Peter Rushforth: I'll start. Okay, there we go. So, hello. Welcome to the day 8 of the workshop, the joint or OGC/W3C workshop on maps for the web. I'm Peter Rushforth, and we're going to, we have a concluding program today that is very exciting. And then we're going to
Peter Rushforth: Have a have a discussion about conclusions and next steps that we hope will have a bit of an open mic opportunity during that session so
Peter Rushforth: I just like to remind everybody that we can chat on Gitter, live chat is welcome there and
Peter Rushforth: I pasted the the URL to the Gitter chat in the in the zoom chat, we can use the zoom chat for
Peter Rushforth: comment. Comments if you don't have Gitter access but mostly just for zoom problems or what have you, audio problems, that kind of thing.
Peter Rushforth: And we're going. We do have a Discourse forum category set up for
Peter Rushforth: web mapping and I'll paste that into the chat as well. And I've been, we have been
Peter Rushforth: Adding topics for each of the each, each of the sessions and linking to, presentations and videos and slides as they become available. And we'll go ahead with doing that for tonight's work as well. And with that, I would like to introduce our,
Peter Rushforth: Oh, first of all, I'd like to remind everybody that the
Peter Rushforth: Workshop is operated under the
Peter Rushforth: Standards of Professional Conduct of the W3C and there is a link it on our workshop website at the bottom of every page that I'll paste into the chat window.
Peter Rushforth: And
Peter Rushforth: And remind everybody to adhere to that, please, so that we don't have to mute you
Peter Rushforth: Alright. With that, I'd like to hand over the the stage to Terrence Eden, who's going to give us a talk on
Peter Rushforth: You user research on maps in HTML from his perspective. And that's the only talk to the night, we'll pass it into the panel after Terrence is finished. So Terence if you'd like to share your screen and take over that'd be great.
Terence Eden: Thank you very much. So in time honored tradition. I'm going to start by saying, can everyone see my screen. You should see a big 'Hello my name is' is that on the screen.
Peter Rushforth: We do. Yeah.
Terence Eden: Yep. Fantastic. There we go. What, what an auspicious start so
Terence Eden: Good morning, good afternoon, good evening participants from around the world. Thank you so much for having me today.
Terence Eden: My name is Terrence Eden and for a bit of background on me. I was the UK Government's representative to the W3C's advisory committee.
Terence Eden: But for the last 18 months I've been working with the UK National Health Service to improve their digital capability. This talk is in a personal capacity and doesn't represent the views of my employers, past, present, future or theoretical so
Amelia Bellamy-Royds: Sorry to interrupt, Terrence we're actually seeing your presenters view, which is okay, but
Terence Eden: Oh, interesting.
Amelia Bellamy-Royds: If you just want to just share the slides.
Terence Eden: Thank you for letting me know.
Terence Eden: Right. Let's see if I can share my entire screen or if that's just going to be black.
Peter Rushforth: Amelia.
Terence Eden: That's just a black screen. Brilliant. Thank you, zoom, I...
Terence Eden: Right. Okay. Well, in which case. Just give me one second.
Terence Eden: I have too many monitors and not enough fingers, it would appear
Peter Rushforth: My apologies. I should have caught that for you, but
Terence Eden: That's all right.
Terence Eden: So hopefully you can just see a thing which was mapping our users.
Peter Rushforth: Yeah.
Terence Eden: And you can't see all my top secret notes I've written to remind me what I'm saying.
Terence Eden: Illusion.
Terence Eden: The illusion is maintained.
Doug Schepers: We only see part of your screen.
Doug Schepers: Oh, it's it's it's cut off at least
Peter Rushforth: Yes that's true at that at the after the OH and mapping our
Terence Eden: Is that any better.
Terence Eden: Yeah.
Peter Rushforth: Slightly better
Doug Schepers: Well, not it better, but not-
Terence Eden: Fine. You know what, then, in which case
Terence Eden: I will if you just give me one second. I know exactly what I will do, I will do it the old fashioned way. I'm so pleased that you're recording this, Peter, because that means that all of this will be preserved for
Terence Eden: Generations.
Peter Rushforth: We have a very capable editor in Ted Guild
Peter Rushforth: who clips and snips, as appropriate.
Terence Eden: So,
Terence Eden: Is that showing just 'mapping our users'.
Peter Rushforth: No, no.
Doug Schepers: No, you're still cut off.
Terence Eden: Sorry, just give me one second. This you wouldn't believe that this did work in rehearsals and
Peter Rushforth: So is there a window in your view options thing.
Terence Eden: Just give me one second. Let's
Terence Eden: See what's going on here.
Peter Rushforth: Ted,
Terence Eden: Well, in which case I will just do it.
Peter Rushforth: We'll do it off the notes page.
Terence Eden: Yeah, there we go. So
Terence Eden: No, no.
Terence Eden: You shouldn't you can probably just see the whole
Doug Schepers: In fact, wecan only, it is still cut off.
Chris Little: Cut off in the same place.
Terence Eden: Okay and I'm very sorry about this. I'll try doing it off the PDF. Who knows, maybe it's just an issue with my Chrome
Gobe Hobona (OGC): So, oh, if Peter has a copy of your slides, can I
Gobe Hobona (OGC): Have
Peter Rushforth: I have the PDF on the on the website here.
Terence Eden: Okay, well, I'll just give it a go with this. I'm
Terence Eden: Location window.
Terence Eden: Is this cut off.
Doug Schepers: I hate to tell you it's worse, when now we're only seeing a quarter of your screen.
Terence Eden: Amazing.
Peter Rushforth: Why don't I do the PDF and you can say. Next slide.
Terence Eden: Next slide please. That sounds like a brilliant plan, right, I shall, if you want to take control the screen share
Yeah.
Chris Little: If it's any consolation I had a similar problem from Firefox.
Peter Rushforth: You see, do you see the slide now.
Terence Eden: I can see 'Hello, my name is' I hope everyone
Is
Terence Eden: Right. Well, listen, let's, let's just cut in here. Hi everyone and welcome to a completely seamless presentation. My name is Terence Eden, and
Terence Eden: Bit of background, I was previously the UK Government's representatives to the W3C, but for the last 18 months I've been working with the UK National Health Service.
Terence Eden: This talk is in a personal capacity and doesn't represent the views of my employers. Next slide please.
Terence Eden: So I've been following along with bits of the workshop as my timezone allows, and I've really enjoyed what I've seen, but I think there's a way to make these proposals, even better, and that is talk to our users.
Terence Eden: And I've read through lots of the documentation that's been presented and some of the proposals and I see lots of use cases and user stories.
Terence Eden: Or what I don't see is any evidence that people have spoken to end users. And in this talk, I want to explain why engaging early with users is a good idea. Next slide please.
Terence Eden: So developers are users too! We often think of the user as being the person who is scrolling on their screen or, you know, interacting with the map, but
Terence Eden: Actually for the sorts of proposals that we're talking about here. We're not just targeting end users, not just targeting people who interact with a web page.
Terence Eden: We have to make sure that we have done the hard work to ensure that developers can adopt this and wants to adopt our proposals to make sure that they are included in our user stories and
Terence Eden: Next slide please. And we were all nerds here, right, I can say that because we're all good friends, but you know we're all nerdy about something.
Terence Eden: And I think we recognize the sentiment expressed in this mug, which says 'Linux is user friendly. It's just very picky about who it's friends are'
Terence Eden: Like this. This isn't about Linux. But specifically, but all communities about are picky about who they let become users.
Terence Eden: If you don't know the right jargon, or if you're a newbie, or if there's a high barrier to entry. It can be really intimidating getting started with something new.
Terence Eden: And this is a problem that HTML has as a language. And I want to take a look at a couple of examples of fairly newish HTML elements to see what we can learn from them.
Terence Eden: So, next slide please. So here are video and audio. So these are I say they're new, they were introduced in HTML5, but they still feel new to me.
Terence Eden: They have basically identical functionality. You know, it's, it's a, an element. And there's sources inside and they are expressed in identical ways and they both take control options and they've got a similar layout to the classic image element.
Terence Eden: When I teach HTML to newbies they can easily understand what's going on here. And once you've learned how the video element works you automatically know how the audio element works like they are
Terence Eden: You know, really simple to understand really simple to use and they get a lot of traction because of this. Now let's take a look at an opposite example. Next slide please.
Terence Eden: So this is picture. And if you're not familiar with this, you know, it's a it's an extension to image, sort of
Terence Eden: It's kind of confusing to try and explain it to people. It's sort of superficially similar to audio and video but it contains some really weird syntax. So it has significant white space and significant punctuation.
Terence Eden: There's a media attribute which sort of looks a bit like CSS, but isn't really and it's not used elsewhere. And what does this use
Terence Eden: Source set when other things use when image uses source, the src, there's no alt attribute natively, even when you teach this to people they find it confusing. You look at stack overflow or MDM documentation,
Terence Eden: People find it, developers users, find it hard to use picture element. And I think that's one of the reasons why it doesn't get much use
Terence Eden: And I think part of the problem is that when this was being developed and I had a very peripheral role in talking to people, while it was being developed is
Terence Eden: Web you know standards experts talked amongst themselves. They didn't perform research with developers and we need to because
Terence Eden: If we need to make sure that people want to use our product. And the only way they'll do that is if they understand it. And the only way they'll understand it, is if we do the research to see what their needs are.
