and a sphere there is no top, there’s no side because it’s it’s a ball. It rolls right. So we’re in a it’s obviously a sphere.
Speaker5: No cylinder.
Frode Hegland: We’re on a sphere. We’re in a cylinder, I think.
Andrew Thompson: Yeah, technically we’re in both. We’re in a cylinder inside of a sphere. So that’s two people both agreeing, but seeing it from a different perspective.
Speaker5: That’s what it’s all about. That’s what it’s all about long is.
Dene Grigar: Both the perspectives are. Attended to, right? I mean, I I’m not seeing myself. I’m situated only because I’m sitting in the chair. But if I’m standing, I’ll have a different perspective. If I’m sitting on the on the floor of my office. But where I am right now, I’m surrounded by text. I can put some everywhere in this space and there is no. One perspective about that.
Speaker5: Absolutely.
Mark Anderson: This is just occurs to me. Is it the case? This probably is one for sort of Andrew and Fabian, I guess. Is that just just as a sort of because I ask a lack of knowledge, is that if one chose to so and obviously, if one says if things are in a sort of cylinder or basically they’re equidistant all around us, then it’s at the same, essentially the same visual distance, same presentation distance. And I assume that’s one of the the attractions of doing that. And also I don’t have well, I can rotate, but I don’t have to move or I can, I can rotate it around me and I’m not changing the view distance. But presumably if one wanted to, I could say have a wide presentation. Yeah. I mean, you know, one of these things where people sort of talk about seeing ten pages, where in a sense you scale things, you just scale in perspective as it goes out. So in other words, it remains all readable. It just means that the the outer end have to be scaled as drawn. I’m not saying I want to do this. I’m just trying to sort of understand because I’m a bit like the thing about, well, Nordstrom’s a print and what we use because we’re used to them as opposed to needing them. Is that funnily enough, the sort of spheres turn up quite a lot in sort of sci fi and things as the, in a sense, the obvious display space. And I think partly because the idea was it’s all equidistant around you. But whether we want or need to do that this idea that maybe we just teleport to somewhere else is, is a possibility. So I’m just interested in. In a sense, what the really, you know, what holds us to using a sphere or cylinder against other things?
Speaker5: Yeah. I mean.
Andrew Thompson: You kind of hit it, Mark. It was it’s so that you can read everything without getting the sort of compressed text. We did try flat spaces very early on. Fred, when you were here in person, actually testing what it was like to read text on a plane and yeah, it very quickly became we realized that, yeah, you have a whole bunch of theoretical infinite space, and very little of it is actually useful at all. So we’re like, well, if we’re only using this much anyways, that’s all we can read. Let’s wrap it. And then you actually get more space on the sides.
Frode Hegland: Yeah. One thing though, we have decided that I think. Anyway, the interaction is swivel chair 360. Not moving about and stuff. So that’s these are constraints that we have to evolve, obviously, but also consider but it is on a kind of a side note, today I was sitting trying to concentrate, working in a cafe where I was at a counter in front of a window, and I was working in a native vision app. Off author, obviously, and it was really nice. I could concentrate, but when people would go by the window and basically stand in front of me, then vision would make the window transparent. My my word processor transparent, right? So there’s all these weird things we just don’t think about before it’ll be. So when I’m talking about up and down in this room clearly there is no up and down as such, but the fact that it feels a bit different here and there, it’s just so exciting for all of us to see how our physical heads are in this environment, to see how it can really make it wonderful.
Mark Anderson: But it’s funny having a sort of that there is a sense of doubt. So maybe when you look down, there should be an awful sort of eldritch, you know, Lovecraftian monster staring back, and now a pair of jaws staring back up at you from the abyss. But also, you could look down and you could see all the things you discarded, as you know, like crumpled paper on the floor.
Dene Grigar: I’m going to say this. There’s some wonderful Portuguese electronic literature artists, Rui Torres being one of them, whose imagined spaces like this in a 2D environment. I mean, a lot of his work looks like this, right? But he was trying he was envisioning it for 2D. I think he would really enjoy seeing this, his work expressed in this environment. Don’t you think? Andrew, you remember Ruby’s work?
