Frode Hegland: I can see your Bluetooth. Connecting. Up to pair a trackpad with vision, even though it’s connected to a mac. Oh, hello. I’m just trying to pair the trackpad here.
There we go. Did I? Oh! Argh!
Frode Hegland: Now it’s Where is it used? Okay. Turn off the music. Okay. That’s better. Hello.
Peter Wasilko: Good morning.
Frode Hegland: Morning, morning. You can probably see yourself here. In the reflection. Yeah, I just realized I hadn’t managed to get the. Trackpad to. Here, but that’s okay. Hang on. Take this. Back off. It turns out Apple only has a black keyboard with numerics. You can’t get a black keyboard with just the basic keys, which is kind of weird. Anyway, how is everyone this morning so far?
Peter Wasilko: Very good. I will have to bow out after the first hour to provide transport for mum. So if there’s anything that needs my input, make sure it’s covered within that window.
Frode Hegland: Okay, that sounds reasonable, I will too. Okay. Yeah, well, I’m glad you guys are here. So we’re having an introduction from Fabian joining us today as well. I the French name Alain Marcel would be the English version. So they’ll be coming in to show us something that I quite don’t understand yet. But it’ll be interesting. And Yeah, I think there’s quite a few things to discuss today, so it’ll be. It’ll be fun.
Mark Anderson: You saw I put a couple of things in the Project Slack thing that I come in. So bridging across the future text, I thought Julie’s post about keyboards in the Ars Technica article. Yeah. I don’t think any of it surprise would surprise us. I mainly put the Ars Technica one because I just thought it’s it’s all useful collateral to to the information. We’re building it.
Frode Hegland: Yeah. No, it’s absolutely worthwhile. And okay to kind of start on that note.
Peter Wasilko: Well, can we see your persona fraud?
Speaker4: No. That bad?
Frode Hegland: How would you be able to do that?
Speaker4: Well, didn’t they.
Peter Wasilko: Say that you could interact with people who are in zoom using your persona from the Vision Pro?
Frode Hegland: Yes, but I’m not in the US, so I don’t have an American iTunes account, so I cannot download any apps, including zoom.
Frode Hegland: Yeah, so I’ll have.
Peter Wasilko: To fix that.
Frode Hegland: I have learned over the last few days I have what we impolitely call resting bitch face. That’s what the kids call it these days. And in my persona, it’s even worse. I look like I’m, you know all the time. So. Yeah, it’s it’s not the best.
Speaker4: Yeah.
Frode Hegland: It’s. It’s been interesting. Hey, Fabian.
Speaker4: Righto.
Frode Hegland: So here’s the thing. I’ll just throw it out while there’s just the six of us. Having spent time now, as you also have Fabienne in the vision and you Rob. In terms of.
Speaker4: Work.
Frode Hegland: If you have an opportunity to sit at a normal nice screen, there’s absolutely no benefit to using the Vision Pro at all currently. So obviously, hopefully we can fix that for part of the time.
Speaker4: Am I wrong? Discuss.
Frode Hegland: Well. So one one silly things I did, especially.
Fabien Bentou: At the beginning of the pandemic. I bought some kind of a laptop holder that I could fix on a tree trunk and on the glass windows at home. So I do have a need to interact with digital content, including documents. But I don’t like being. I don’t like to have to be behind my desk. And including in some, let’s say, strange things like being reading in the middle of a park. So it’s the, the comparison goes only up to a certain extent. And I also, it’s not the same as a phone like I want also something that has quality input. So I’m, I don’t know if it’s if it’s only that. I don’t think it’s interesting. I think it can do more than that. And I think just the ability to move around to read in, in a stranger place than just behind my desk. To me, there were some benefits to it. It’s of course, limited in terms of, yeah, price and reliability and even the battery, etc. but yeah, I still think there are some things to explore with, let’s say, just that model.
Frode Hegland: Hello.
Speaker4: Hello, hello, hello.
Frode Hegland: I’m Norwegian English, so I’m going to be very careful with pronouncing French names. So I’m going to go with. Hello.
Speaker4: Hello. Bonjour.
Frode Hegland: We look forward to hearing hearing from you in a bit when everyone is here. Currently we are talking about I was just throwing out the statement that so far, even though I love the Vision Pro, I really don’t find any benefit working. I usually take it into town when I’m at a really wonderful coffee shop, and I tend to find that even a 13 inch MacBook screen, which of course is far from ideal, is okay putting on the headset and moving it up. Apple has done some things really well with the text screen, and I now have a working version of author native version. Of course, it doesn’t do everything, but just getting the screen right and all that. It doesn’t. It’s shockingly to me, doesn’t provide me with that much more affordance. And I think that is why what we’re dealing with on the Wednesday with the Sloane, which is trying to use the space, the width of the space more than the depth because depth, we often talk about putting things there. It kind of gets messy in the background because even in the vision, if you have two apps semi overlapping each other, it becomes a bit messy. I don’t know what you guys think about that so far.
Fabien Bentou: Out of curiosity, did you play with in the browser specifically with different windows? So to clarify, for people who are not necessarily used to it, but you, of course, so you put the headset on, you have the browser window as you have on your desktop. You can have different tabs. But you can have different windows from Safari open. So you can have, let’s say a PDF there page there, another page, I think even like you can freely position them. So I’m wondering if you if you played a bit with that.
Frode Hegland: No, I haven’t played much with that. Not with the browser, but I have had different PDFs open the browser. So I’ve done many things, but even now, when I’m working on my PhD revisions, I don’t often need that many documents open that wouldn’t necessarily fit on a 27 inch screen. So that’s why the competition for me is very much the screen, because we talk about a headset being like your thinking cap. And similarly, I guess we can say that our big screens that we have at our home or desk is also like a thinking cap compared to a laptop in a casual situation that mark.
Mark Anderson: Yeah, I find it really interesting, especially this last couple of things, the Ars Technica article and the Ulis one about the keyboard. It is interesting because I find myself reading that thinking I agree. But also I’m thinking to what extent do we do this classic thing though, that we take the thing we’re really used to and then complain massively that the thing that we’re very used to now doesn’t work in the new space, which could lead you to one of two conclusions. Either the new thing is useless because it doesn’t work with the old, or merely. We’re reminding ourselves that we need a new form of interaction. And that’s sort of the thing I think we’ve been beginning to discuss in the Wednesdays. It’s not going to replace everything, but I’m also thinking, so if I’m working in XR how much of that actually has to be typed input literally typing text. Or is it more consumption and triage and moving things around or tasks that could be done within a sort of, you know, positive affordances? I don’t know, but I think this is the interesting exploration. I don’t doubt that the text input will lag physically, perhaps more than the sort of software side, simply as we work out what’s practical in the new interactive experience. Fred.
Frode Hegland: I don’t actually have a problem with text input, because using a physical keyboard with a native app is really actually quite fine. It’s. Yeah, it works flawlessly. Now, I was thinking earlier today about introducing a joystick into our XR experience, and I don’t mean a proper flight joystick, I mean like a toy joystick that just has, let’s say, four degrees of movement, because one of our big issues, of course, Fabian will have one. Let’s have a look.
Speaker4: Well, no, that’s.
Frode Hegland: Yeah, that’s not a bad idea too, because, you know, when in history have people decided when they’re thinking they want to have less in their hands? We always want to have something in our hands. So. So the input is fine, but even just Yeah. And this is not a complaint. It is just for me to say that I think what we are doing is more important because we are not coming at this purely as fanboys saying XR is better, end of story. So it makes it more interesting. You’re saying please.
Hussain Panjvani: I so I’ve recently just bought the quest and it’s my first experience of using VR where I’ve used it as probably a different user to you guys. Because I’m not actually typing much, I’m using it more for consumption. Rather than having my laptop on top of my quill and balancing it along with everything else that I’m balancing, like a cup of coffee or something, or maybe not coffee at night, but rather than having something there, I’m just having it in front of me and placing my screen where I want it. Has been really useful to me in that sense. So from my current setup, okay, I don’t have a 27 inch screen that I use for my normal sort of use. I do in the office, but that’s more for the staff. But just my normal use is usually just a laptop with me reading a PDF document. But this has been good that I’ve been able to just set up my screens with the desktop and a reader in the in the quest, so that’s where it’s benefited me. When I tried on your Vision Pro, it was a lot clearer, as in just the general view of a of of the sort of I use was a lot clearer. But yeah, that’s my use of the VR so far.
Speaker4: It. Yeah.
Frode Hegland: One thing that has become clear to me over the weekend, while thinking about all the views and stuff is and I know it’s obvious, but interaction is really where we should be focusing. So without going into too much Wednesday work we we now have a nice way to view a row of text like, excuse me, a column of text, like a reference section and interact with it is coming along really well. So now what we need to do is allow you, wherever you come across stuff, to be able to do stuff. But you know, we can’t have that to be a static world. And that comes back to. You know, the question of how much of this can be done on a on a screen. Now Adam has made me do panoramas in Photoshop that I then put into photos and then umboo on the headset to give a sense of a lot of text around, which is really, really useful. And what I have seen already is the mock up we have with 100 references just in a row is kind of readable in in the vision in XR, but when I try to fit it on.
Speaker4: Screen.
Frode Hegland: In keynote, it’s just it gets too tiny. So that is the beginning of understanding how the space helps. Hello, Leon. So I’m not sure about Adam and Denny, but Just doing. Just a few text messages.
Speaker4: Yeah. Peter.
Peter Wasilko: Yeah. I think there could be a lot of potential for programmers taking advantage of that extra special dimension. So much of what we do is based upon what nested scope we’re operating inside of. So it’d be very interesting to see if you could imagine starting at the highest level of the program and then just have the environment around you zoom in so that you’re at the next nested scope, and maybe have some secondary side panels to tell you what current bindings of variable names are floating off to the side.
Frode Hegland: Yeah, I think
Speaker4: So I know it’s.
Peter Wasilko: Not the standard scholarly use case, but it’s so closely analogous that looking at coding in XR. As opposed to coding for XR. Could be an interesting maybe a follow on grant down the road. To dive into it and later point.
Frode Hegland: Yeah. No, I think that makes sense. Peter, using the space for that now.
Frode Hegland: One thing that I’ve been asking a few friends recently is there used to be something called flash, and it used to be something called HyperCard. Is there any software today where you can make something with interactions without programing? Please mind and then share that.
Brandel Zachernuk: There is reality composer. The scope of the interactivity is you’ll need to use a reality composer on your iPad today. A Reality Composer Pro on Mac is a slightly different.
Speaker4: I’m.
Brandel Zachernuk: But that’s something that people are interested in in a lot of places. And and there are other people following that kind of model in a lot of other different ways. So the reality composer has a sort of a click on, like a card based interaction schema where you say, like, when I click on this, I want that thing to turn big. When I click on that, I want or when I move close to this, then I want it to play a sound.
Speaker4: It’s.