Terence Eden: So, next slide please. This is a slide which Peter presented earlier and I love it. If I was designing the map element. This is almost exactly how I would develop it.
Terence Eden: And that's why it shouldn't be like this, because I am not the average user my opinions shouldn't matter, even more importantly, the opinions of everyone on this call.
Terence Eden: Shouldn't matter we are all maps nerds and standards nerds and HTML nerds, nerds of some sort, and it's not about us. It's not about
Terence Eden: People who spend their days, thinking about maps, we need to think about people who were being told by their boss, hey, I want a map on the website and needs to do this.
Terence Eden: And they don't want to get into the semantic discussion of, you know, 100 different things. They want an element which they can really easily pick up and use and yeah, is powerful, but hasn't necessarily been designed for the needs of mapping nerds and standards nerds and
Terence Eden: So I want to talk about how we get around to doing that. So my plan was actually to spend the last six months, talking to developers and other users to see what they expected from new elements, but, as some of you may have noticed, the world has had other plans for the last six months.
Terence Eden: So, next slide please.
Terence Eden: Here are some of the questions I wanted to ask, and I suggest before going much further
Terence Eden: With these sorts of, with this proposal that the group engage either a proper user research or do some, you know, really quite structured albeit informal user research to ask these questions
Terence Eden: And others because these are the questions I would want asked me if someone asked, you know, what should be in a new element. So, next slide please.
Terence Eden: I guess, the very first thing is, what should this element be called. So I saw map earlier. You know, I've been doing HTML since, well since, before my beard was a lot shorter and does it make sense to recycle map.
Terence Eden: You know, I saw a map ook earlier. That's interesting to people understand what a map book is, is essentially a geo thing, is this about geography, would this make sense as an emoji. Why can't we have an emoji element, would that be interesting, would people get that
Terence Eden: Intuitively. We know that the majority of web developers in the future will speak Chinese. That's sort of almost demographic certainty, will this
Terence Eden: Idea get more traction. If this is the first Chinese element. The first element with a Chinese name. I don't know the answer to that. You don't know the answer to that. The only thing we can do is speak to users to find out what they expect this to be called. Next slide please.
Terence Eden: There are some things inside
Terence Eden: The element that Peter had shown, lat long
Terence Eden: Why lat long, is that what developers want see. Do they want latitude and longitude. Not everyone speaks English. We're not trying, you know, we don't have to save every single byte.
Terence Eden: What other coordinate systems are being used, people in the UK might want to use Ordnance Survey grid references, rather than latitude and longitude.
Terence Eden: So perhaps we need to talk to us as to say, you know, not just the design of what mapping things we can put there, but the actual names that people want to use. I mean, if I'm Googling for something. Do I know what lat and lon are, why isn't it lat and long?
Terence Eden: It's worth talking to users to find out what they want. Next slide please.
Terence Eden: The zoom thing I found really interesting. So I did chat very briefly to people just before the pandemic kicked off about what they thought zoom meant
Terence Eden: And it's really interesting. Some people think that zoom zero means there is no zoom. So your zoom all the way out, and others say there's no zoom. So you're zoomed all the way in.
Terence Eden: What's, what's the consensus here there's what people use at the moment there. But there's also what people expect
Terence Eden: There's also what other systems use, should zoom be an integer, should it be a number at all, or perhaps when people are saying, zoom, what they mean is fit it to the screen.
Terence Eden: Again what our expectations are of what zooming is in a mapping context can be very different to what a developer
Terence Eden: Thinks, very different to how other elements work and that may be where their mental model is if we don't understand the mental model of people who want to use this element, we're going to wind up with a with a mess.
Terence Eden: Next slide please.
Terence Eden: Here's another one. So controls. This is something which I think
Terence Eden: The video and audio element have got wrong. They have a binary state controls on or off. And what that means is that little bar at the bottom, which allows you to play and pause and fast forward and stuff you can turn it on and off.
Terence Eden: But that is that fine grained enough, what controls do users expect our end users might expect, so end users might pinch and zoom and
Terence Eden: Pan and things. But what are the controls might have developer wants to see in there. Which ones would they want to turn on and off. Does it make sense to say I want this map to be panable, but not zoomable
Terence Eden: Maybe, but we won't find that out unless we actually talk to people and find out what controls not only do they want to do they expect to be able to enable and disable
Terence Eden: Next slide please.
Terence Eden: So there are lots of existing solutions out there. I mean, I used at least half a dozen different HTML map things in the last year.
Terence Eden: Can we say, let's just take the best bits of each. Well, yeah, but then we're, again, we're only scratching our own itch.
Terence Eden: We need to speak to the customers and the users of the existing solutions and say, hey, what do you love about the current mapping solution.
Terence Eden: And then, even more importantly, what do you hate about your current mapping solution. If any of you have done product development or worked on a
Terence Eden: New business, you know, sometimes the most powerful thing you can do is look at your competitors and ask the customers. What don't you like about that. And if you can be the opposite of that. If you can do the one thing well, that others can't. And that's how you get traction
Terence Eden: Because
Terence Eden: I'm literally wearing a t-shirt, which says 'users first', we can't rely on our own instincts our own instincts are wrong, because we think too much about this. If you look at you know, there are some
Terence Eden: Bizarre HTML elements which have been put in as far as I can tell, because
Terence Eden: nerds like us said, 'oh, it would be really good if' without actually talking to people and seeing if there was a absolute user need on whether they would get traction
Terence Eden: And I think that mapping is too important to fall by the wayside like that. This is why we absolutely have to do user research. Next slide please.
Terence Eden: So who are our users. Do we know?
Terence Eden: I mean, have we spoken out to people. If this proposal gets into the standard, but nobody uses it, it will be a failure.
Terence Eden: And you know will kind of have wasted all our time. Yes, we'll have done something that's very technically interesting but if it doesn't get use. Then there was no point to any of this.
Terence Eden: So I want to make sure that we together as a community, create a proposal that developers will love and the best way to do that is to speak to them before we go too far down the path of creating something which satisfies our needs. Next slide please.
Terence Eden: And that's the end of my presentation. Thank you very much for bearing with me through a technical difficulties, and I think we probably have time for a few questions.
Peter Rushforth: Sure we could take a couple.
Peter Rushforth: That's a great presentation, though. Thank you very much. I agree. We need, we need to get get
Peter Rushforth: Our, our domain in front of more developers to
Peter Rushforth: Explain the possibilities and I'll have people help influence it. I mean, part of the part of the challenge of a community group is a low profile, you know, battling a little profile.
Peter Rushforth: And I know there are vehicles for reaching out to users. Typically the the Mozilla, Mozilla web DNA survey is
Peter Rushforth: A vehicle for reaching users, or reaching developers, but I think part of the challenges that web mappers, web map developers don't see themselves as web developers.
Terence Eden: In some cases,
Peter Rushforth: You know that there are some high end professionals who are high end web developers, but they are the intermediaries for web developers, right. They just want to have have that element.
Peter Rushforth: That hopefully is not a dip.
Terence Eden: Yeah, thank you for that. So I'm just taking taking a quick look through the Gitter thing it is entirely possible. I'm mistaken about the history of the picture element, it turns out, but but I will reiterate, you know, when I teach HTML
Terence Eden: Again, I see it as a that particular element is huge barrier to entry and I appreciate. There's a lot of historical baggage, but we potentially have a new chance here of
Terence Eden: Creating something which doesn't have that. And I see Brian doesn't want to have to type latitude every single time.
Terence Eden: That that's a valid thing. But again, Brian, you are you are possibly not the average user
Terence Eden: I don't want to type latitude every time. But again, I'm not the average user, it's worth chatting to people, especially people who, where English isn't their first language to see whether they actually understand why things are abbreviated the way they are.
Terence Eden: So do we have any
Terence Eden: Questions either here or in the chat or should we move on to the next.
Peter Rushforth: Well, maybe we can move on and you can take some questions in the chat, because it's good form
Terence Eden: I shall do. Thanks for listening, everyone.
Chris Little: Actually
Chris Little: It's Chris
Chris Little: Hi, one of the things about asking users really made me
Chris Little: Change my view. We're having a discussion in a geospatial community about the third dimension like GeoJSON allows you to add
Chris Little: 'Z' or 'Zed'
Chris Little: The virtual coordinate, and as a Spanish speaker, he said we should use altitude, not elevation because elevation's use for angles above the horizontal
Chris Little: but he said the real reason is because it's because it rhymes with latitude, longitude.
Terence Eden: You know what mnemonics are a really powerful
Terence Eden: Thing though, yeah I would
Terence Eden: I'd go with that. I've never heard an argument of development from from poetry before, but I'm all for it, Chris, I think that's a brilliant idea.
Brian Kardell: So I also asked a question in the channel, can I ask, Oh,
Terence Eden: I'm sorry, I missed it, Brian.
Brian Kardell: That's okay. Um, so I actually, I really liked a lot of your presentation, I
Brian Kardell: Agree with very much, even if I disagreed with a couple of details in the Gitter
Brian Kardell: But, uh,
Brian Kardell: I guess, I would like to know more about what you propose in terms of what it means to talk to users, because this has been of a thing that I have been really focused on and it is
Brian Kardell: Like actually tremendously hard if you can't give them a thing because
Brian Kardell: Once you give people a thing.