Andrew Thompson: Could you mention the name of one of them?
Dene Grigar: Carissa. Is one of them. I’ll show you today when you come to laugh.
Andrew Thompson: It’s not ringing a bell. Yeah, yeah.
Dene Grigar: There’s two of them that are more Carissa and a few more that are just like this, but flat. But this is so much better what we’ve got here for him. I could see him designing in this space.
Speaker5: So we have.
Frode Hegland: Four minutes left and it’s been a really glorious day. All I will ask you all to background processes in this view, forgetting everything else we’re doing, if you can. What do you want to do and how? You know I’m going to be plugging ahead with what I’m doing, so let’s not have that as the default. It’s very, very productive when we disagree about these types of things. So Denny and Andrew, you’re not around on Monday then? Because of the conference, right?
Speaker5: Right.
Dene Grigar: Okay, not the conference. The exhibition. We’re doing the exhibition. Sorry.
Speaker5: Yeah.
Frode Hegland: I’m just put that in my calendar. So we will just do, I don’t know, maybe just talk about this view on Monday, maybe keep it short and then on.
Speaker5: Yeah.
Frode Hegland: I don’t need to put anything. So we don’t meet for a pre meet you and me on Wednesday or on Tuesday. Right. This next one?
Dene Grigar: No, I’ll be so busy.
Speaker5: Fine.
Frode Hegland: As long as it’s not in my calendar. And of course, on Monday there is the Apple WWC kind of a thing.
Speaker5: So can I mention about that?
Dene Grigar: I dropped in the article the the idea that they’re probably going to be announcing some sort of AI integration in the headset. So it’s Siri inside instead of Siri outside. And that’s what I was referring to. Fabian. So if you looked at that article.
Speaker5: Oh. By, right? Yeah.
Frode Hegland: No, that was interesting. Thank you for that. Dini. Of course there is a basic AI. The voice is much better in Siri vision than it is on any other device we have. Yeah.
Mark Anderson: So now Siri can be in the room with you telling, telling you they don’t know what the answer is, as opposed to telling you that the speaker on your desk.
Speaker5: The level of.
Frode Hegland: Ineptitude with Siri now is legendary. We’ll go in history books. Okay. I’m also going to go I’m going to have a family birthday dinner. I said we could do it early today due to Short-Term eating. Is there anything else that you guys. Because I’m changing the prompt. Is anything you’d like to say to our dear friend Claude? As in something that you would like to see if it turns up in the transcript.
Dene Grigar: I wish I had money to bring Rui Torres to my lab so he can make art with this environment.
Speaker5: But can you.
Frode Hegland: Can you say it using that LMS name? I want to see if it’ll catch that you said this and this to it.
Dene Grigar: I’d like to bring Louis Torres into the.
Speaker5: No no no no no, you.
Frode Hegland: Have to say as though you were talking to Emily in the room. He said, Emily, I would like blah, blah, blah. So you use that name? I won’t say it now because it’ll confuse it. Let’s see how it works.
Dene Grigar: Hey, Claude. I like to have Rui Torres come to my lab and do some art with this new environment that we’ve been building.
Frode Hegland: And I think that is a good idea. Anyone else have any messages for Claude? Other than saying thank you, Claude, for helping us with analyzing our record.
Mark Anderson: Right. I just want to wish. Denae. And, Andrew, good luck with the exhibition.
Frode Hegland: Yes. And the odd picture, if you don’t mind.
Andrew Thompson: Thank you. Mark.
Dene Grigar: What’s really nice is we’re working with people that have the tech already there. So all the Apple S’s and the color classics are going to be sitting on a table for us. I don’t have to ship anything that was like $5,000 with the shipping. So it’s so nice. And then we’re driving. The big 2014 Imax.
Dene Grigar: With that. So I’m very excited. It’s gonna be a lot easier to mount. Well.
Mark Anderson: The one thing I didn’t get is whereabouts? Victoria. Where in?