Brandel Zachernuk: Not amazing in terms of its the ability for that to be able to scale today. So if you saw the do like, mudkips thing that I wrote I programed that with the same underlying mechanism using the same data format that that graphical user interface constructs. But it’s certainly something that people have a lot of interest in being able to pursue and improve in the future. Yeah, but.
Frode Hegland: Apple Reality Composer is only available inside Xcode, right?
Speaker4: No, I mean, you.
Brandel Zachernuk: Have to download Xcode, but it’s not. It’s a standalone application just for making stuff. So it comes as a package with Xcode, but it’s not obligatory to.
Speaker4: To use it
Brandel Zachernuk: To do coding with it. You can build things that are independent of having to create compile projects and files and app stores and stuff like that. You can distribute Usdz.
Frode Hegland: Okay. That’s interesting. That’s obviously 3D based. It is. It is a bit baffling, and I think we need to think about how we want to go forward with this because you know, I am mad about metadata these days. So the question is the authoring system going back to Macromedia Director and, you know, go live and all these wonderful things, they either author to their own thing or they author to another open thing. And if they author to an open thing, if the metadata is in many different applications can be decided for how the interaction should happen. So I think that’s kind of interesting. But I think we can. Yeah. Mark, please.
Mark Anderson: I just couldn’t stay in one place that those well, they’re not specifically apps, but I mean, I think there are lots of things happening around hypertext and knowledge tools and something where essentially you’re working in a surrendered space, whether you’re turning it into formal HTML pages or not, is almost neither here nor there. Whether it’s doing quite the same UI type interaction as something like HyperCard does is a different thing. But I think there’s plenty of writing going on in that modular form, and where people are actually giving quite a lot of thought to supporting metadata that that enriches the various views. It’s just it’s not particularly bundled up as an app that screams at you. This is, you know, in the way that HyperCard is for making HyperCard stacks.
Speaker4: Yeah.
Frode Hegland: So. Okay, I was going to wait until Dini joined us, but she’s had a heavy weekend, so she may not. In the discussion. Mark, you have often said to your credits that documents today are basically modeled on paper. You know, we need to go beyond that. And of course we all agree. And what does that mean? And then this weekend, I realized that the notion of electronic literature. Really comes into this because and it wasn’t actually through Denny’s work or the ELO. It actually came up when when looking at different levels of digital. And there was an archival company who said that a digital document can just be a scanned PDF with no inherent, you know, interactivity. But electronic literature is kind of born native digital. So that becomes A very interesting component on that. So that’s why also with a slight bit of rewriting I have to do for my thesis, I’m basically saying, can we make academic literature become electronic literature to enable these hypertextual affordances? And the hypertextual stuff can only happen if you have the right data. So that was kind of fun. I’m looking at the sidebar here. Of course, we should be using slack, but fair enough.
Speaker4: There’s a lot of.
Frode Hegland: Stuff about webXR and Yeah, I’m just looking at your comments. Right. Alain, would you like to present or rather introduce yourself? Maybe we should just briefly everyone else introduce themselves so you know who you’re talking to.
Speaker4: Sure, sure. No problem. My pleasure.
Speaker4: This.
Frode Hegland: On my side. We are working with since a long time. That’s a long time. A few years, but a product that is focusing on the knowledge. And this my background is software developer since about 40, 40 years, not a developer, but a part of
Fabien Bentou: A company that.
Frode Hegland: I found in the US about 40 years ago where we create what we called at that time groupware, if you remember that. And so we were we have released some software called Thinking Time. Marco Polo document.
Alain Marsily: Management markup also at that time. But, you know, the market for Macintosh was very, very small. And then we, we moved to, to different market from windows. And now we are leading in the education market. We are we have developed a CMS for a university where the company, more than a campus, is now the full leader in the US and also in, in Canada. Thus thus we I have a good background in edtech market. And this my, my the since I’m going old was to really create something that is that’s I feel really good about document management, but also knowledge management, learning tools and so on. And as I start two years ago, a software with a partner called Skill Pass, where we, we were really focused on the knowledge for transmitting knowledge inside company. The software for the application was called a skill pass. But since the market is really red ocean with the Covid, we moved to something that is now called quidsi and Quidsi. It’s something that is like aggregation of all my, let’s say, understanding of what is good for transmitting knowledge and inside company, but also in groups. And the idea is to create something that helps people to structure their, their content, but also to try to keep the context of their content.
Alain Marsily: And that’s the idea of this. If you don’t mind, I can also make a quick demonstration of of the software for the application. So we have already the the word software in mind. But the point is we we are now in beta. This the application is working. It’s really let’s say not a it’s a lot of feature inside the application just from the authoring tool that was the, the the background of the learning tool that we have created in the past to something. Now it’s more to a shareable, a shareable content. So we we went to the idea of sharing content between a members in a groups and also share to web and so on. But the the point with Quidsi is to the idea is to mixed all kind of file type inside one player, let’s say one organizer, where you bring Excel, word, PDF, text, image, video embed file, everything in one place, and you structure your document like you can’t do with folders. And you can also add some comments inside the, the, the structure to help the user of the attendant and also or the recipient to understand what you want to share and let them comment if it’s necessary to do the first point. Sharing.
Alain Marsily: But let’s say you create your content, you organize your content in a kind of what we call smart file, kind of shareable folder where you structure as an author, you contain and you share this content to your audience or for yourself. And the idea is to let the user and or the recipient understand what you want to share. That’s the idea, is to help user not to to go to a drop or a Dropbox folder or a Google Drive folder. But the idea is to keep the same structure that you have created, and you share the structure to your recipient. And on that side, they have no need to have any application because everything is visible. So what you see, what you get or what you send, what it’s what you get. So the idea is to say with, with the new new generation, the user doesn’t understand what is a folder generally? Also, they don’t want to play with a lot of application. The idea is to have a meta player where the user can really read everything that was sent by. The the the author. And when we are we have not said the the file format for display. We plan to have the the idea is we have the kernel, we have the core.
Alain Marsily: And the idea is to let now the user understand what we have in mind, because it’s what my partner in the US, it’s William, a real perceptual tool. And we want something very practical. And we we are now going down, going to something that helps the user to share the content. Is this we we would like to simplify the, the the interface and make it smooth, intuitive let’s say powerful. And we plan to have intelligence or let’s say smart or intelligent insight. Quincy, in the next few months, when we have really established the foundation of the application with the user. But that’s the idea of coincidence. I’m following a lot of work in the future of work. But also I’m, let’s say a son of Ted Nelson. Doug Engelbart. And, you know, in Belgium, we have palatal that was also some, some one that was really involved in the the way to access to the information or document this I have a long path in this in this content because I like this matter. Well, that’s the idea. If you want, I can share a few minutes and you have access. Also, if you want to have access to the application, it will be my pleasure to to let you in.
Speaker4: Do?
Alain Marsily: Do I go to to the screen? Fraud.
Frode Hegland: You should be able to share a screen. Yeah. Okay.
Alain Marsily: Let me, because I’m not working a lot with. You have my screen on my on your on on your side.
Speaker4: Yep, yep.
Alain Marsily: This. You have something that I have a notification this inbox and outbox inbox is is updated. At the moment this is hidden, but the idea is. You can have inbox for the users, can brings some smart file, or you can also share page like you do with notion. And if you go to something like this one, you click it. You have something that helps you to understand that it’s a kind of a Spotify look and look and feel. And the idea is that’s something that was shared with my friend Will William and I can co-edit this this. I have two coeditor here. I have an item. It’s two megabytes long and I can also view. I have this content view different with different way. So you have card view and you have also the idea of. Screenshot and content. That’s the idea. And you have also the ability to have kind of presentation. And you have a player like you do with PowerPoint or any presentation tool. This This you have for for way to view your your content. This that’s the inbox. You have the board. That means if you want to organize your tool and you have columns like Kanban, and you can create your own column and you can follow the content that you receive in the way you want to track them or follow them or approve them.
Alain Marsily: Smart fun. It’s You the. I will create one for you in a few minutes, but the idea is you have a smart file that you shared, so that’s the one. Sorry. This version is my test version. This don’t use. If you have the the URL that’s internal use. Sorry, but don’t play with this but that. That’s for having a lot of content. I want to share my development version. It’s also a smart file that was shared to me, and I can also follow the the user if he has read five item on the 21. And if you have a check mark, that means it has read read all the content and you can also use the favorite, favorite if you want. Library. We have an internal API of a KPM inside. This is where you have page like you have in notion. So if I go to the page. I can play with this one. Just so you can edit and you have a tool, let’s say a subset of notion that if I create one, let’s say new page.
Speaker4: Typal.
Alain Marsily: And you have blood minus. That means you can always add some stuff. You can also go to your library and if you want to add some some content from this library, you do like this. This fight type is not supported. Sorry. Thus, you can create easily. Page. You can as you do with notion. You can change make a title. You have all the, the normal tools like you have with medium, the same ID is, is, is included inside. That means you can create your own content if you want. The library. You have slides. That means that will be free from like HyperCard. The idea is to have free form modules and text and create your own slide and make link with different slides or different page or different item that you want to share. So it could be Excel files or whatever. So it’s not yet done. It’s a development. Right now you have component. A component is let’s say if you have several text or content that you want to to, to mix, that means a video with text, with PDF, you can create a component that will be stored as an entity. And this entity can be added to any page. And the idea is how to use that to see if I want to do that like this I can. Sorry. I can select this one. And it’s not working. I know it works.
Speaker4: I guess whatever.
Alain Marsily: This is, you can create a component, but normally you can add different content. But I don’t know why it is sometimes that’s the kind of stuff. But the idea is you can create a component and this component can be added to any page. And if you have add this component in several pages, it will also give you the connection between all the page and the component you have in add in your your page. This the ID, if you change the component and it will be spread in all the page and also in all the smart files that contain page and so on. So it will completely nested. Okay. You can you can add and manage all your PDF if you want media you have. Audio, video, image, embed file also. That’s that’s also a link if you want to add some URL from this here we have URL and you have embed file and files. That means could be word Excel. But we will add some other format also from Adobe because people expect some Adobe tools feature. Here you can create channel like you do with but I will show you how it’s, it’s organized this inside a channel, I create my member this you click here for member info. That’s also all the information about your your channel. And then you you can easily open. Each one. And each time this one. There are nothing. Sorry. Does. You have an item here? This 16in this case. You can also add to your favorite. You can see detail That means all the, the some metadata from some from from each smart. Smart file. But it’s the same thing for any item that you have inside your library. So if you go to library, if I go to page. If I can select this, I can view details. And you can see that this page is used in several smart files. You have also if I’m not going too fast, but if you go here you have one sometimes in other place there are one also. But that means, you know where this page is used in some of your smart file. What I can say for my team.
Speaker4: Sorry. Peter you’re.
Frode Hegland: Not muted. That was funny.