Brian Kardell: They begin exploring it and
Brian Kardell: Yeah.
Brian Kardell: Yeah yeah and and like once they have a tool and they can explore it, they often do more interesting things than you intended, which is impossible to do in early process right
Brian Kardell: It's one of like also why you get like one of the famous quotes about like, faster horses, like people
Brian Kardell: for faster horses right, but if you give them a car and you say, like, what do you think about this, then you know they begin doing interesting things and giving you
Terence Eden: Agree. Good.
Terence Eden: So part of this is, yes, we have to invent the future and you know, we are much like Star Trek, going where no one has gone before. And sometimes we're going to have to produce something which is so new and confusing to people that it will take time for them to adopt.
Terence Eden: That being said, one of the problems I see, and you know, I keep an eye on
Terence Eden: incubators and I'm also on the AMP,
Terence Eden: Google AMP advisory board as well is people come along and they say, we've got this brilliant idea for a new element.
Terence Eden: And it sounds really interesting and you say okay do
Terence Eden: What the developers think about this, and they go, Oh, we don't know, We just thought about it inside Google or inside Yahoo or inside, whatever. And that's, that's no way to do standards driven development. So what I actually propose is is a couple of things. One is, make a very
Terence Eden: user focused part of whether it's GitHub, or somewhere else, where you can say, look, here are some of our proposals, what do you like what do you dislike about them.
Terence Eden: Go and find user groups where people are either
Terence Eden: Getting Started with HTML or are quite experienced and ask them very specific targeted questions. Now, completely up front, I am not a user researcher.
Terence Eden: I don't even play one on TV. I would recommend finding either freelance one or professional one, and getting them to design, you know, even just a questionnaire that you can put in front of people, so that you can start gathering
Terence Eden: Information about what people want. And again, you know, looking at the demographic of the participants.
Terence Eden: Yeah, I don't see a lot of names which
Terence Eden: Are sort of outside Europe, I see a few which look like they might be. But, you know, we are not, you know, it's not the Western world, wherever it's the worldwide web. So we need to find a way to go out and reach web developers in,
Terence Eden: All around the world to ask them what they're mapping needs are, one particular thing which I find interesting, and again, this isn't a
Terence Eden: professional employer comment, is different countries recognize other different countries and, you know,
Terence Eden: I'm not going to get into which countries recognize which but if you're a mapping developer and you're asked, Would you like this HTML mapping solution one of your first questions might be
Terence Eden: Does it show that this country is separate from that country because if it does, I legally can't use it.
Terence Eden: So that's why we need to speak to a wide variety of people, because otherwise we'll end up with something which can't be used in certain countries. Sorry, I've gone off on a bit of a tangent there
Peter Rushforth: Okay. Um, well, we do have the Maps4HTML community group then we convene the
Peter Rushforth: Workshop of
Peter Rushforth: There's a worldwide workshop. So like everybody is welcome to join the community. Um, I know that I realized that that's not like
Peter Rushforth: User surveys or anything, but we hope that people will come and help us solve these these issues right, debate them, put issues in about them, you know, suggest ways to to get definitive data on stuff. And I know that there's HTTP archive and so on. So there's lots of tools.
Brian Kardell: Yeah, I think in a lot of ways your custom element
Peter Rushforth: okay sorry
Brian Kardell: And other and other mapping custom elements go
Brian Kardell: Further, because it, it meets developers, where they live. Right. And it says, like, here's a thing,
Brian Kardell: And what do you think about it, it's very easy for them to explore and comment on what they don't like about it. And this is the interplay with what
Brian Kardell: He was just speaking about that. I was curious about, like, if there were thoughts on how we could do that better, because like talking about real, concrete things, and what is your experience. What don't you like about them. What
Brian Kardell: Like, how could they be better.
Brian Kardell: Is actually the economics of getting people involved is way more practical.
Peter Rushforth: Okay.
Peter Rushforth: So Simon has a has a question.
Simon Pieters: Yeah, it's more of a comment really on the
Simon Pieters: The
Simon Pieters: Topic of how do we get feedback from web developers before it's like shipped in browsers. And one way to do that is
Simon Pieters: already part of the process for Chromium and they call it origin trials, that's
Simon Pieters: opt in mechanism so you can
Simon Pieters: Use a new feature before it's
Simon Pieters: shipped to everyone and they collect feedback from the people who opt into this origin trial and iterate on the API design.
Simon Pieters: So that's, that's one. It's a bit maybe a bit. It could be earlier or late in the process, depending on what happened before that, but this could be early if you start with an implementation in
Simon Pieters: Chrome. Yeah.
Peter Rushforth: Yeah, cool.
Terence Eden: Good point. Yeah.
Peter Rushforth: Let's start with canada.ca. Okay, that's great. So I think we have a really interesting panel coming up, so we can move on to that and so
Peter Rushforth: If there if we do do questions for that panel, maybe we could raise our hand in zoom. I think there's a hand raising facility there and
Peter Rushforth: We can organize, organize questions that way if there is time for them, there will certainly be like an open mic session later on where we have to look
Peter Rushforth: Where we want to decide on future future next steps, kind of thing. And that might be a good way to proceed. I'm sure I've not gotten to everybody who would like to have been heard here, so I apologize for that, but
Peter Rushforth: Let's, let's, hand it over toTom Kralidis and his panel.
Tom Kralidis: Great.
Thanks, Peter.
Tom Kralidis: Okay so great. So this is the panel for for open source, for for web mapping. I'm the moderator for the panel and we have Daniel, Simon Andreas and I'm not sure is Will online.
Tom Kralidis: Will Mortenson
Tom Kralidis: Okay, I sent them an email as well. Let's see if he's able to show up.
Tom Kralidis: Okay, so basically this panel, um, you know, in the context of this this workshop
Tom Kralidis: With regards to open sources what and web mapping open source has been a longtime proponent of the web mapping
Tom Kralidis: Ecosystem, open source software so you know there's servers, clients, serializers, parsers, you know, all sorts of different open source software to in support of web mapping. So we've been doing this long enough, for a long time in various Foss4g projects as well as the OSGeo community.
Tom Kralidis: Lots of web mapping going on there for, for, for decades. In regards to this workshop and some of the discussions around
Tom Kralidis: Sort of making maps and the map element, let's say, more of a first class element. There's a lot of discussion around, how can we get this standardized in HTML to, you know, it's all about lowering the barrier. And, you know, similar to the previous talk, lowering the barrier, users,
Tom Kralidis: Making it, making things easier for users. So how do we make maps first class in HTML and get that, you know, get that mechanism standardized. We all once we do get that standardized. We do that needs to be, you know, implemented.
Tom Kralidis: Open sources also obviously part of the, the browser ecosystem. So adoption from those two communities is obviously is obviously key.
Tom Kralidis: So here, the overall discussion is is I guess what the relationships are what the synergies are between those communities, what opportunities exist to try to move forward.
Tom Kralidis: You know standardizing maps and in HTML in, when open source web browser projects, given all the lot of the experience that we also have him on a sort of Foss4g side. So, as Peter mentioned, we can take questions on the on the chat if folks want but
Tom Kralidis: We'll, we'll start now. So at this point, I'd like the panel to
Tom Kralidis: Introduce themselves and you know your name, organization, background and your role and relationship with with open source, mapping and browsers. So maybe we'll start with Simon
Simon Pieters: Thank you. I'm Simon Pieters. I work at Bocoup, a web platform consultancy firm. I have experience working on web standards and conformance tests for browsers.
Simon Pieters: Bocoup participated in the OGC testbed 16 program to review the MapML proposals.
Simon Pieters: So that's
Simon Pieters: That's the
Simon Pieters: Context that brings us to mapping, but the experience with
Simon Pieters: Adding stuff to the web platform is
Simon Pieters: Also relevant
Tom Kralidis: Great, thanks Simon.
Tom Kralidis: Daniel?
Daniel Morissette: Well, I'm Daniel Morissette a software engineer and President at MapGears. MapGears is my own company. So as the company grows, I spend less time coding and more time just doing admin stuff.
Daniel Morissette: Been involved in open source software since the mid 90s. I was introduced to the idea of releasing my source code instead of selling it
Daniel Morissette: By Frank Warmerdam. At the time I was writing GIS data translator and I've been involved in the GDAL libraries with him
Daniel Morissette: When he created it. Then later on I started doing web mapping, tried a couple of proprietary black boxes. Some people may may remember the names of
Daniel Morissette: Map Objects and MapXtreme. The early versions, I mean, the later versions were better but I mean I'm talking about the ones in the late 90s. I quickly gave up and came back to open source for web mapping as well and I never looked back after so
Daniel Morissette: Around 2000, I joined the University of Minnesota MapServer project, I helped set it up as a tree collaborative project with Steve Lime who was the only developer at the time.
Daniel Morissette: Then I've been involved in OGC for some, for a little while and the days of the WMS standard and web map context stuff, which may remind some stuff to some of you guys, web map context.
Daniel Morissette: And then I've been also involved in the open source geospatial foundation when it was created in 2006 and involved in some of the committee's, the incubator, been on the board. So to sum it up my connections are open source software development, mainly in MapServer, GDAL and OGR,
Daniel Morissette: OGC Standards and OSGeo foundation, kind of three hats.