Dene Grigar: It’s at the McPherson library. And on the downstairs floor there’s this beautiful exhibition space. Yeah, but we’re working in Victoria, BC.
Mark Anderson: Oh, right. Okay. Yeah. Okay.
Dene Grigar: And John Durno has built a it’s called the Computing History Lab. And so he’s but he’s an expert in telidon. So you might know the telidon or that technology, the videotex machines in the 80s, the Minitel was the one that was popular in Paris.
Mark Anderson: We had Ceefax and stuff here.
Dene Grigar: Yeah, yeah. Ceefax. Yeah. So every country was building these environments. So he’s an expert in Telidon and we’re showing for Telidon works.
Speaker5: Yeah.
Mark Anderson: So? So I’m really sad I found out about Plato after it got closed down, because that’s something else you don’t see.
Dene Grigar: But, Andrew, this is amazing work today, and I’m bringing the headset with me so we can take pictures. I want some photos.
Speaker5: Oh.
Frode Hegland: Thank you. I was I was a bit annoyed that we were keeping you. Bettina. Yes. If one of you has an iPhone 15 Pro, do some spatial videos, please. And if you can’t do that, record video with the headset standing really still because the spatial video can then share with us, it’s a really wonderful way to keep record. I did film a little bit of the guys this weekend, including Leon ten minute testing. It’s just so much fun. Anyway, thank you all. Rush rush rush.
Speaker5: Bye bye bye.
Chat Log:
15:59:38 From read.ai meeting notes : Frode added read.ai meeting notes to the meeting.
Read provides AI generated meeting summaries to make meetings more effective and efficient. View our Privacy Policy at https://www.read.ai/pp
Type “read stop” to disable, or “opt out” to delete meeting data.
16:04:46 From Frode Hegland : https://public.3.basecamp.com/p/RU5WajhQcG3tm1eCcXEceqWN
16:04:53 From Fabien Benetou : Have a brief demo to share, small but should interest most.
16:05:06 From Fabien Benetou : (not on agenda as finished 10min ago)
16:08:22 From Frode Hegland : It really does not look new, so no worry about that 😎
16:11:35 From Fabien Benetou : (also have a meta demo to show)
16:21:41 From Frode Hegland To Fabien Benetou(privately) : How long will you need?
16:26:37 From Andrew Thompson : Thank you Mark!
16:26:47 From Dene Grigar : Yay, Mark!!!!
16:27:19 From Mark Anderson : Reacted to “Yay, Mark!!!!” with 😀
16:27:22 From Mark Anderson : Reacted to “Thank you Mark!” with 😀
16:29:10 From Mark Anderson : Peter W also sends apologies but will watch the recording. 🙂
16:29:15 From Frode Hegland : Reacted to “Peter W also sends a…” with ❤️
16:29:59 From Frode Hegland : Reminds me: I’ll add a quite note to ‘Claude’ to see if it can catch it. I’ll be modifying the prompt to see if we can do this.
16:34:37 From Frode Hegland : For the record, it does not seem like we can do this with native apps in visionOS. We are trying.
16:35:00 From Fabien Benetou : this i
16:35:11 From Fabien Benetou : this is what I use to mirror from VisionPro to Linux https://github.com/FDH2/UxPlay
16:35:44 From Frode Hegland : https://futuretextlab.info
16:40:42 From Fabien Benetou : did you see me sharing?
16:40:51 From Fabien Benetou : (the weekly experiment now I meant)
16:40:54 From Leon van Kammen : yes
16:40:58 From Fabien Benetou : ok great, thanks
16:41:09 From Fabien Benetou : thought it’d be good for people without HMD and the record
16:41:21 From Leon van Kammen : Reacted to “thought it’d be go…” with ❤️
16:42:11 From Mark Anderson : Channel/embody our inner octopus → more finders/legs
16:42:25 From Fabien Benetou : Reacted to “Channel/embody our…” with 🐙
16:58:00 From Fabien Benetou : actually gotta run, bye
16:59:47 From Mark Anderson : Best wishes to Dene & Andrew for the exhibition.