Alain Marsily: Let me create one smart file just to understand how it works. This if I go to smart file, I say future.
Speaker4: The forwards, the forward. Award.
Alain Marsily: And if I go there, I can show you that I have several kind of file type at. I’ll take I just select it. This afternoon I drag and drop.
Speaker4: Was too fast. Right. Sorry. It works. Not usually.
Alain Marsily: Those are easy to to add. I can.
Speaker4: This one.
Alain Marsily: And what you can do is you can move this any place if you want to do it on on the view you have a preview and as you move it where you want. If you want to add some. Some text. So this is a comment. Well, sorry that I make a mistake. I can also move this comment where I want. I can add a, I can create title and so on just for making having good structure. And as a viewer, that’s the idea how to work this. The point is you drag and drop all your, your files and you can reorganize them and create structure and you can share this. The idea is if you can download it you can share it and you can share with to a channel with someone. And you can also add limits. If you want the time limit, you can co-edit with some user the C file. I want to, to have this I share I can share share channel share to web. You have a user and if you want to have user that can make additional comment that you can co-create a smartphone and share to other like this, this the idea is to really co-create and you can drag and drop and from your own library as a recipient or a coeditor, and you enrich the smart file with your with your coder. That’s the idea. And what I can show you more. I don’t know why it’s always coming to here. Well, that’s about, let’s say, the philosophy, the, the idea is to have a container that where you add content and you can restructure your content and you share your content, you can have a user adding some, some comment if you want. And the the idea is to give to the user some, let’s say tools that is not explained here but is easy to, to show this if I’m going here edit.
Alain Marsily: You can see that you can create comment, but you can also hide blocks that as an author you can create your content, but you can also hide some part of your content, and you share only the the content that you want to to share with your user. But you can keep some internal notes inside your, your organizer by your authoring tool. What you can also do and I will be that will be the last feature this if I’m here. And I will take something from one another the same time. Go to Nebula Project. I can be here. This does the kind of stuff we exchange with with my my friends. And if I want. That’s why it works here. You can add this content if the user has given access to this, you can like a Spotify. You can take part of their content and create a new smart file. Or you can add the content to one of your smart files. The idea is to to create a kind of playlist. In fact, when we released the first version of of Quincy, it was called playlist. And sometimes for me it’s some difficult to use, but the idea is to have the same philosophy of that Spotify, that you take the content part of the content, you can create a new content and you can share this content with other user. And if you have, keep the link of the content, even if the author or the author change something, it will spread all the the change inside all the the the smart file that you have shared.
Speaker4: What?
Alain Marsily: I don’t know if you have any question but that.
Speaker4: That’s good. I’ve got a bit.
Frode Hegland: Of a cough, but I’d like to start off with. What is the technical side of this? What is the framework of wrapping is a proprietary open. And what’s it based on this?
Alain Marsily: This is create with a normal framework react. But behind it’s graph database. Neo4j’s database with additional component that we have created for to to make it smooth and easy to manage. But we have no idea at this stage, if we have 100,000 users inside the the Quincy, how will be the speed that the kind of stuff that we want, we plan for next month is to stress the database. But the the architecture is, is mainly a graph database.
Speaker4: Are you still there?
Mark Anderson: I think Fred may have stepped away from his thing. I’ve got a quick question. So what’s your notion? This seems to be following notions emergent at the moment of basically blocks. What do you how do you define a block other than technically, in other words, to someone who isn’t a programmer? What what do you describe a block as?
Alain Marsily: Does that mean the kind of text behind its part, when the block is is create its markdown compatible? And but we have found some limitation for some stuff, but we, we, we plan to be even more than markdown. That means at least markdown, but also to, to enrich depending of the version. But the idea is normally if you you copy something from markdown, even copy paste, it will be we keep the formatting of what was done, but we want to improve that also. But yes, it’s markdown.
Mark Anderson: Okay. One other quick one. Well, I’m here is so this this inclusion into so smart so effectively you’re translating contact content in, in the sort of in the Nelson sense if I, if I’m understanding correctly.
Alain Marsily: Yes. That’s right. If I have one of my friend is, is let’s say someone that has worked a lot with some engineering in the US about transclusion and so on inside the graph database. But it’s not as, as a purist. It’s not really transclusion as is, but we want to improve that. But at this stage, we are learning. That means we want to have a foundation and have people like you or your team that help us to make it not only just conceptual, but also to be practical. And that’s that’s the point. Because if you go to and strict to the theory we have found difficulties. This that’s the reason we have spent a lot of time because this interface evolve and we we try to make it and keep it simple. We want to make it powerful, but we know that if we want to make it, let’s say something for knowledge worker. It cannot be just for PhD. It must be someone easy to understand and to play with.
Speaker4: Thank you. What is your.
Frode Hegland: Primary use case talking about?
Speaker4: Haha, that’s.
Frode Hegland: A.
Alain Marsily: Good question. In fact, we have no persona. That’s that’s something that is difficult to explain because we we have something that is really rich. Probably some some say too complex for some usage and, and and so perhaps so powerful for some stuff. But the idea is to let the user share easily, like we do you do with, with transfer. And at least you can do that. If you want to go a little bit further, you can, because if you want to cooperate and if you want to exchange with some members, could be supplier, could be customer, and you can easily manage. And we have several use cases. We have notaries, we have real estate company, we have insurance company that play with, with quidsi. Because it’s so easy for them to structure what they want to share and have also their recipient adding some photos for insurance company or sometimes it’s easy for for some people, for some employee and notaries because when they have a lot of Wetransfer or Dropbox, they have no structure of what they receive, and they spend a lot of time to restructure or to rearrange. And with Quidsi, you can easily let’s say, help the user to understand what you want to, to, to exchange. And that’s, that’s the idea. And we want in the future, let’s say what we will work.
Speaker4: On.
Alain Marsily: Is not only something that the author can reorganize and organize for their recipient, but we want also the recipient to add some comments that for him and to store the the context of what he have received. And the idea is like you do with any paper you receive and you add in your binder, you have some post-it and so on, but something that enrich your original content and that the something that user ask or require. That means not only sending back and make comments to the author or to other members, but also keeping a personal tool for them and, and and storing the, the whole binder or let’s say, a smartphone with all the comments and notes that he has taken and to keep the context. We want to improve also the date of the that let’s say the.
Speaker4: The date.
Alain Marsily: Does that mean everything that you create in it? With timeline, a smart file can be also viewed as a timeline in the future. We. I know you have. I have something that is is good for me to say. I discover a few weeks ago was a tapestry. And I think they have something in term of philosophy are looking for something to, let’s say, sharing content. But in this case, you can do it with other members because you can you can really group or you can cooperate with the user. And we want to, to have this idea of meta player, this trying to preview all the content that you have inside your library and share to your recipient and not having them looking for something that, let’s say, rendering the content. You must do that. Thus, we have some work to do for, to improve and to, to add more file format. But the idea is to be beyond a PDF and having a video embed file and everything, and also comments and that, that will enrich what you share.
Speaker4: I have a question? Sure.
Frode Hegland: You first.
Leon van Kammen: Of all, thank you very much for this presentation. I think it it looks very beautiful. And it also the, the interface and it looks like you said Spotify. It gives me Spotify feelings. Spotify meets g suit and and Dropbox or G drive. It’s a it’s a nice mix of things. You you mentioned notaries. And that made me think because I, I’m really sensing a very fluid fluid document ecosystem here where documents automatically update themselves based on these blocks. And I was just wondering you know, what do there must be something. What notaries or lawyers don’t like about fluid documents. And there must be something which they love about fluid documents. And I was wondering if you had some insights on that, in which cases fluidity is great and in, in which cases, fluid fluidity is not so great.
Alain Marsily: When we listen to the user they ask for having a lock. That means you can see the here you can lock and unlock a. That means if you share a document, it cannot be download. But, you know, screenshot can be done anyway. But that means the document cannot be modified by the recipient. If you don’t lock the document, the recipient can. Make his document for him, but he can he can also fork the document to make it really for him. Or he can keep the relation with the author. This it’s part of the decision of the recipient to say, okay, if I have access, do I have to keep all the the legacy or even perhaps the the relation with the author? Or do I take this document because I will re reinvent it or I will take some snippets and recreate the content. Notaries, in fact, they don’t play with deep feature with what they use is the way to organize when they receive some stuff from real estate company. They, they, they can see the order with comments about document that was shared to them. And if they want to share with some of their user and not all of them are using them because it’s a beta version, because they are scared to use a beta version they just test with some, some user what the love is, that they can share something that is not a folder with comments inside their Gmail or their email where something, all the content are not fully structured.
Alain Marsily: This what they love is the way to the user has to read it and also to mix excel file or PDF file or word file if it’s necessary. And the user can also download the word file and, and also play with it. We can perhaps to integrate really office inside Quidsi. But it’s not our focus right now. But the idea the, the user, the recipient can download, modify and, and make a drag and drop back to, to the document to, to to the, the notaries. This we are listening. We are. But it’s not really deployed in, let’s say, real work environment fully. That means some of their user beta tester also. But we are listening and what we have a request is and and encryption. That’s some of the user expect for any reason. They don’t want to have some of their content inside the cloud and they are scared. But we have designed to to let let the user create a local instance in the future. That’s for company, for university, but also for, for user. It’s not yet an application. That’s if when Quincy will be an app, it will be probably user also easy to create a local DB for for the user.
Frode Hegland: Thank you very much.
Speaker4: Thank you very much. My pleasure.
Frode Hegland: You as well. Can you unshare your screen, please?
Speaker4: Sure. What? There we are.
Frode Hegland: Yeah, it’s definitely worth thinking about. Hang on. The screen got very reorganized. There you are. About how we put our information together, how we share it. It’s been interesting. I learned the first time I think I saw your work was 3 or 4 years ago now, depending on how we count. It’s good to it’s good to keep the dialog going for sure.
Frode Hegland: So today in general, what’s the consensus in consensus in the group? Do you want to talk about XR text or other text or any specific feelings? Otherwise we’re going to talk about metadata. Gosh darn it.
Speaker4: Okay, I want.
Frode Hegland: To talk about HTML. Then for a second because in in my personal world, there’s Sloan and there’s the thesis and there’s same issues but very different aspects. So it’s, you know, messing with my head. But I wrote it a little note on encodings and it’s interesting and I really want to hear what you guys think. It’s interesting to think about HTML as a format whereby everything is written in the document, dot HTML, everything is there. Of course, you bring in images and stuff. That’s not what I mean. However, it is explicitly written for the reader to render it in a specific way. That’s a really powerful thing.
Speaker4: If you look.