Tom Kralidis: Great, thank you, Daniel, and
Tom Kralidis: Andreas?
Daniel Morissette: You're muted Andreas
Andreas Hocevar: Thank you very much.
Andreas Hocevar: My name is Andreas Hocevar. I've been involved in Geospatial web based mapping projects that are open source since 2007 I think,
Andreas Hocevar: Or maybe even earlier if I also think about community map builder, an old web mapping project that is retired now, but the biggest project of those is probably OpenLayers, which I'm also a very active developer in
Andreas Hocevar: I'm self employed, basically, but I'm also director of a small company called W3Geo which provides development and consulting for everything in the field of browser based web mapping and web based GIS systems.
Tom Kralidis: Great, thank you. It looks like Will is trying to connect, so I'm sending him emails as we speak.
Tom Kralidis: But notwithstanding that, you know, once he does get in. I'll sort of break and ask him to introduce himself and get him wired into the, wired into the discussion so
Tom Kralidis: Let's see what happens there. So thank you for the introductions, everybody. And I guess my first question to the panel would be, you know, Daniel, we heard you talk about OGC and
Tom Kralidis: You know, with regards to open source.
Tom Kralidis: In in the Fiss4G community or the geospatial community and sort of, the web mapping ecosystem
Tom Kralidis: When we relate that to the standardization of maps in HTML. What are the kinds of things that I guess the Foss4G web mapping community can do in terms of helping with with standardization.
Daniel Morissette: Who should go first, like, should I dive?
Tom Kralidis: Sure.
Daniel Morissette: Yeah, well I guess what, that's one of the things that I was, that I did with the MapML stuff, is testing the specs with the reference implementation so
Daniel Morissette: In my case anyway for myself, reading a draft abstract document, you know like a raw a spec document doesn't tell me much until I see it running so kicking the tires, I guess.
Daniel Morissette: And making it possible for others to see what the thing looks in real life, is probably one of the means by which open source developers can help, test driving the specs essentially
Tom Kralidis: Andreas?
Andreas Hocevar: I think developers of web mapping related projects in particular can
Andreas Hocevar: Inform standardization process for web maps, but just
Andreas Hocevar: Talking about, or providing different use cases not only use cases, but also challenges that are related to implementing things
Andreas Hocevar: Right now we're implementing implementing things on higher level API's. There's a chance of rendering web maps more efficiently with better performance.
Andreas Hocevar: If they are deeper baked into into browsers. And I think this is an interesting process. There are very smart folks, folks in the involved in the
Andreas Hocevar: MapBox GL JS mapping library. They wrote a brilliant WebGL based render but WebGL is very low level, you know, in Open Layers. We use a higher level API, Canvas 2D
Andreas Hocevar: And I think the, the biggest chane in this case is to have a discussion about use cases, what's needed in rendering web maps, the whole topic of styling vector maps of
Andreas Hocevar: Change.
Andreas Hocevar: With a rasters are displayed and there's definitely a lot of experience from developers of web mapping projects, whereas on the on the browser developer sides
Andreas Hocevar: They know how to tweak performance on a low level in the browser. And I think there's many synergies. The problem I see right now is
Andreas Hocevar: And we have even on this panel, several developers that know about web mapping from the higher level API side and from how to use browsers, but we haven't really started the discussion yet with web browser developers. There are some
Andreas Hocevar: Things going on in the right direction like today, for example, there was an issue in the OpenLayers GitHub tracker
Andreas Hocevar: About event sequences of touch events. And it was the first time that a browser developer suggested by themselves
Andreas Hocevar: To look into how its implemented in the browser to make it better suitable for OpenLayers instead of the other way around. So that's an interesting development, and I think we have to move way further in that direction.
Tom Kralidis: So basically bringing forward, you know, more and more use cases. And I'm also hearing you know implementation is good proof of helping with the, with the standard and standardization. Simon, do you have any comments sort of from the, from the browser side.
Simon Pieters: Yeah, just want to agree with the last thing as Andrea said that we really need, I think browser developers and the map
Simon Pieters: Map framework developers to talk to each other and get an agreement on
Simon Pieters: Use cases and requirements and how we can change the browser to better
Simon Pieters: address those needs in a you know, well designed way.
Simon Pieters: And if you want to help with standardization, what you really want to do is to participate in the standardization process.
Simon Pieters: And that's often the first case, the first step is to talk about use cases and requirements.
Simon Pieters: And
Simon Pieters: Evaluate existing solutions and come up with new solutions.
Simon Pieters: And then just discuss and get consensus or review proposals.
Simon Pieters: And after that, the next step would be to write conformance tests and write an implementation in a browser engine and get it in the hands of
Simon Pieters: Of users, so for developers, to try it out.
Simon Pieters: And the, this is not something that has
Simon Pieters: Or is very inclusive necessarily, or this can seem like a world of
Simon Pieters: You need a lot of know how to
Simon Pieters: Participate in some in standards.
Simon Pieters: And I'd like to share a link to a guidebook that we've written at Bocoup
Simon Pieters: To create an onramp to
Simon Pieters: Standards work essentially
Simon Pieters: I'll share it in the Gitter chat, but it's W3C.guide.
Simon Pieters: And this goes through all of the necessary steps for
Simon Pieters: Participating and standardization at different levels.
Tom Kralidis: How far do we go along in
Tom Kralidis: You know, when you standardize something, do you show up with the implementation first or do we come up with, do we get the standard to a certain point before people do the implementations, a little bit of a chicken and egg but what do you think about the sequence of this?
Simon Pieters: I think the sequencing should be to get alignment on use cases and requirements first, with the all of the relevant stakeholders at the table. So you want both
Simon Pieters: Browser implementers those who are going to
Simon Pieters: Own the implementation on the browser side, if you
Simon Pieters: If that's the goal. Like if you want the feature to end up in browsers, you need
Simon Pieters: Stakeholders from browsers early on the process, I think, but you also need to have
Simon Pieters: The developers who are going to use the feature, because without them, the browser implementers might come up with a solution that they think is great, but then it doesn't really solve the needs of
Simon Pieters: Of the developers, of the users so
Simon Pieters: And that's something that happens sometimes, there's like a lack of communication early on in the process.
Tom Kralidis: And Andreas and or Daniel, are there, I guess we have, like, you know, a bank of of use cases, throughout the years, one way or another that
Tom Kralidis: That we can we can all work on sort of on the on the open source web mapping side, that we can bring to the table with regards to
Tom Kralidis: You know end user cases. Is that, is that accurate or do we need, you know, specific targeted work to come up with all these use cases because that you know that's effort as well.
Daniel Morissette: Definitely an effort to build the complete list of all the use cases, I believe.
Daniel Morissette: Amelia did some work on that with Peter and I saw some of the work, and it's quite a job to try to identify all the use cases and as Ivan showed in his
Daniel Morissette: His presentation a couple of days ago. You can imagine a bunch of use cases that you've never thought of, so it's it's, it's tough to have all the use cases, but I think it really starts with communication, that's the key thing
Daniel Morissette: Simon pointed out that, you know, get the web and geo people to start talking to each other. That's probably the big thing. I mean, in my case, I had never really talked to any
Daniel Morissette: Web, I mean browser people and until this exact size with the, well this workshop this week, but also some of the work that I did with Peter on using the spec so
Daniel Morissette: That's, I guess, workshops, like this one, and you're getting geo and web people talk to each other is one thing. And then as Terence said in his talk today, speak to users is a good lesson, not just for this exercise, but for specs in general. And if I may share an anecdote on that.
Daniel Morissette: On the OGC side, if we had spoken to users in the WMS 1.3
Daniel Morissette: Working Group when the axis order fiasco was done. For those were in OGC on those days, so the WMS spec was a very, very simple way to create a URL to request a map for a given region.
Daniel Morissette: And to meet ISO compliance, the WMS 1.3 working group decided that we had to abide to a big database of
Daniel Morissette: Projection coordinate system, which meant that latitude and longitude, wwhat used to be longitude and latitude, X and Y, would now be latitude and longitude because when you talk about the
Daniel Morissette: That
Daniel Morissette: The coordinates and degrees, you have to talk about latitude and longitude because that's the way the database has it. Besides, it anyway,
Daniel Morissette: This sounded like a good idea for compliance, but for the users, for the simple
Daniel Morissette: Implementations like the simple WMS clients that were trying to implement the spec, it became a nightmare. And it's been a nightmare for about 10 years of trying to explain to people why does my map not show up at the right place.
Daniel Morissette: I even wrote a rant on my blog at that, I've had that one thought about this. So this is a perfect example of speaker users to validate your
Daniel Morissette: Ideas, maybe not get all the ideas, but at least validate them before you, you adopt your final spec. That's what communication, a good point. So good lesson.
Tom Kralidis: Yeah, I'd have to agree with that, I will say on my side that the the the current evolving OGC API standard, I think, or maybe you're going in the right direction with, with regards to that, by share your pain about the axis order and those and those memories, Daniel, for sure.
Andreas Hocevar: Here and unfortunately, it continues with other OGC standards as well.
Andreas Hocevar: But I couldn't agree more with Daniel about communication here being key and
Andreas Hocevar: I would say compared to other big players on the market, OpenLayers, the mapping library that I'm involved with, is a small player.