Frode Hegland: At an ATF, the ATF is very simple. This is bold and whatever. If you look at PDF, we know what a mess that is. At least it presents it visually, superficially, cleanly. But HTML is designed not to be read as is. Isn’t that a phenomenally interesting philosophical position? It’s defined to be rendered even if you have the raw file. You know, the browser wars would have very different ways to render different things. But what we’re living with today are, you know, over a decade, decades of cultural agreement and how certain CSS and other stylings to be presented, isn’t it? Oh, please. I’m not sure who is the sound in the background, but please mute if you’re not talking. And Leon, please unmute because you will be talking.
Leon van Kammen: Yeah, this is a topic to my heart. I also noticed this weird jump between plain text RDF rich text file and suddenly HTML. Well, this is of course my small little brain. And how I experienced experienced this jump. Suddenly we went from readable text to sort of unreadable text surrounded by markup and magic symbols. So in a way, this even RDF was still what you see is what you get. Sort of like you can see it, you can read it, you can type it. But the moment we went to HTML, which was also great, don’t get me wrong. We we took a pretty, pretty big step, and I, I noticed this when I started to explore visual meta which is based on BibTeX and the whole philosophy behind BibTeX. And again, the same difference between, for example, BibTeX and JSON is that BibTeX, you can sort of write, I can write it on a, on a napkin, but I can’t really it’s not really comfortable to to go beyond that. Like the moment you start to write notes in JSON to people, then things get go wrong pretty quickly. When you think about Passability and all that stuff. So yeah, I think this was a very quick transition in, in just a couple of years. But I have a feeling that in those couple of years this step was so big that there might be a lot of stuff to discover inside that, that step, so to speak.
Speaker4: Don’t.
Frode Hegland: Yeah, that makes a lot of sense. Randall.
Brandel Zachernuk: Beyond merely being a document that defines its presentation, which is the case for a PDF as well as, to a lesser extent, RDF as well, and things modern HTML is supposed to be responsive in that it’s composed in a way that means that it has the ability to live on different screens of different sizes and across a range of capabilities. So we have media queries. They are primarily to say, if you don’t, if you have a monitor or a screen display above a certain perceptual size, then do this. And if you if you have it below a certain size, then do that. And for the most part, people just use that to say this is what it should look like on the phone. This is what it should look like on laptop, on Mac and PC. But it’s not unreasonable to also have other targets on there. Most people don’t have a very good reason to have web pages run on an Apple Watch, but you can and you can use CSS media queries to make sure that to whatever extent there is information that is relevant to display there, you can do that and you can have, if not the exact content, then the sort of the conceptually correct experience of that content in there.
Brandel Zachernuk: But, you know, the same should and will continue to be the case in visionOS. So if there are things that visionOS can do, not that it is a visionOS device, but that something that, that visionOS or, or, you know, one of these or one of these can do this is quest three vision. So then tell the user, tell the document, and then we’ll do whatever we want with it. So, you know, I am building the HTML model element. There are special characteristics of the model element that are only available and present on on visionOS today. Hopefully quest will follow up. And and so it’s not just the sort of there’s a document that tells you how to display. There’s a document that tells you how the display should flex to the range of surfaces and the configuration of of capabilities and preferences that a user has. And I will grant Leon that, that that means that it can come at a cost of, of legibility to a certain extent. But I would also say that that’s at some point level, a function of the the authorship of the document. If you actually delete a set, the CSS file from Apple.com it turns into a surprisingly readable page. You just delete it, it straight up.
Brandel Zachernuk: And that’s because the styling is the, the actual styling is rendered more complex for the benefit of making sure that the document reads correctly for accessibility and for for sort of performance and, and legibility and, and a range of different contexts. But people have broken that very, very terribly over the years. So I, you know, it’s that’s, that’s a challenge that like, an ongoing challenge is how do you how do you make something that is is still a document, it’s still a text document, but has the ability to to look the the best for the capabilities that are put on offer. But yeah, like, I just wanted to add that this, this flexibility is, is an intrinsic component of it. They didn’t come at the ground floor CSS, the cascading style sheet sort of model. And beyond that into media queries for the sort of flexible constraint based display. They came in like 2020 ten, respectively. But, you know, if and where we have to add things to the idea of, of a document and of styling which we almost certainly will have to in order to make these things spatial in the future. I believe that people will let me. So hopefully you can feel confident in that.
Frode Hegland: Thank you. Brundle. Mark.
Mark Anderson: Yeah, I think I’m with something like HTML, part of the other part of the sort of unseen part of it is the semantics it puts in. Which some of which is used to some which is obviously used to help with the with the rendering, but also just to give some prompts as to as to what’s in there. So it might help with the way the remediation is done if you’re putting it into a different format. I think the interesting transitional phase we’re at is that Where we’re sort of trying to make sense of the post paper document. In two different ways. One is you, which I was thinking as I was listening to Anna’s presentation, because of course, we can make all these wonderful fluid documents and one of the or presentational systems. And one of the challenges we then have is, well, at what point how do I fix that without losing all the flexibility that’s in it? So it works wonderfully well. It’s all flexible and a challenge often comes with these things is or even even with something like a Google doc, how you Record the important sort of waypoints as something more than just this edit on a continuum, because at the moment, too many things fall back to programing sort of nostrums that don’t actually make sense to pretty much anyone but a programmer. And that’s problematic at the moment. And that’s why a lot of tools, I think, struggle, because the wider user base don’t have any notion of forks and edits and all this kind of thing. They’re more used to they’re they’re probably still swimming in the world of documents, as you would think of in a paper term. So there’s no no error in that, I think. I think it’s quite normal that we’re going to see this as we, as, as we transition through. But I do think having documents That records some of the structure are incredibly useful. If you don’t if you if you accept. Rather, Brandon was brought up. You don’t necessarily know where the render surface environment is going to be.
Frode Hegland: I thank you, Mark. Any instant. Look, guys, just to make it simple, if anybody has an instant response to something someone says, you know, just say so. If it’s if it’s brief, obviously. Peter.
Peter Wasilko: Okay, before I leave, I just want to mention that it seems like we’re getting a lot of interchangeable file formats now. We’re having a proliferation of it. You have RDF, you have people who are using word, then you throw it through Pandoc. Pandoc turns that into markdown. You fiddle with the markdown throw it back through Pandoc, turn it into LaTeX source. Then you run that through the typesetting system to get a final PDF. If you really care about Microtypography and you want that real fine grained control. So. Well, I think we should start to think about is maybe a meta document which consists of a whole bunch of potentially interchangeable formats. You can use the mathematics of lenses, which are bidirectional filters, so that you can edit in any one of the formats and automatically have that change get mirrored in all of the other formats. And on that note mom’s ready for me to take her out for the day now, so I’m going to have to drop off. See you on Wednesday.
Speaker4: Thank you. Peter.
Peter Wasilko: Bye bye.
Frode Hegland: So just really briefly on what Peter said. You know, one of the ideas discussed, of course, is to have an entire HTML version of a document encoded in visual media in a PDF. I know it sounds horrendous, but or at least the interactive elements should be able to. Visual meta doesn’t have to be BibTeX, you know, if they’re specific things. So then we have to ask ourselves what should be interactive? Fabian. Sorry.
Fabien Bentou: Yeah. Another thing the I think my, you know, my bias for HTML or at least the web. So I do think yes, it comes at a cost, but in terms of responsiveness, it and that also is not free like in order to make responsive content. Yeah. It is challenging to, to work from yeah. Any kind of headset or laptop or phone or the form factors and to keep it accessible. For example if you cannot even read then you must have the content to be semantically annotated. I mean, it’s a proper challenge. And to also show the the size of the challenge. And I already discussed with Leon about this a bit, but I’ll, I’ll share my screen briefly. If I find the button. Yes. Please let me know when you can see it.
Frode Hegland: It says you’ve started. It hasn’t come up yet. There it is.
Fabien Bentou: Yeah. Super. Thank you. So I, I have another headset, the lungs. I think I maybe mentioned it a couple of times, and unfortunately, the browser in it doesn’t support hand tracking. There was a bug. Basically, it got partly fixed, but nobody made the build. Basically the browser for the the lungs headset. So I had to do it myself, which is actually not trivial. So just spoiler alert, it works. It’s not perfect, but basically now I can manipulate content with with my hands. But why do I mention all this? Is because it took about 50GB on my hard drive. Not any video. Like there was no video. Just code basically. So it shows a little bit. So that’s the actual outcome. I have controllers as virtual hands, so it works. And I describe a bit the setup, but building a browser. So that’s how do you say that’s an immersive browser? But in practice, it’s built, I would say I don’t remember, but 98% at least is the normal browser. So based here on gecko for Firefox, there are switching also to chromium the same way Chrome and Edge and other browser are being built. And it’s huge. I mean, 50GB of source code of text, basically it’s gigantic. Most of the time when you build something, you have maybe 500MB of source.
Fabien Bentou: And where does all this come from? It’s the JavaScript engine. It’s the CSS rules. It’s and so it shows a little bit also to get all this responsiveness. And okay, yes, there is more, there is some networking etc., but it’s, it’s, it’s the complexity of the stack is huge. And most people can’t even build like such a thing. It’s, it’s yeah, it’s both scary and exciting. And that’s also why to be honest. And I mentioned it also to Leon before it’s I was happy to be able to build that browser just because it feels like it is still a bit of a black box in term of complexity, even though it’s open source. But it feels still, despite all this, like, tractable. So that if at some point I want to change some part of the browser itself even though today I don’t have the knowledge for it, I know that it can be done. It’s not something that just sits there and you have to accept the way it is. So it’s also, even if once want to change the browser to support something let’s say visual meta or another format or. And that person is not Apple or Google or meta. It is still so despite the complexity feasible.
Frode Hegland: That’s just very, very Brandel.
Speaker4: A ride a.
Frode Hegland: Browser. You guys are just geniuses beyond compare. Both of you. That’s. So you wrote a browser, Fabian, for the headset? That’s. Yeah, that’s kind of amazing.
Fabien Bentou: I did not, I did not, I just built it. Oh, sorry. Built it in the technical sense of somebody wrote it, but between the step of going from that browser as text to an executable that the device can run it it’s a series of steps and I had to abstract away, like, do I mean, you don’t want to know the detail, so I don’t want to say it’s trivial, but I also did not write it. What I what I did modify it a tiny, tiny bit, like the bare minimum to show that, hey, I’ve wrote it or I built it and modified it rather. But yeah, it’s it’s the it’s as if somebody took, I don’t know, a book and then they wrote their name on the back page and then they shared it. That’s that’s the amount of work in term of creativity or let’s say that’s but then it’s as if I printed the whole thing. I don’t know, it’s a little bit analogous or limited, but I did not write it. I will put the reference there in the, in the chat, which does include the, the source code of the browser itself and the, the people who wrote it. Also, like I was saying, they wrote maybe 2% of it, like 19% is from gecko, from Firefox. And then people who built Gecko and Firefox, it’s I don’t know how many dozens, hundreds of people over decades it’s it’s a huge it’s a huge endeavor. So I did not write it. That being said, also, as I mentioned, I think here before and to Leon, there is something called the browser book which I browse through rather than read properly, which specifically is like, can you actually write a browser? And it’s there were the argument. There is, it’s not that complicated. If you don’t bother with a huge amount of how do you say Yeah. If you keep it minimalist, I’ll say if you just do the bare essential, then yeah. Students can write browsers too, if they. If they want.