Andreas Hocevar: And I think way bigger players like Google with Google Maps or Apple with Apple Maps, and both are way closer to browser development than than any of our first Foss4g projects are
Andreas Hocevar: So it to get those big players on board would be awesome. I don't know what it would take to achieve that. But then the whole movement would really gain momentum, I think.
Tom Kralidis: I think, I think Will has joined, Will, are you online?
Tom Kralidis: Think you're muted.
Tom Kralidis: Just wanna loop Will into the conversation here.
Tom Kralidis: Okay, Will
Tom Kralidis: Oh Will, can you hear me now?
Will: Oh, I'm sorry.
Will: I've got all kinds of
Will: Technical difficulties and it's kind of wonky. So I apologize for being late.
Tom Kralidis: No problem. You just want to sort of midway here, do you want to quickly introduce yourself.
Will: Yeah. So my name is Will Mortenson
Will: National Geospatial Intelligence. Largely what I do is is focus on a lot of open source technology capabilities that are out there, to basically help our foundation mapping missions at NGA
Tom Kralidis: Thanks. Thanks a lot Will.
Tom Kralidis: Um, okay. So we're, you know, carrying on. We hear, we heard a lot about, you know, use cases and and and you know the differences in size between
Tom Kralidis: You know, Foss4G sort of web mapping community of developers, as well as, you know, as well as you know, the bigger, the bigger browser players. We talked about
Tom Kralidis: Process, legitimising ideas, coming up with proposals and working together, but we especially talked about communication, to talk, to talk between these these various communities which is which is obviously key.
Tom Kralidis: What are people's ideas or thoughts around how do we raise, how do we cross pollinate and raise participation between these between these two communities. I mean,
Tom Kralidis: Working on that, you know, open source web mapping software projects is, is you know one type of thing. Browsers, I'm not really sure, but I'm interested to learn
Tom Kralidis: How do we, how do we raise participation of a web mapping folks inside the, inside of the browser projects and you know and and vice versa. You know, beyond, beyond, beyond just the standards. I mean, how can we, how can we lower, lower the barriers to getting those two groups to work together.
Simon Pieters: So I can go first. I have
Simon Pieters: One additional thing that I didn't mention and but I think will help with getting participation or interest from browsers at least, is to try to look outside of the map box and apply use cases for maps to
Simon Pieters: Other problems, right, because the
Simon Pieters: Feature that benefits maps, might also benefits three other
Simon Pieters: Use cases that have nothing to do with maps.
Simon Pieters: And for example, if you take
Simon Pieters: The
Simon Pieters: Panning functionality that maps, have which is kind of similar similar just scrolling, but it's a bit different, as well, because a map you can pan around and then end up at where you started.
Simon Pieters: Which is kind of like infinite or looping overflow and
Simon Pieters: So that's something that maps need, but there are other use cases that also want to use that feature, if you take an image carousel.
Simon Pieters: Where the
Simon Pieters: Part of a page that has images are automatically
Simon Pieters: Flowing by,
Simon Pieters: At the end of that you want the first image to show again. And that's kind of hard to do right now because there is no concept of looping overflow.
Simon Pieters: But finding those common use cases between different needs, automatically
Simon Pieters: gets an interest from
Simon Pieters: Other
Simon Pieters: Areas that
Simon Pieters: aren't even related to maps in the first place.
Andreas Hocevar: That's a very good point, I think, and I can tell from my experience with what people use OpenLayers for
Andreas Hocevar: That one thing that happens very often is, people use OpenLayers for deep zooming into images. So they provide their images in from low to high resolution so they can really zoom in very deeply, OpenLayers has been used for displaying microscopic images from electron microscopes.
Andreas Hocevar: And way more even dashboards for warehouse robots which are maps to some extent, but also not geographic maps in in the sense that we know. So that's, that's a very good point I think.
Brian Kardell: There's a lot of knock on effects to that.
Simon Pieters: Yep.
Tom Kralidis: Are there, are there any opportunities.
Tom Kralidis: Like you know, do browser projects have you know they have code sprints or or hackathons where where potentially some of the, the web mapping crowd can
Tom Kralidis: You know, can, can attend, or can we are there opportunities for colocation. I know in
Tom Kralidis: You know, OSGeo and the Foss4g project community. We have a number of you know code sprints and hackathons to work with software, that's a nice way of cross pollinating between
Tom Kralidis: You know, the, the open source, open source geospatial community. What kind of things are opportunities exist, sort of in the browser side to to to do that kind of thing.
Daniel Morissette: I saw that
Simon Pieters: So,
Simon Pieters: Oh go ahead
Daniel Morissette: Well, I saw that Brian wrote in the chat that again. Yeah, has the as a hack fest. And actually, the, the hackfest for Igalia, I looked it up.
Daniel Morissette: Was taking place the week before the OSGeo code sprint, which was planned for MAy this year and was cancelled. And I think Igalia's event was also cancelled
Daniel Morissette: There may be potential, I was thinking about that potential for getting people from both sides to meet in one of those events, and I'd add to what you said Tom that
Daniel Morissette: In the OSGeo case, we call them code sprints. But really, in my case anyway, I don't write much code I spend a lot of time discussing features,
Daniel Morissette: Not talking about beer. I see you smiling Tom,
Daniel Morissette: Discussing features, discussing specs. Like in the last one,
Daniel Morissette: It happened that I use the one of the last codes springs, OSGeo code sprints to friends to introduce a bunch of people to the MapML stuff and get some feedback, which I brought back to the working group after to Peter, so
Daniel Morissette: It's, you know, even though we call it code sprint. It's not just about code, it's really more about getting coders together
Daniel Morissette: To make decisions.
Daniel Morissette: I don't know about the Igalia ones, or others. It would be interesting to hear about the other type of events on the web browser side to see where we can fit there.
Brian Kardell: Yeah, so I was just mentioning that we have the hackfest that you can definitely you can definitely come to, and what we had a few people from mapping community who were planning on coming to that actually
Brian Kardell: But that that's, that's a good way to do that. There's also other events like there. There have been historically some like extensible web events that are intended to bring developers.
Brian Kardell: And implementers and standardistas all together, and it has like breakout sessions that talk about things and you'll get some fast feedback, but it will ultimately, probably be a lot like the feedback that
Brian Kardell: You've gotten so far, like that. There are
Brian Kardell: useful and interesting things that are smaller, more known quantity things that benefit many, many, many use cases. And those are good steps toward breaking your problem down.
Brian Kardell: That they would be the things to start with probably
Will: I would just add, I'm not a programmer type either, I more gather requirements from customers and users, but I would think that
Will: As a government guy, I don't necessarily know everything that's great and wonderful in the world. And so if I had the ability there, or if we had people that could
Will: From the public side that could stress why certain things are important and get those advocates inside the government that actually have a pocketbook that can afford to
Will: contribute some dollars or some money towards these efforts. I think that would be important, too, because I know it by our particular organization. We have held
Will: hackathons in the past for specific events and we do put out challenges now as well, for certain geospatial type activities. So it's my thoughts.
Brian Kardell: But like, I guess the other thing I was trying to get to is that they're like, if you want
Brian Kardell: People who can help you find avenues in the web platform where there are things that could be specced like there are a lot of people who do that, Amelia
Brian Kardell: Does that and Bocoup does that and the Igalia does that and there's like a small community of people that you can reach out to to do that.
Brian Kardell: Like if you want to actually get implementers time and talk about why things are complex, and like what their costs are and what what kind of technical challenges you would have
Brian Kardell: Igalia's a consultancy and you can like if you want to talk with implementers, we're implemented so you can contract and
Brian Kardell: Talk to us about that.
Will: No, I understand that. But what I'm saying is, I don't know that people inside, I don't know that people have that same technical know how inside of some of these organizations, especially the people that are making decisions. And so my point is, is that, is enlightening some of these folks.
Will: Through events like this, and other types of events,
Will: Writings, documents, things that up to the public, where people throw eyes on that I think would benefit your group
Will: Soliciting more input or soliciting for getting people to be more excited or interested in helping out because like I said, I don't know that everybody has the same technical know how as a lot of you.
Tom Kralidis: So we've got a question from Amelia and the question is based on implementation headaches, what are some of the specific browser features that are missing, which would make a Web Map libraries easier, easier to build.
Andreas Hocevar: In, in my opinion, the thing where
Andreas Hocevar: Process could help most would be
Andreas Hocevar: Rendering API that's more suitable for maps. I mentioned already web GL, which allows you to build very fast, very responsive maps, but it's a very low level API.
Andreas Hocevar: And at least in the OpenLayers project. There's not really anyone with with good enough WebGL skills. Maybe that will change. So we tried several times to implement a WebGL renderer
Andreas Hocevar: But for now, we still stick with the canvas to the API, which is a generic drawing API for for drawing on a on a canvas, arbitrary geometric elements, but a renderer that
Andreas Hocevar: understands how, how maps work especially vector maps where you have geometries and you have styles and you apply styles to those geometry based on attributes and based on your resolution, that's something where existing mapping libraries could be made much smaller by leveraging
Andreas Hocevar: Higher level map specific rendering APIs that might exist in a browser.
Tom Kralidis: On the other side, is the second half of Amelia's question.
Tom Kralidis: Is there anything we can do on the geo side to make to make things easier without necessarily changing the changing the browser.
Tom Kralidis: In terms of maps or features. What are the folks have any, will that have any impact or are we at the point where we need to dive deeper into the deeper into the browser.
Amelia Bellamy-Royds: To jump in, in particular, I was thinking, well, about the issue about latitude versus longitude and those sorts of
Amelia Bellamy-Royds: Inconsistency is between different data sources, that just make everything more confusing or different URL templates and formats.
Amelia Bellamy-Royds: Is that a
Amelia Bellamy-Royds: Solved problem or can we make
Amelia Bellamy-Royds: A lot of the work a lot easier by getting better standardization on the geo format side.
Andreas Hocevar: That's, that's a good question, which is, there's no no cure, yes or no, I would say, I've been following standards and formats that are available that could be used for expressing maps, starting a long time ago with with SVG, which
Andreas Hocevar: Provides a lot of what's interesting for maps.
Andreas Hocevar: But because SVG.
Andreas Hocevar: Relies on the DOM and implements the whole DOM events. It's only suitable for a relatively low amounts of data. And if, I if I look at at a vector, vector tiles. What, what we do today and and consider it as the best way to transport vector data for rendering to the browser.
Andreas Hocevar: There are many similarities that to to SVG, but the gain in performance and
Andreas Hocevar: Better compression of the for-, of the format for transport is because we don't use a markup style language like XML and because we
Andreas Hocevar: Don't need to need to hold DOM and and the DOM events. So I also saw a comment from Ivan Sanchez Ortega, on, on the chat right now where he says that you can get performance flexibility and community standards which without putting more code in web browsers.
Andreas Hocevar: That's, that's definitely true. The only question is how, how long do you want to go API-wise. So the standard answer these days is probably WebGL, which is very common for game developers.
Andreas Hocevar: There are game engines that run entirely in the browser and don't have any performance problem, but
Andreas Hocevar: WebGL is not accessible to to all developers, it's, it's, it has more people familiar with in the gaming community than in the GIS or mapping or geospatial community.
Daniel Morissette: I agree with what Andreas said about new better compression or mechanisms that would allow us to pass data from the server to the browser.
Daniel Morissette: In a more efficient within a verbose, the HTML tree
Daniel Morissette: I mean, we've been dealing with, you know, in digital world we're dealing with large very large data sets. It's been like this, since like, we've always been dealing with very, very large data sets. So
Daniel Morissette: If a mechanism was no like in this case in the MapML proposal, we're dealing with. We're still dealing with markup with issue with an XML type of syntax and the, more efficient than coding would definitely help if there was a standard way of, of having it, of support.
Daniel Morissette: Efficiently streaming or having a progressive loading of binary encoded version of the data that would probably help
Brian Kardell: So, but, like, that's the wire format, but like, how do pixels get to the screen
Brian Kardell: Is a different question and like what
Brian Kardell: Yeah, I mean like inside an element.
Brian Kardell: You would want to base it ideally on something that exists in a platform. And I think question is, what exists in the platform that you would build it off of today and from the presentations here.
Brian Kardell: And just my own like looking around, I could be way off, but it seems like there isn't entirely agreement yet, like are people who want to use Canvas. There are people who are interested in maybe using SVG, because it has some good qualities
Brian Kardell: Like what you send across the wire isn't necessarily what you have to use to render, right?
Peter Rushforth: Can I say something?
Peter Rushforth: Here, I guess I would just pitch the web, you know, the web is ubiquitous, the web underlies everything, and it allows humans to express thoughts, allows companies to express everything they need to express and
Peter Rushforth: The world we live in is everywhere It's everything we touch. It's every place we go with every place we are everything has a place in it, and we need to be able to describe it in much the same way that we describe ideas with the web.
Peter Rushforth: And so we need to leverage an existing standard platform using the same technology.
Peter Rushforth: That integrates into the standard fabric of the web, not the JavaScript layer of the web to describe our world. And I think we've heard that this is a desirable goal in the past couple of weeks from people in the AR and accessibility
Peter Rushforth: Communities and
Peter Rushforth: So, so like
Peter Rushforth: I just want these things to have a plan for for getting there from here. And so, like, that's all, that's all I want to say.
Tom Kralidis: Yeah, when you-
Brian Kardell: When you crack open many of the more complex elements inside our other bits of the platform right, like fetch underscores
Brian Kardell: Like everything that use to fetch data and we have primitives like streams and things like that. And even elements like video, like the players are built up of other elements inside, but you don't like have to send the insides across the wire or something like they can
Brian Kardell: Like they can expand into more stuff.
Brian Kardell: And if you do that, if you, if you separate those concerns. It doesn't mean they can't be the same but it means they could not be the same. So you can split the problem at the joint.
Tom Kralidis: I see one question here from Bryan Haberberger.
Tom Kralidis: Asking about HTTP2. Does anybody. How would, what kind of implications would we have for putting maps in HTML and maps deeper into the browser.
Tom Kralidis: With, from the perspective of HTTP2. Does anybody have any comments or thoughts around that.
Andreas Hocevar: On also only a small bit of experience with with OpenLayers. What really works better with HTTP2 is that a loading data from multiple sources in parallel, like loading tiles for a map is way faster with HTTP2 then with HTTP1, and that's something we benefit from already
Brian Kardell: I can tell you, like as proof of this. Just today I was looking at a Google map on a device that only supported HTTP1
Brian Kardell: And
Brian Kardell: Like there was, you could pull the same thing up on the same device but supporting HTTP2 and the difference is
Brian Kardell: Incredible. It's really, really big deal.
Brian Kardell: But what's the question about 3 because 3 is
Brian Kardell: in sorta process
Simon Pieters: Yeah, the question was about 2, but 3 is indeed the next thing.
Simon Pieters: But,
Simon Pieters: Uh,
Simon Pieters: Ivan said that it shouldn't affect HTML and I kind of agree with that. It makes performance better if you're loading a lot of images but
Simon Pieters: It doesn't make anything else, but you still want to have an interaction
Simon Pieters: In a way that's maps work, with panning and zooming and want to have
Simon Pieters: APIs that that you need for for maps and so on.
Simon Pieters: And that's how those things have nothing to do with what goes over the network that's
Simon Pieters: More what what APIs and primitives are available in the browser to use
Simon Pieters: Or in the web platform.
Tom Kralidis: Another question. So imagine. Imagine a world where we start moving toward getting getting map deeper into into the browser and
Tom Kralidis: Can we think about what the future could look like when, you know, once sort of maps become first class in in HTML and the browsers themselves what are, what is everybody's thoughts around, you know, what does
Tom Kralidis: What does my web mapping software look like, once we once we get there and, you know, how do we how do we see the future of those of those communities moving, you know, moving forward.
Tom Kralidis: So let's say we gave you know the browser, you know, we added the map element. What does that reduce in a web mapping library, let's say.
Andreas Hocevar: But it definitely lets us focus more on what we can do with maps, how we can make nice maps, how we can support users in creating nice maps. If we can take the focus off the rendering and
Andreas Hocevar: The more I think about it, the more I like what has been said earlier in this discussion, that it might help to see a map or a map like thing different from what we are used to
Andreas Hocevar: And and really bring it to the basic concept of what we call a slippy map which is panning and zooming and an image.
Andreas Hocevar: And maybe help with creating image from for images from geometries with attributes which is a bit more specific to maps, but not not really to to geographic
Andreas Hocevar: And with that in place, mapping libraries could probably develop in directions that we that we cannot even think about yet. But this strong focus on the rendering that we have now is definitely that is in the way of leveraging resources for other kinds of developments.
Daniel Morissette: And the same type of evolution happened about 15 years ago when the Google Maps and Bing maps came to the world.
Daniel Morissette: We on the server side, people were telling us, well, you know, your mapping stuff that's gone, that's going to die because now there's maps are everywhere. And actually that just opened up access to maps to
Daniel Morissette: Democratize access. So I guess having maps as a first class citizen and browsers in HTML would probably just open up a bunch of possibilities. Just like you like it happened for the
Daniel Morissette: you know, before those days, producing maps was really a challenge because access to data was hard.
Daniel Morissette: And, you know, for those who just needed to have a couple of points on the map and all the Google and Bing and other service providers just solve the problem for them.
Daniel Morissette: And our server side mapping engines became used for the more interesting and more advanced stuff. So I guess the same stuff with the same situation would probably happen on the client side. Once the browser takes care of the basic stuff as Andreas just said.
Daniel Morissette: You know we developers will be able to focus on on the more interesting features and solve the problems instead of just dealing with performance all the time.
Tom Kralidis: And deal with axis order and all that.
Daniel Morissette: Well, I know. I don't know if that one's ever going to go away.
Daniel Morissette: I mean, that's the difference between geographers and
Daniel Morissette: you know, software engineers. You don't think the same way.
Tom Kralidis: But an example, you know, videos, used to be in Flash, if I recall, right. And then that became an element and you know I'm sure whoever was doing JavaScript around those things was a, you know, they they kept going with interacting with that element instead. Yeah.
Andreas Hocevar: That's, that's a good point. That reminds me of another thing I wanted to say in this discussion.
Andreas Hocevar: All relatively new browser elements that we saw appear in the last couple of years have relative simple ways of interacting with them for a video. It's basically making it full screen, going back and forth, going slow motion, going time lapse.
Andreas Hocevar: The only, the only format that I remember that had more interaction with it was VR ML, which was a standard that died, and I do see this this problem with maps as well. There are so many different ways to to interact with maps.