Speaker4: Fabian.
Frode Hegland: You sound like one of those people who is just served a really, really nice dinner and we say thank you for cooking. And you go into half an hour of how some of the ingredients were pre-made. It’s still an achievement. And, you know, there are many ways to argue this up and down. Hardly anybody writes assembly language anymore, so. Fair enough. So and here’s the thing that I just want to bang on a little bit and see what the reaction is. So and this comes back to my core value, of course, which is how to deal with the data slash metadata. In most systems, the thing you put into something determines what you get out of it. And I think that’s quite limited. You know, HyperCard was great, but if you made a button on the left side with this view, the viewer would get a button on the left side with that view. But what I’m trying to do with the approach of visual meta is say, here are stuff you’re reading. Software should be able to choose what to do with it. I happen to have an instantiation to prove the concept in reader, of course, but I’m wondering, especially with HTML as a model, if we can think more about very, very open and discoverable ways to put more into our work that different systems can can use. I mean I would like author to be able to export documents and have someone else say reader is absolute rubbish. Considering the metadata, I should do this, that and the other, that would be useful considering that web browsers today, you know, they all competed on trying to be the same in a sense. You know, I don’t know of any web browsers that are look, pages are displayed very differently on this web browser. We know what the underlying is, but the user community needs something else. So reader.
Speaker4: Mode is.
Brandel Zachernuk: Reader mode does not exist on Chrome. Reader mode.
Frode Hegland: Yeah, I’m familiar with reader mode, and that is useful. But when you’re in reader mode, can you? I’m just hang on. I’m going to click on it now. Okay. So I’m going to. It’s been a while. How do I enter reader mode?
Brandel Zachernuk: You click on the the AA on the left side of a page that is document documenting. So if you go to something on NY times or something, then generally you’ll have articles that look look document, document like enough that it’ll allow you to enter reader mode view.
Mark Anderson: It’s the left hand end of the navigation bar in Safari. So just to the left of the of the URL.
Speaker4: Not for me.
Frode Hegland: No. I’m in addition that CNN. Looking at an article. New York Times. Let’s have a look. Supreme Court rules Trump can remain on the ballot. Oh, yeah, there’s an icon there.
Speaker4: A day of.
Brandel Zachernuk: They have broken it. So news one of the challenges which is sort of central to this. So I am on that same article and it is not working for me, and it’s likely because it is in an iframe. So one of the challenges is that, you know, news articles, news websites don’t want you to use reader mode because it ruins their ad revenue. So I’m curious, I’ll file a radar on this, but let’s find another one, then go news.google.com. One of the things that reader mode does is when it identifies, let’s see if the CNN article, it’s better for it.
Speaker4: That’s better.
Brandel Zachernuk: That’s not as reader moody as I would have expected. Or maybe so. I use an internal version of macOS with internal version of Safari that not for any cool kids club reason, but because I’m responsible for the development of it. But reader mode is supposed to make a single column layout that has none of the advertising and has all of the figures and illustrations kind of displayed inline, as close as they can kind of measure, can kind of anticipate would be appropriate for it. And that’s because sort of similar to what I was saying about responsive design. And the fact that there’s these well-established norms for being able to build, you know, Mac versus iPhone and, you know, Android and Windows and respectively as well. The. It has been by turns understood that you don’t get to decide how things are going to display on people’s devices. And so you’ve got to make decisions about what’s appropriate to do on that basis. You know, so you really like the sort of surveillance advertising models that we have today where they need to ensure that the ad was on screen. And, you know, if they had their way, they would want to make sure that there were eyes on the screen, eyeballs on it, that kind of thing.
Brandel Zachernuk: Is somewhat at odds with the infrastructure of the web. And so there are there are deep challenges there. But yeah, so like, but the basic premise of, of a web service like that is that they should be relatively open. And so that’s where scraping, you know, becomes not just possible, but very, very difficult for people to, to remove the capacity for, not that that it stops people from trying. Scraping is the process of having an automated browser a script based thing that is sort of pretending enough to look like it’s a real thing, to be able to make inferences and determinations about web page content. And we like some scrapers, we call them crawlers. They’re the things that are responsible for making search engines work, work much as they do today. But we don’t want scrapers that are sort of pulling all of the data off of Reddit unauthorized in order to provide a large language model. We only want to do that to paying customers. So not not we. I’m not representing Apple. That’s not that I don’t that’s the joke.
Brandel Zachernuk: But it’s but it’s it’s also true. So that’s a sad one. But yeah. So like that. That is a sort of a balance of capabilities and concerns that sort of come out of the substrate, the, the metaphors of the web. What’s funny is that while we sort of understand at some level the what the web is, because we’ve all spent so many years, some of us decades on it now. The underpinnings of it are relatively elusive for a proper intellectual understanding of it and, and what the practical consequences of those are. It’s not simply that people oh, they don’t understand what a car is like. There’s very little benefit you get while driving a car to know exactly what a carburetor is doing, that kind of thing. But when you know what the web is, then you can envision some genuine alternatives and opportunities as a consequence. So thank you for entertaining the diatribe.
Mark Anderson: Well, Fred stepped away. I it’s interesting that Well, I’ll pick up a point he made which was you know, we’re talking about the visual metaphor. I mean, in a way, you’re talking about a visual sort of CS, because essentially CS was, and I remember it when it first arrived. It was marvelous. You won’t have to write font tags everywhere. And it, you know, it was it was something that allowed some, some flexibility. But the point that made me stick my hand up earlier was I was thinking here in discussion of sort of, you know, things like rich text. I’m thinking how little time we’ve actually had it because rich text sort of happened in my working life, I think. I mean, before you used to type stuff, you went to the typing pool and they underlined it, or you were sent off the typesetter so that we have almost a sense of so, so, you know, several decades of working with word have perverted our sense of our understanding of document and typesetting and stuff. And that’s not a jibe against against sort of real design and typesetting as in a nice book, but just the quotidian stuff that we push out. I find I actually, I, someone asked me to write my own word the other day. I just gave up. I just had to use something else. It was so horrible, flashing things on and off and telling me I was doing the wrong thing wouldn’t let me write.
Frode Hegland: Author.
Mark Anderson: No, no no, I find I have trouble with author, so I actually wrote it in well, I wrote it in Bbedit and in tinderbox. Because I just find it easier to write in plain text and use HTML, because then I control the markup. And if I wanted to be bold, I’ll put any bold tags in. Thank you very much. I mean, because a lot of things I find, funnily enough, I find myself in terms of trying to write meaningful documents, moving away from Wysiwyg, because most of the time I’m writing for somebody else. So my my intent is clarity for the reader, not not what I want as a designer, in a sense. I’m not really interested. I’m not invested in the design. I’m interested in the clarity to the reader. And the readers are often quite different. So I’m I’m.
Speaker4: Contradictory.
Mark Anderson: My readers. And that’s been an interesting thing. That’s just come over time in practice. You know, I didn’t think about it, but I reflect on it now. That’s that’s definitely part of my my writing approach, and which is why I’ve gone back to much more bare bones writing. Because I write the text. I mean to write. I write minimal styling text, and one of the most useful ways of doing it is there are various ways, but they’re all sort of along the lines of marked up text. So I don’t really want to write in a rich in a rich text environment, because I just find it gets in the way now.
Speaker4: On that point.
Frode Hegland: Mark having bold text. B bold italic text b italic. How does that get in the way for you?
Speaker4: Okay.
Mark Anderson: Well, it’s it’s what it was when I was writing Werderm. It kept wanting to to had all sorts of bizarre notions of what was good grammar. Well, maybe not in this, not in the country I’m in, but, you know, so there are all sorts of defaults were getting in the way and it wanted to change things. And it was.
Speaker4: I’m not I’m not.
Frode Hegland: Talking about the.
Mark Anderson: Wanting to restyle. And I was thinking, I don’t use a lot of I mean, I don’t tend to use a lot of styling except at an output. I don’t really think about the styling until I want to go and output the document. I tend to use the, you know, it goes back to the I think back to the early HTML markup when we had emphasis was bold, bold, but but italic took over from from the Em tag, which I think was the intent because the idea is that you were using this as semantic meaning, so you were emphasizing something to make the point that it just, you know, it needed to be clearer to the reader than the main text. And somewhere along the way of writing in Wysiwyg editors, I think we’ve rather lost sight of that. And it’s it. Just become a push button. It’s become a sort of it’s become a push button. Laziness for anything. Okay. No, no, this this actually, no, this needs to stand out or this needs to be in a slightly different typeface because I need to I need to indicate that this is a piece of actual code that you type or, you know, this particular character absolutely needs to be rendered like this and not like something else. But otherwise I find most of his fluff.
Frode Hegland: But so for instance, in author, what would be the thing that would get in the way of you just writing?
Mark Anderson: Oh. The only reason actually I struggle with author is just because it’s very it’s obviously very attuned to you. I just find all the, the way I’m used to doing things is, is, is so things like adding a link, I want to write something and then turn it into a link, but the application wants me to say I’m going to write a link and do it. It’s very minor things, but it just creates massive friction.
Frode Hegland: Mark, if you write a link and author, it becomes a link.
Speaker4: Yeah.
Mark Anderson: Yes, but I don’t I don’t like that. I want to decide. I actually want to do that because I use, I tend to write quite complicated sort of technical English where these, these affordances, if I’m writing, you know, letters of counsel or something are fine. If I’m writing very deliberate stuff, I don’t actually want the editor trying to second guess what I mean, because I spend my entire life going back, having to undo it. So and critique is.
Speaker4: That.
Mark Anderson: It’s not a critique of the tool, it’s what you’re trying to write. And I would admit that the writing I tend to do is probably complex in its in, in, in, it tends to have a lot of quite structured language that needs to be carefully nurtured in it, which is admittedly not what probably most people are writing.
Frode Hegland: So do you mean that you find it annoying that it underlines a link? Is that what you mean?
Mark Anderson: No, I it’s hard to say. I just find it hard to use because it doesn’t work. How? I sort of expect a a rich text writing tool to work. And that’s not a critique, it’s just that it doesn’t work how I expect it to work. I can’t put it more plainly than that. I mean, I haven’t kept lots of bullet points or something, but I just find after sort of 15 minutes I just have to go and use something else, and then it’s easy, and then I just copy and paste it back, and if necessary, then I’ll go through and redo the formatting, which is quicker for me to do.
Frode Hegland: I understand that I just find it fascinating and interesting because when you write an author, you cannot change the font. So a lot of those things and it doesn’t try to format things for you. So I found it interesting that you found that it does too much. And the.
Speaker4: Example.
Mark Anderson: Things okay. So the thing is, I find it I find it really, really hard. The white insets, the insets, the headings because it just having having the body text sit to the left of. I just find it odd. So I’d rather not see it because it just, it disturbs my flow when I’m writing. So it’s much easier to not have all this styling on because it’s not actually helpful to the work, if you see what I mean to the to the writing you’re doing. Because the the intent is to be is to have your mind in, in what you’re doing. If you’re fighting the formatting the whole time. Because whether it it doesn’t matter why, but if it’s, it’s drawing your mind away from the actual act of writing meaningful text for another person, which is which is why I find myself going back to writing the plainest sort of text which which wouldn’t be my expectation, say, ten years ago.
Frode Hegland: Yeah, no. That’s interesting. I know you don’t want the indenting, so we’ve actually removed that on export. You can turn it on and off depending on what you want. But it’s very interesting how we view formatting differently. So you, you feel that in Bbedit or something if you write like this is bold and then you write the word and then bold, it has less friction for you than making the word.
Speaker4: Bold.
Frode Hegland: So now that’s.
Speaker4: Often. I won’t add.
Mark Anderson: That till the end anyway. I mean, as time goes by, it’s almost easier to go and go and actually gloss the style on when you’re done the Wysiwyg method. It has taught us that what we do is we go along and we make it look pretty as we go, which is actually an incredibly inefficient way, I think, of actually.
Speaker4: Writing this.
Frode Hegland: Is this is a very important point in our community, because you’re using the word pretty and you’re using the word Wysiwyg. Wysiwyg, of course, was applicable to printing. But I would say that what we have now, not Wysiwyg, it’s what you see is what it is. And I was a, we could call it rich text. That would be more appropriate because
Mark Anderson: But it’s the same. It’s the same thing. I mean, in the sense that.
Frode Hegland: It’s not it’s not the same thing in the sense that when you are writing some things, you just want to write as a big block of text. Absolutely. And sometimes you want to change it for yourself or the reader in some way. But that’s not for the experts. But there is, you know, Bob Horne and James and Jim, they talk about really wanting a huge amount of formatting, which goes even further. And I want, you know, like a huge thing so that it’s, it’s a it’s an important range of. Mentalities we need to consider. Randle, I think you had your hand up a long time ago.
Brandel Zachernuk: Yeah, I was down because I wasn’t sure I. I didn’t totally lose the thread, but what I was going to say there was about the ability to simply not display some things depending on the capabilities of a device. So in the same way that visual matter has the ability to be printed. There are visual queries there, sorry, media queries for print, technically. So that you can have a fundamentally different document which, you know, the web is that it should be philosophically the same document. But if you have a particular point of view about what you want to be committed to physical and durable media, then you could print out your meta information, your, your visual meta within that and merely use it within the context of a live rendered browser. So that that I think is interesting. I mean to weigh in between the two of you on the other subject of, you know, what is formatting? I do a lot of stuff through bullet points and, and outline hierarchies. And so I do need to have type styles of heading level one, heading level two and tabbed in bullet orders in order to make sure that I am covering points, you know, so at some point, hopefully we will get the release notes for Safari 17.4 out.
Brandel Zachernuk: They might be out today. I’m not sure. But we also have a more thoroughgoing, like explanation of what transient track point, end point is that we need to have for webXR. And you know what? My writing is best when I sort of cover the bases, when I figure out what what has to go into each piece, and then I will flesh it out. So to that end, it’s useful to have formatting, but at the same time, that’s actually kind of a false work. It’s something that goes up over a paragraph, but is not constitutive of any of the parts of the paragraph. And so there are some interesting challenges and tensions there between what I want in terms of having formatting for the purposes and benefit of being able to kind of enunciate the the scale and hierarchy of that stuff. But at the same time, I’m somewhat in March camp in that I don’t want formatting on the prose that gets kind of constructed afterward. And if anything, that actually makes my ideal, ideal text editing platform even more complicated because I, I want to have text that I don’t want. So I don’t know where that puts us.
Frode Hegland: Well, many things that we keep going back to is different levels of reading something and also different level of authoring. You know, you might author for someone young, someone in your field and all of that stuff. But of course that takes work and it’s not often done. But if we look at a simple example of references now, I don’t know why references are handled so badly by publishers. It may be that they don’t want you to look at other documents. I don’t know what it is, but if we just take the notion of a reference section in either HTML or a PDF, specifically PDF currently is essentially dumb data. It isn’t something that can easily, automatically be parsed. Right. So I think it’s a good example to look at the notion of including metadata that was there in the manuscript anyway, and then have the reader software decide what to do with it. Because otherwise it can get a bit.
Frode Hegland: Big or. Yeah. So for instance, when you, when you download a PDF document and you’re reading it two things. Number one, you can’t easily programmatically see what the sources are. But also, as we’ve discussed in our Wednesday meetings, it might be useful to include more information than that’s what is available in a normal academic document, for instance, the link type. Why is this being referenced? Maybe a note from the author saying this is particularly interesting. And so on. So when I’m talking about having stuff in the work and the different software reads it in different ways, that’s kind of what I mean. But that kind of thing. Yeah.
Mark Anderson: Mark, there’s an interesting just just your last point about the the thing about linked types had an interesting conversation, actually with Dave Maillard the other day, which led us back into the subject. But he made the point quite fairly, said, well, arguably the context of the references is made by the is made by the, the the context of the writing. And I thought, well, yes and no, because I think as we as we’re showing in the experiments we’re doing with the XR and things, there’s a there’s a bit more to it than that. It’s probably true. Well it’s, well it’s in the, in the realm of, of paper or something that’s pretending to be paper, but certainly in a, in a, in a sense in a plastic phone call it that environment. In something like XR, there’s no requirement to stick within those conventions. And so actually having the information is useful. Problematic though, is that we we’ll say something like, well, the author will will indicate what their intent is. But you know, we’ve we’ve been through 20 or 30 years of this whole debate of between the author’s view and the reader’s view. So it’s of course, it’s it turns out to be a little more complicated. I the reality, I think, is somewhere in the middle that rather than forcing everything to have necessarily extra information I think if the tools or the methods we, we use allow for glossing where it, where it might be useful to the reader. Makes sense. Otherwise you fall into the the bear trap of having a fixed system and wasting time trying to work out, you know, which which which box you’re going to tick before you’re allowed to enter this, this bit of text into the, into the flow.
Speaker4: Yeah. I mean.
Frode Hegland: What I’m talking about is including that mark, because there’s no reason why a document, a manuscript cannot find the sentence that referred to the citation and add it to the citation at the back of the document. You know, we have looked at ways of being clicking on the references, the number, and it shows where it is in the document. But what I’m really trying to advocate here and get more dialog on is the notion of quite simply stuffing as much as you can possibly can into an extremely open way in a document and then having different ways to read it. You know, if this is PDF with visual meta or if it’s HTML or whatever, I don’t really care that much. But I do care about things being very closed. I mean, we had just the fact that we had image maps and they’re all gone. You know, the average user cannot make an image map anymore. That’s absolutely craziness. Right. And it was we.
Brandel Zachernuk: Well, so an image map was a was a solution to a problem. We we can do SVG. And arguably SVG is doing a better job of conveying the content. So there’s no.
Frode Hegland: Authoring systems for it.
Speaker4: There’s Inkscape. Yes.
Brandel Zachernuk: Illustrator.
Speaker4: Yeah I mean how.
Frode Hegland: Much do you spend per month on illustrator? Right.
Alain Marsily: If you play with Figma you can have also SVG working for that. And that’s in our mind we think that could be a figma for document because everything is also managed with SVG if we want. And on my side I will we, I prefer to be focused on how to manipulate and how to handle text and formatting myself and help and being helped by or sustained by some intelligent to help me to format today we have you have to format and then express and I think in the future that will be a more free form where it will be easy to create some snippet and some text and so on, and ask the intelligence to reformat your text and make suggestions. And I think SVG will help to also create big titles and you just expand your titles. And if you want to be really coded with your own pattern, you will be able to use kind of kind of fonts and kind of format. And you can have some, let’s say, something personal, personalized, and you, you, you play with that. And if you want to increase the tool will help you to go to other kind of format. But you can also play with your own format. And if you if you try to write something, you know, when you start with a word processor, you have to start on the beginning. That’s strange. You have to play to quit, to to directly type where you want in the page. You have to define the size of your page the way you want to, to express your your ID.
Alain Marsily: And then if you want to make a good. With only a stuff. The intelligence must reformat for you, and we have to create some template to reverse and not on the beginning on. That’s the kind of thing I will be probably helpful for the the author, but also for the the the recipient, for the reader. I think it must. It can receive some text and some content and organize the way the author has done. But after he can remix the way he wants to to view and read and keep the information. And I think for me format cannot be a constraint. Any, any format must be free form and help the user to create the way he wants the document and the want to keep the document. But format must cannot be for me a constraint. And if you look for some kind of format it is difficult. But as I think it will be, could be a good path for manipulating objects. And I think document is more object than just type of, of, of let’s say type of, of object is it, it must be a general and let the user play the way you want. And after he can, if he wants to convert, okay. Will convert. I don’t know if it really will explain my ID, but I think going to a format would be for more constraint, but going to the way the user wants to have the document shared, or the way he wants to keep the document as a reader and after if he wants to go to the original document, he has the two version.
Alain Marsily: And also for me just to to explain that the component must be far more stringent than today. You know, when you share a document it’s sometimes some document has a connection with other documents, and document can be content. And if you want to select one part of the document or one part of the content, and you explain the meta and also the origin of any content, I think you can must be able to bring not only the text, but also some connection that are behind some some content. And we we always think about part of and snippet and something that is part of a of a whole. But I think the value of, of text, value of content and value of context means that you must keep some link between some, some item of, of content. And when you want to manipulate something, you must make some link with your your real content, your own library, but also with some connection you have create, even if for you or from from the author. There’s some. It’s not just. It’s a computing. It’s. It must be intelligent behind any kind of content. That’s, I think for me, the future of creating document sharing document and let the user manipulate the content without being constrained by the format. Yes. It’s semantic.
Speaker4: Yeah.
Frode Hegland: I also noting Mark’s comment here about esthetics. I don’t see it as an esthetic, what our discussion is at all. And I’m not going to try to sell you on offer, but to tell you a design thing of author is that you, the user, can you can set the typeface for the headings and for the body text, and that will apply on that computer for any document you open. So the user controls the esthetic for themselves. Absolutely. But that’s all it is, you know? You can’t do any esthetics while you’re writing. You can bolt to highlight. You can italic. That’s about it. There’s hardly anything else you can control because I agree that it can very, very quickly become an issue of just fiddling with fonts, as we put it, you know, moving it over there, moving it over there. So I completely agree with that. So while you’re writing, you don’t want to just write one continuous, massive block of text. Of course you will want to have different breaks. And of course there are many different ways to do that. But I do think it relates to the author’s semantic intents of how to go through the documents. And as you’re talking about Ellen with the, with the linking, you know, that’s a huge topic here of linking both with web links and also the names of documents.