Andreas Hocevar: And to add controls to maps. So if we think about a standard for that we should maybe leave that to the application or a mapping framework around it and really focus more on the rendering than on controls to interact with the map.
Simon Pieters: I think there are
Simon Pieters: Maps, map implementations can vary quite a bit in their feature sets, like in some maps, it's not enough just to pan and zoom you also want to be able to rotate the map until the nap. Maybe you want to view a 3D view of the map. And how would you do that if
Simon Pieters: It was a declarative
Simon Pieters: Feature of HTML, you wouldn't be able to go through 3D without
Simon Pieters: Using Web GL again. And then the question is,
Simon Pieters: How much do you need to reimplement if you need to do that. Are you still able to do use the panning and zooming
Simon Pieters: interaction model of the browser and forward that to the Web GL
Simon Pieters: Rendering or do you need to re implement the panning and zooming as you
Simon Pieters: Have to do today. We need to be a bit careful about how we design these things and so that
Simon Pieters: If
Simon Pieters: Web developers want to use
Simon Pieters: A solution that is slightly different. They won't need to reimplement everything because that means that the solution wasn't ideal.
Simon Pieters: Does that make sense?
Andreas Hocevar: Definitely yes, because the way you control or you interact with something is always opinionated in some way and
Andreas Hocevar: I think native browser, I meant should, should not be too opinionated in in how you interact with them.
Tom Kralidis: Great. I'm wondering, do we have any other questions from from the crowd or the or the listeners.
Chris Little: Yes, I have uh, Ok.
Amelia Bellamy-Royds: I was just going to read out Fred Esch's comment which kind of could be a question is that he says I don't hear anyone talking about how you relay the semantics of a map to a screen readers. So with the focus is all on improving
Amelia Bellamy-Royds: Rendering
Amelia Bellamy-Royds: Like Web GL and canvas that, where do we get the idea of communicating what the map actually means. And are we still relying on map libraries to somehow do that.
Simon Pieters: Okay, this is a great question and something that we need to seriously consider when designing solutions in the space. >aps should be accessible.
Simon Pieters: Also to people who use screen readers and it should be easy for web developers, in my opinion, to do the right thing.
Simon Pieters: That is, if you
Simon Pieters: Create a map for the web. In the simplest possible way, then it should be accessible by default.
Simon Pieters: In so far it can be, right.
Simon Pieters: But
Simon Pieters: For
Simon Pieters: More custom rendering when you use Canvas or Web GL and then
Simon Pieters: Accessibility is
Simon Pieters: It's hard to get by default, but you will need need to make it possible to layer in somehow
Simon Pieters: I know that Canvas has
Simon Pieters: Already accessibility features that you can leverage with the focusable elements that you can tap between but
Simon Pieters: aren't being rendered on the Canvas
Tom Kralidis: I see a question from Chris Little on augmented reality, saying does the panel think that maps will only take off when AR does?
Simon Pieters: I don't think there is necessarily a connection between that, maps are popular on the web already
Simon Pieters: And I don't see any reason why it would have to wait for AR, but maybe AR takes off before native web maps happens
Simon Pieters: I think it could go either way.
Chris Little: I think, the my thinking behind the argument, the question is lots of people can have a bit fed up with the blue dot not being where they are on the map, and when AR takes off one assume they're going to have much better anchoring and better precision.
Chris Little: Because there's no going to be good infrastructure to
Chris Little: Until the next generation of satellites come up
Chris Little: And maybe that'll give us synergy,
Simon Pieters: Maybe
Tom Kralidis: Any other questions or topics that people wanted to bring up or or discuss? Very interesting panel discussion.
Tom Kralidis: So hopefully one day when we have maps in in in the browser and maps in HTML. I mean, there's many, you know, data providers out there. There's over you know, certainly on the OGC side.
Tom Kralidis: Earlier this week I was at Apache con and George Percivall was mentioned that there's a over a million
Tom Kralidis: OGC websites services out there. So I think as you know as we move forward and making those things easier on the OGC side and making them easier on the browser side, maps can can certainly has a place, core place in the browser.
Peter Rushforth: Can I ask a question?
Peter Rushforth: Yeah, maybe I missed it. It was addressed already but like I heard on Twitter somebody say never in 1 million years, will Google allow maps to be integrated into
Peter Rushforth: not just, you know, maps would never be integrated by the what working group, for the simple reason that Google, Microsoft and Apple have spatial data infrastructures, private spatial data infrastructures and
Peter Rushforth: I personally have
Peter Rushforth: Not got that belief, but can, can the panel or maybe talk about that from an open source perspective.
Peter Rushforth: Can they see
Peter Rushforth: Anything.
Simon Pieters: I think I can talk to that. I mean, I can't talk to what Apple or Google will decide to do but
Simon Pieters: They can't really
Simon Pieters: Gatekeep what goes into HTML,` more than not prioritizing, implementing something in their browser engines, but
Simon Pieters: So the standards themselves are open source and have their own governance models and no one company has full say what what goes into HTML.
Simon Pieters: But
Simon Pieters: Also, the browser engines aren't one to one to
Simon Pieters: These companies. Chromium is an open source project that several companies contribute to and have
Simon Pieters: It has a governance governance model that is
Simon Pieters: Not completely tied to Google. Google certainly has
Simon Pieters: A lot of influence, but they don't have 100% influence over Chromium.
Simon Pieters: With the Igalia also contributing and Microsoft or big contributor also Opera and Samsung, a lot of companies are maintaining this project.
Simon Pieters: And of course, Mozilla is
Simon Pieters: Or the gecko engine that sits behind Firefox is also open source and for WebKit
Simon Pieters: The engine behind Safari.
Simon Pieters: Each of these have their own
Simon Pieters: Teams that decide what what goes into each project. And if something is implemented in two browsers or in two browser engines, then it's defacto part of HTML or part of the web.
Simon Pieters: And at that point.
Simon Pieters: No company can say that it shouldn't go into the de jour standard because it's part of the process to define things that are implemented.
Tom Kralidis: So, I'm sorry, go ahead
Andreas Hocevar: Also, I think.
Andreas Hocevar: This statement on Twitter comes from a point of view that sees a map as, the data behind the map and like in Google Maps,
Andreas Hocevar: Key part of the map is all the information that you show us with a map itself the map content, but we're talking about a way to display data in a map like manner and also OpenLayers is a mapping library that's only about the code
Andreas Hocevar: To display maps, not, not about the data. So you can use OpenStreetMap for a street map or you can use your own data and visualize that. But data is not part of the offering and
Andreas Hocevar: Just the fact that you have a spatial data infrastructure which is more about the data than, than the software, you need to display this data, shouldn't stop anyone from making that part of the of HTML or of what browsers can do
Tom Kralidis: And I imagine, Simon, I imagine these are the open source browser communities or the project's themselves, they, they have like I know in a lot of OSGeo projects we have project steering committees.
Tom Kralidis: I'm guessing there's the equivalent in the in the open source browser browser projects.
Simon Pieters: Yeah and there's
Simon Pieters: Usually a defined process for adding new features that needs
Simon Pieters: Agreement from
Simon Pieters: The
Simon Pieters: API owners or what what they call us for for each project.
Daniel Morissette: Alright, if if there really is pushback until it's hard to tell if it's true or just proceed from our standpoint, if there really is push back from those organizations those companies.
Daniel Morissette: It kind of reminds me of the vendor lock in the world, we had in the GIS world about 20 years ago.
Daniel Morissette: Where it was common practice for GIS software companies to have their own format, their own data format, their own data system
Daniel Morissette: To force users into using the whole suite of their software and prevent them from migrating to another one and open source has been
Daniel Morissette: The way to break that so I guess open source implementations of whatever map standard with we decide on even if there was pushback
Daniel Morissette: From some companies at some point they will they will probably end up being forced to just follow suit because everybody's going to end up using the open source implementation.
Daniel Morissette: A good example of like a standard that did work and nobody could could kill, even though it's simple and could be called flawed, it's the web service, WMS
Daniel Morissette: Every vendor in the GIS world supports it. They have to, they can't not support it and they can't even support it in a way that they break it, because it's too simple. So
Daniel Morissette: And another example. The GDAL OGR, the file format library.
Daniel Morissette: It's breaking the si-, these data silos, by supporting all the file formats and allowing people to go back and forth between the formats and all the GIS vendors, all of them.
Daniel Morissette: Integrate. I don't think there's any one of them that, maybe Manifold, but I haven't heard of Manifold in a while.
Daniel Morissette: But I think all of them do do use the GDAL OGR libraries to read to read each other's file formats. So open source software can be a way to break the silos and
Daniel Morissette: And create that options. So maybe we just don't listen to those who say that there's pushback, we move forward. And if it's good enough people are going to use it and it's going to when
Daniel Morissette: I'm talking about. If the map integration in HTML is good enough. It's gonna win over time.
Tom Kralidis: Okay, we're a couple of minutes over time. Any other final comments from the, from the panelists.