Frode Hegland: I find having the names be very useful, but a lot of this comes down to. You know, I’m author and reader are quite successful. You know, I get emails from people saying they want this and they’re happy and sales are fine, but it doesn’t suit everybody. As we just looked at, you know, with Mark and me, and that’s absolutely fine. But we really need, I think, as a community to work towards a way where. Any one of us can really write a writing system. That anyone else can really write a reading system, and there is enough richness and robustness and open standardness in the thing what we write. And it’ll just work. You know, maybe HTML could work for this as well. You know, unfortunately, it’s not really used in academia. Now, that’s a bit of a stumbling block, but it is bizarre. And when we look at things like epub, it doesn’t seem to be rich enough to to have the answer. But at least you know, from the stuff we’re talking about on Wednesdays, at least we should be able to see what a document connects to and what connects to it. Right?
Speaker4: Yeah.
Brandel Zachernuk: But so I think at this point, for better or worse trivially, like if somebody has the ability to write a reader or a or a or a writing app by definition, effectively it will be able to be done in an HTML plus, you know, some, some non-trivial quantity of JavaScript that’s just like the, the feature parity that’s available for it means that anything you can do, unless there are really clear sort of system limitations and capabilities that you must lean on, like being able to to trigger multiple cameras simultaneously, that kind of thing. Then you should be able to do it over the web. And if and moreover, like if there are any requirements to do that kind of thing for any enduring sort of duration for people, then it’s not impossible for those to be sort of jammed into the web to respond to demand. Go ahead.
Speaker4: Yeah, we.
Frode Hegland: Can kind of do that. But with the example of a reference section that doesn’t seem to be standards, even in HTML that says this is a reference section.
Speaker4: No, no.
Brandel Zachernuk: And so yeah, exactly. Like unfortunately that is that why I said trivial is because that’s neither here nor there with regard to what that thing is or what sort of priorities philosophically, that web based, that tool, web based or not, actually has for for better or worse, the web is such a flexible platform that it doesn’t really provide an intrinsic sort of necessitation of certain priorities at this point. You know, Google has built innumerable like very clever but fundamentally web breaking ways of doing the web, the Google Web Toolkit being one of the most interesting and important. But, you know, I, I’m not a fan of react for some fairly significant reasons in that regard as well. Like it is not the web, it’s something else that is built on top of it. And there are benefits, but there are costs. So. Yeah. One of the things that I, I don’t so I don’t know that it necessarily needs to be a goal of everybody to be able to write a writer and a write a reader right now is because I want to change what writing and reading is. And I don’t know that we can we that everybody can start that. I think that there are some very fundamental things that happen, particularly as a consequence of doing things like endure, like, like substantial reading in Vision Pro that just make me want to do different things that that haven’t been picked up yet.
Brandel Zachernuk: So, like, I’ve been continuing to read Marshall McLuhan Understanding Media in Vision Pro sitting on the sofa back there with, you know, the screen pitched up like that and kind of looking like I’m dead for three hours at a time every night. And it’s amazing. And but there are some really interesting things, really interesting provocations that come to mind as a consequence of doing that. Like, I would love to, for example, annotate in a social way and, you know, Amazon Kindle and read Amazon.com. Allow me to to make notations that contribute in some diffuse way to like lots of people highlighted this part. So maybe you want to feel clever about it too, but something else that would be pretty awesome is to actually leave bookmarks in there to say, like, if anybody wants to talk about this, then maybe let me know and we can actually literally chat. You know, the idea of having an intrinsic social first kind of mechanism, and if it’s social first and spatial, then that that makes some really incredible sort of opportunities. And like I was saying last Wednesday being able to build a substrate upon which people can actually have a trusted network of annotations that can live in situ would be something that could not just transform. Something like Wikipedia is that but if you had if you built academia that way, then people could write papers that literally were annotations of other papers and it wouldn’t be, you know, on the it would the owners wouldn’t necessarily be on the the reader of both papers to do that reconciliation between them.
Brandel Zachernuk: You know, you could publish an annotation and that would just go over as a slip on on the original originating paper. And, you know, the real estate doesn’t permit then you can revise and rearrange it. But what you have is, is a literally derivative work. Something that so but but yeah, like that model, you know, requires a lot of trust and and and copyright negotiations things like that. But but that would change reading along with the proper integration of audio visual sort of capabilities. And I don’t think that stops being fixed simply because it’s not exclusively glyphs. I know we sometimes disagree on that, but I think that there are even if there isn’t text, I think it’s worthy. And I think it’s worth worth thinking about the practical consequences of. So with all of that in mind, like one of my worries is that like, I think people will be able to participate in that soon, but I think that there are some some fundamental groundwork in infrastructural things that need to be done to the basic concept of reading and writing and books that we have to do first in order to render those artifacts imaginable to the rest of the public.
Frode Hegland: That was a wonderful speech. And now I’m going to go flying into it. Headwind. Yeah. Okay. So pardon my French people are listening, but not French. Pardon my bad language. Let’s fucking solve this. Right. I think we can, because forget BibTeX and HTML as being an either or. Let’s just do two things in this community. And I put a note here so I don’t forget myself. Let one of them can be. We define what a reference section can look like. Right. And again if it’s an HTML or BibTeX or if it’s written in Urdu, I don’t give them monkeys. Right. But we say author first name equals or just author name. This is stuff that Marc and I have gone through for years, right? But we can also have lots and lots of exciting alternative field link type equals. If it’s empty, who cares? Ignore it. Right? It can be sentence. It is found in the document. If the authoring system can automatically put it in, all of that stuff can go in. And the only reason it doesn’t exist is because it hasn’t been built and implemented. If it’s built and implemented in more than 2 or 3 systems, they can extract this stuff.
Frode Hegland: It’s a standard. It may not be an immediate worldwide hit, but it is a standard. And similarly, on our Wednesday talks, one of the things we’ve gotten to the point of saying, and I’m using the term library extremely liberally. Imagine having a document where the only thing you care about is the reference section. Everything else is irrelevant. It’s just a library that gets really close, frankly, to what you’re talking about. A list of glossary glosses. But there’s nothing wrong with that. That is very, very important. You know, if you look at the history of books, writing, printing, some of the earliest books printed were basically that they were glosses. That was all it was. So if we do a little bit of linking, a little bit of connection, but if we just implement this, I will implement it in visual meadow. But in visual meta it can be anything. Right? Let’s say Fabian says that’s nonsense and he wants it in HTML. That’s fine. I’ll just write the HTML inside visual meta, but that’s how I’ve stolen it and then the software can extract it. So I really want to hear more about just doing this stuff. Leon, solve all my problems, please.
Leon van Kammen: So yeah, I, I am one of these people who is just fucking doing it. The problem is that I get cricket sound or the sounds of crickets here, and I will I’ll show you what like, for example, I am I’m adding or I have added descriptive metadata in in my Excel fragment spec, which basically means if I have a 3D scene, then it will scan. It will say, let’s, let’s do this, let’s just scan for it.
Speaker4: Because it’s so.
Frode Hegland: Important what you’re doing. Can I ask you a clarifying question?
Speaker4: Yeah.
Frode Hegland: In a sense, would it make sense to think 3D first? So XR fragments first and then the other stuff is just like a shadow of that rather than the other way around? Maybe. Is that what you’re thinking by saying XR fragments being important?
Leon van Kammen: Well, well. No. Well, I don’t even know that. What I’m.
Speaker4: So sorry.
Leon van Kammen: For. It’s such a big space. So? So I just see, like, this world of descriptive metadata. And the moment, if I erase my fingers, there’s, like, thousands of people who are going to say like, no, we talked for years about this. Go to the RDF guys or go to the BibTeX guys, or if you don’t like those, then you have to go to the Dublin Core guys or, or maybe just go to the the area guy. So I anyways, I talk with a lot of people. I was like, okay, I’m just gonna sort of Polly adopt all of these things Dublin Core open graph. And so I’m now scanning for the let’s say Exp’d header. What kind of license is on this document or creator Dublin Core? Then we have OG title Open Graph, which is also supported in DC. And we have well, basically every standard has some different unique ones. There is some overlap. But yeah. So my my point is a bit like I’m doing this, but I have a feeling that anybody who says let’s fucking do this like me or you like we get cricket sounds because it’s such a loaded environment that I yeah, I feel like it’s you almost have to go full activist mode to really pull something off. And I’m not really an activist. I’m just making something which I think is is useful. And I’m just adopting, probably adopting all the the most popular specs. And to make it even more complex, I even noticed that the RDF movement actually started out as a sort of poly metadata kind of you know, they had this polyglot vision, and now they decided to no longer do this and sort of, like, rule out some older ones. So that, that that also scared me. I’m scared. Fraud.
Frode Hegland: Come to London, we will do another barbecue, I hear. By the way, is finally grown back. I just want to show you that Mark has written on this, and I completely agree with him. And also, before I give the mic over to Mark on our Sloan work, we are making this stated assumption that the metadata for the documents is perfect. So obviously we’re faking it, right? But none of it is magic thinking all of it is the reasonable expectations that document can be proper, full electronic literature document and not just full paper. Right. So as part of that dialog and it doesn’t even mean you guys have to be there on Wednesdays. This is important, and I would love to try to work with you, Leon, on incorporating some of this into author and seeing if that works, and then see if we can get into the reader. But also it would be great if, for example, Fabian with your HTML world, if we can look at a way that whatever we put into our horrible system could easily be opened up to what you’re doing, you know, we could very happily put HTML stuff in there to find a way to render that in your world. It has to be bigger than just a to be Mark.
Mark Anderson: Yeah, I was going to just loop back around because I the your passing comment earlier was saying, you know, well, Epubs don’t really work, but I think that’s only because epubs aren’t really finished. I mean, I spent some while on those when I was making the one to Ted Nelson work. And the things you trip over are things like it’s the unfinished business, so no one’s really thought about, well, what do you do with footnotes? Or, you know, where do the where do the end notes go? Do they go to the end of the page? Do they go into a separate article? Do they go into a separate document? I think these things are fixable. It’s just there doesn’t seem to be I maybe there just isn’t. The caucus is having the conversation about that. And I suspect the publishing world, they were all sitting back hoping somebody else is going to do this because, well, they’re commercial organizations, you know, why do they want to spend the money to do it? So I see what happens. It’s frustrating because you know, Epubs are are probably there. I mean, there is a separate issue compared to a book that I just think the handling of images at the correct resolution and presentation is there’s a technical issue there that I don’t quite understand how it is, but most ebooks, I find sub-par in that sense, which is why I’ll read a novel in a in in a Kindle or a e-book reader. But I wouldn’t read a textbook I really care about in in an e-book form at the moment, because once you get away from the text, the quality of the presentation tends to fall apart.
Mark Anderson: Then I believe that again, I believe that’s addressable. It’s I guess it’s I don’t know where the impetus will come from for the conversation to move those things ahead, but if we there are lots of useful things, for instance, that have evolved in the hundreds of years we’ve used the, you know, print format and things like, you know footnotes, for instance, are really useful in the right place and just sort of to leave them in the wrong long grass would be wrong. Therefore, we have to have a conversation as to where that fits in a non-linear document. You know, that isn’t in a sense predicated on being printed out, even if it never is, which is sort of that’s the bind we’re in at the moment. And we haven’t moved our understanding of. So we haven’t gone and done is sort of gone back and say, okay, this is broadly what a document used to be. Where do where, where do we find a home for all the parts of it. So we’ve done the easy bit. We’ve sort of taken the body text and we’ve done some very good stuff with, with sort of HTML. But there are other there are other bits of the richness that we haven’t yet ported across. And I think that’s what’s holding back some of the digital representation. So it, I don’t think there are any insurmountable problems. It’s, it’s where the conversation of all the necessary interested parties can take place in.
Speaker4: Yes.
Fabien Bentou: I have to to run in 30s now. I guess it’s just in term of I’m not going to say the swear word, but in term of doing it, I made a tiny example in in the chat as a Dockerfile. So it’s making a container. And here it’s to use caliber which has a converter. So you basically convert whatever in term of. Document, text based document, Epub and whatever to PDF and then potentially PDF to JPEG and then using it in the headset. So it’s very rough. Let’s say it’s not something beautiful, but it’s something usable in term of including kind of I don’t want to say whatever, but document based kind of content in the headset. And, and then I had other things to do. But I’ll keep it as a teaser for next time and want to thank you everyone. Gotta run. Bye bye.
Speaker4: Thank you. You know Randall.
Brandel Zachernuk: The subject of RDF isn’t finished. I agree. One of the things that is so challenging about the position and history that we have is that we just haven’t had the time with any of these technologies to really understand what we want out of them. You know reading Fisher and Ong and other folks who give have the ability to convey a deeper sense. And and Blair having the ability to see the deeper history of writing systems and the practical implications of these other sort of seemingly happenstance things that happened centuries, centuries ago or for four centuries. You know, the fact that, like, it’s arguable that Latin, the Latin character set, was better positioned, but like, actually less better positioned than you would have expected than Arabic for a movable type printing. But that had and that was just like happenstance that that it went through changes because like, there are benefits for the way that Arabic is joined up and has uses ligatures and diacriticals for for vowels and things like that, but it meant that it was worse position to be able to leverage offset movable movable type printing. And there were massive consequences for that in terms of how that meant that, that that printing could kind of spread things tinted through the particular kind of lens that was going to come from it. And in that same way that like. But that was because of changes like small evolutionary changes to the way that like like there’s not that much there’s much less about English to Arabic. That would have necessitated that the writing be the way that it is. But it underwent genuinely very long period, long form evolutionary changes into a thing that meant that you could punch keys and get individual differentiable glyphs out of it.
Speaker4: And.
Brandel Zachernuk: We just don’t have any of that time for people to kind of live with the stability of a set of features in order to figure out what, what a set of capabilities are. And so, like the book is just not closed on on Linotype, let alone TF, let alone any of these more modern things. But, you know, we linotype’s only, what, 100 and some years old and that it’s it’s just really frustrating because.
Speaker4: We.
Brandel Zachernuk: A lot of people will say, well, we only need to look at the last 2 or 3 years of innovation. And it’s more like we we probably need to be taking seriously the possibility that we’ve left stuff on the table from the past 2 to 300 years in order to be able to seriously consider what we ought to be able to do with technology now.
Speaker4: I have a book.
Frode Hegland: Mark, are you ready?
Speaker4: Okay. So
Frode Hegland: Stephen Mitton British guy has written a new book called Language Puzzle and he refers to the. Functional specialization and domain specific thoughts rights, which is related to my concern that we may be able to not have certain thoughts because they’re going in different sections. This book is great for that section is great for that. But yes. Brandel, we have for some of this, it’s been very incremental and very, very slow. Absolutely. But I do believe that so Leon and me are going to have to collaborate a lot more. And I do believe that we’re at an incredibly interesting time in history because before this year, because of that Vision Pro thing the world was we talked about rich media and we talked about plain text, like, there is such a thing as plain text. Of course there isn’t. There’s no such thing. You still have to have a typeface. I think we’re at a point now where we can accept that the world is more than 2D. As the base. So if we now. So when I look at ZR fragments of course I’m thinking extended reality. If we now just put a stake in the ground and say that for the way we’re going to be handling data, metadata, documents, dialog and so on, we have to take into account spatial positionality and context and connectivity. That. That’s the base thought.
Speaker4: We can start.
Frode Hegland: Noting that down. We’ve talked about it for the last 2 to 3 years in different ways, but now it can mean something. I mean, I found that the Vision Pro for me to work in as a practically useless today, but the only way I found that out is by using it. It’s a wonderful device for many things, but now we’re at the point where we can use and try and realize a lot of it is nonsense. So if we can find ways again, the most basic would be reference rich reference section. Secondly, Wendy Hall said exactly what you said. Randall, a few years ago when I started my PhD, I’d like to be able to compose a document based on other things. Right. And this is very much what Alan has been showing us today as well. Pulled things in. Right. Don’t retype it, don’t plagiarize it. These are entirely possible now. And yes, I will put my version in visual meta, but there’s no reason it can’t also go on HTML and everything else. But as when I did The Future of Text a few years ago at the London College of Communication, when I was a teacher, somebody asked about the origins of the internet.
Frode Hegland: And I started telling the story of the students doing it. You know, Vince surfing them. They were just students at the time. None of us here are students right now. But who the hell else are going to do this? So if we now while on Wednesdays we deal with design because Andrew is doing the coding and we have general discussions on Monday, if we can also deal with just creating these standards, pretending we own the world. I think there’s a good chance we might end up owning the world. Now just imagine in terms of, you know, Denny and me handing in this long report with these standards that we’re all credited on, you know, we wanted it this way, but other people said this. We fought it out. We decided on this. You know, that’s a real valuable piece of work that we can do together because of that. And then we have this dude from Apple here who’s making real things, right. Brandel? You know, these are these are this is not an abstract discussion. What do you say? Should we try to have more standards talk.
Brandel Zachernuk: The use of standards.
Speaker4: What what what.
Brandel Zachernuk: The standards should be in service of. Yeah, that’s something that’s important for me. What I do with standards is mostly strictly around spatial stuff. So that means actual literal 3D models and stuff like that. But but I think that, you know, as we sort of fill out our understanding of how information and space go together, those things can, can, can work in tandem. But yeah, I in order to do that job, I should probably go and start another meeting. Yeah. This has been really awesome.
Frode Hegland: I see you have that other meeting, Brandl. But on Wednesday, you’re going to try to talk some of this stuff standards and stuff this Wednesday as well. So I look forward to that. Well, whatever time you have on Wednesday, we’re doing lots of other things.
Speaker4: Cool.
Leon van Kammen: Thanks. Good to meet.
Speaker4: You. Fine. But see Randall.
Frode Hegland: And the rest of us. I guess we’re signing off. Mark has a comment. Anyone else? Please feel free to chime, but you have a mute comment. Okay, that’s even shorter.
Mark Anderson: Sorry, just a very brief comment on the on the aspect of metadata is, is to build ourselves the time to at least look more than once through it. I think one of the problems with I’m seeing with a lot of projects I’m involved in the moment, everyone’s in a rush to make metadata as the means to get to where they want to go, but the metadata tends up being a bit shonky as a result because it’s a thing and say, what do I want? What do I need to do what I want? And then looking back and said, okay, well, now I know that I need this. Is this, is this actually a sort of useful and robust way to do it? And it’s that little extra step that I think we tend to miss out at the moment. So that’s something our practice should perhaps think more on.
Frode Hegland: Yeah, we should and let’s remind ourselves also that a lot of, a lot of metadata is already there. Just don’t throw it away.
Speaker4: Ellen, please. Yes.
Alain Marsily: Do you know that Adobe has worked on the Rich PDF and it wasn’t really successful, but I think they have already dig the the subject, but it could be perhaps interesting to to understand where they handle this rich, rich format because no reason that it has no success because there are part of the future and I’m pretty sure that they have worked on, on on that matter and would be interesting to, to have contact with the product manager of, of this. Because the rich PDF has hasn’t any success any, any way does there are some reason, probably technical, probably let’s say no agreement with other other format, but it’s for me. Adobe is must be the player of of that that matter because they have so many connections to different connections that it could be designed, it could be text, it could be video. And I think it it must be a player somewhere in, in this in, in this understanding and and the way this stuff must be handled in the future. My, my fear.
Frode Hegland: We have been in dialog with with Adobe on some of this, and the PDF standard is now externalized. Anyway, so we’ve been dialoged with that group that I want to change so much, they want to make sure things don’t break. I’m also was also a member of the PDF association for a while for exactly this reason. And you know, the only reason I’m a fan of PDF is universities demand it and it lasts forever. That’s it. Beyond that, they’re awful, right? I really wouldn’t want to do a multimedia presentation. You have to look at it on an A4 on a computer screen. That would be an absolute joke. So, you know, obviously we need to take the best from each different format and try to put it together. And I think the only, the only way to succeed is through success. We’ll just have to build something nice. Now. Ellen, we have a grant from the Sloan Foundation to to work on reading and writing in XR. It’s an okay amount of money, but it means that at the next two hypertexts forum conferences, we will be putting headsets on people, and it’s mostly theater. People should feel that. Oh, I can see differently and I can work here. And when we say it’s mostly theater, the data has to be there to back it up. Even if it’s not ready. We need to. Yeah, it’s metadata all the way down. That’s right. Mark
Speaker4: No, I mean the entire.
Frode Hegland: The problem I have with finishing my thesis is probably my. My PhD thesis can be summed up in one sentence. If it’s important, write what it is at the back of the document. That doesn’t go down so well with my examiners, right? That is the entire philosophy. And it’s very, very extensible. And so many people in this community have added all kinds of things. Like Peter has added the notion of extra visual methods that can be deleted. That’s fantastic. You know, we use it for AI at the moment. Anyway, I also have to go. Emily’s made barbecue chicken today using her mother’s recipe, which she finally found, which is lovely. I look forward to seeing some of you on Wednesday, the rest on Monday alone. You are welcome anytime. You can come and go as you want. There is no special need to be okay. And also, as you may have noticed, some people today came and left in the middle. That’s life. That’s what has to happen. It’s not considered a bad thing.
Speaker4: All right.
Frode Hegland: Thank you, everyone. And we got work to do. Bye.
Speaker4: Thank you.