Tom Kralidis: Okay, okay, well, I'd like to thank Daniel, Simo,n Andreas and Will for for your participation in the panel, I would also like to thank the audience for for bringing forward the the useful questions. I think we, we've had a good discussions here and I really think we're at a
Tom Kralidis: Crossroads or tipping point in terms of how to move maps deeper into the browser so open source is obviously a big part of
Tom Kralidis: Of what mapping and so as soon as big part of part of browsers as well. So let's see what we can do to try to marry those communities further in terms of the implementations and obviously the standard moving forward. So thank you to everybody and thanks to the panel.
Peter Rushforth: Yeah I echo that. Thank you very much. Thanks for your participation. Your willingness to play along.
Peter Rushforth: Um, yeah. So that brings us to the end of our
Peter Rushforth: Our workshop, essentially. And I noticed Amelia unmuted. So like if she wants to just push me off the stage here that would be fine. But the, the objective of the next segment of our discussion is to
Peter Rushforth: summarize and conclude and talk about next steps. So that'll be a bit more of an open mic.
Amelia Bellamy-Royds: Yeah, I just
Amelia Bellamy-Royds: Wanted to take a few minutes to talk about logistically, so this is the last of our video conferences.
Amelia Bellamy-Royds: But in a way, it isn't the end of the workshop, because of course, we're setting this up as much as possible to encourage
Amelia Bellamy-Royds: Asynchronous participation from people who couldn't call into the live sessions. So, a reminder that
Amelia Bellamy-Royds: All these sessions are getting posted on YouTube, all the discussion is getting migrated over into theme topics on Discourse.
Amelia Bellamy-Royds: And we're going to try and keep those open for about a month. Well, I mean, they'll stay online indefinitely. But we'll be trying to gather up more discussion and comments and questions over the next month so
Amelia Bellamy-Royds: I would encourage everybody who has been engaged in the discussion so far, if you know of anyone else who would be interested in any of these topics. Definitely point them to the videos, point them to the discussion. Let's
Amelia Bellamy-Royds: keep the conversation going. This is not, not the end of the discussion of maps for the web. It's the start or the start of the next stage.
Peter Rushforth: Yeah, thanks. Amelia. So,
Peter Rushforth: So first of all, I guess I
Peter Rushforth: Want to say thank you to everyone.
Peter Rushforth: For participating over the past two weeks, and
Peter Rushforth: To say that it's a bit of a dream come true to have all these really awesome map, Web Map developers and graphics experts and web standards experts participate in a workshop that could talk about this subject and
Peter Rushforth: I apologize if
Peter Rushforth: I dominated when I did. If I did so, but
Peter Rushforth: It's, it's really inspiring to have had all the inputs and the different perspectives. And so I just wanted to maybe
Peter Rushforth: Reiterate some things that I think I learned from the past couple of weeks, and the one of them is that one of the main takeaways for me is that accessibility is why web maps are actually important.
Peter Rushforth: Potentially more important than they currently are, in fact, and you know maybe accessibility is why augmented reality will be important. One day, and so
Peter Rushforth: I think that that that those are two big takeaways for myself and you know it's going to definitely require more reading and and understanding and of course, empathy.
Peter Rushforth: To to come to a better understanding of these these needs that are
Peter Rushforth: Around web maps and
Peter Rushforth: So I guess, you know, part of the first. First thing I would like to ask the audience is, you know,
Peter Rushforth: Are there other stakeholders that you can identify who should be looped into this discussion. I know we've talked about electron microscopy and images of various sorts paintings and whatnot. Things that are shareable
Peter Rushforth: Like image carousels and so on.
Peter Rushforth: So that's the first thing. And the second thing is do. Do people agree that standardizing web maps is a worthy goal. I know some people do not they think it's a bad idea. And that's, that's a legitimate opinion.
Peter Rushforth: I hear something, dog panting or something.
Peter Rushforth: Okay, thanks. And then the. The other thing you know to become should match become a first class object on the web, like video basically along those lines. So
Peter Rushforth: We need to talk about that, and share opinions and respectfully share opinions on that topic. And so, you know, if
Peter Rushforth: If so, what are the next steps. Should we seek a working group at the W3C, should we see see could joint working group between the W3C and the OGC or should we retrench back to the community group and do more research and development and and
Peter Rushforth: You know, what are the next steps in, in, in the audience's view so
Peter Rushforth: Okay, so I guess, finally, I would like to say that part of part of what I've realized is that, you know, my, my intent was to you know, get big organizations to participate in this in this
Peter Rushforth: Workshop in the hopes that they would say, okay, well that's good and get out a checkbook and start writing checks to make maps on the web, a first class citizen.
Peter Rushforth: But what I realized is that I got that backwards. So what has to happen in my, in my view, and I'll share that with you, before we get our discussion going.
Peter Rushforth: Is that the community needs to approach the big organizations and say this is a worthy goal. Can you give us money to make it happen.
Peter Rushforth: And that's the way that this is get, this gets raised up on the radar of the decision makers in governments and funding organizations.
Peter Rushforth: And, you know, with the with the evidence. I think that's being provided in this workshop that that case can be made, but I'd like to hear other people, views on
Peter Rushforth: what that evidences and the
Peter Rushforth: Coordination that's necessary. So
Peter Rushforth: With that, maybe I'll turn it over to questions if we could you know somehow regulate the flow of questions and comments.
Doug Schepers: I have a an observation. This is Doug
Peter Rushforth: Yeah.
Doug Schepers: I, traditionally W3C workshops, the ones I've helped organize the ones I've attended have
Doug Schepers: Have done a summary of the proceedings
Doug Schepers: At the end
Doug Schepers: It's a little bit less necessary since we've
Doug Schepers: Done videos and have everything available, which is not normally how it's done at W3C, but I think that that might be a really useful next step for the committee and for anybody else who wants to contribute to
Doug Schepers: Make some, a findings report.
Doug Schepers: Some takeaways, future directions, things like that that
Doug Schepers: Could help us set the next stage in the process.
Amelia Bellamy-Royds: Yeah, I should have mentioned that and we're talking about logistics. So when I said, leave the discussion open for one month that was
Amelia Bellamy-Royds: Within about a month, we should start trying to write a report because we have committed to have some sort of short summary report of conclusions, so
Amelia Bellamy-Royds: That is on the table, whether that report is just a summary of what was talked about, or whether it actually has recommendations and specific requests that kind of depends on what happens here. And what happens in the Discourse and whether we have
Amelia Bellamy-Royds: An ability to identify a clear consensus on what needs to happen next.
Peter Rushforth: I sort of whiffed on on the following topic, which was internationalization and a comment that
Peter Rushforth: Brandon Liu made in his panel with Doug was that, you know, a lot of the that the rendering text rendering and rendering
Peter Rushforth: Intelligence built into browsers shouldn't have to be duplicated or repeated and what for, for the benefit of web maps, right, we should be able to leverage the existing
Peter Rushforth: Centries, person centuries of work that has gone into those rendering enter engines and and I wholeheartedly agree with that. And I guess I think I heard that from Andreas earlier about rendering in general.
Peter Rushforth: So, yeah.
Peter Rushforth: So, the floor is open.
Peter Rushforth: People with comments about any of the aforementioned
Bryan Haberberger: Absolutely. Hi, Peter. It's Bryan. Hi everybody it's Bryan. Hi. I just want to say definitely you know we're group that works with the digital humanities, you know, so a lot of cultural heritage museums, libraries.
Bryan Haberberger: It's a major stakeholder in this game, you know, because you can just imagine all of them for any item in any collection. There's a there are coordinates where it's been or is or some other way to, you know,
Bryan Haberberger: Bring it to a level of understanding for someone who's just looking at it out on a map.
Bryan Haberberger: So I look forward to trying to bring this to my group and I know I work with a lot of people who just are doing that basic get dots on a map on on the web and
Bryan Haberberger: Get little implementations and try to do some proof of concepts and be there with feedback because I think
Bryan Haberberger: Especially you know I'm speaking as a developer and definitely coming from the developer mindset, you know, when I will be able to most help you is when I've tried to implement it and I know the little specifics, because
Bryan Haberberger: Otherwise, I'm just kind of coming in with trepidation and worries that I haven't even proved yet so
Bryan Haberberger: That's, that's my immediate goal is to try to find a way to get my little group together and maybe see if we can start to work together and show you how we implement you know map stuff.
Peter Rushforth: Yeah, that'd be great.
Peter Rushforth: Be sure to check out the Web Map custom element polyfill on on the Map4HTML community group.
Bryan Haberberger: Yes, absolutely. I did. And I think most specifically we would end up focusing on that feature element, which I know it's kind of a
Bryan Haberberger: taboo topic a little bit during this
Peter Rushforth: No, not at all, adopt it, make it your own, you know, like it needs a champion like there's just a lot in this proposal, right. So it needs champions.
Bryan Haberberger: I just, I've been looking for more and more evidence of sort of GeoJSON on the web, which is why you know this workshop. It was
Bryan Haberberger: Evident. It is a weird de facto thing that everybody's using and to see that, you know, we took on following it, and that it can be parsed down into your markup languages
Bryan Haberberger: Is great. It just means you know a lot of stuff will just work with that little in between to to figure out how to go okay, instead of throwing it into viewer that's gonna absorb a JSON object. How can I realized that is a, you know, MapML object and show it instead
Peter Rushforth: Yeah, again, it's a it's a proof of concept at this point where before it can be truly be described as a polyfill it has to actually work properly. So
Bryan Haberberger: Yes, and I will test that
Bryan Haberberger: So, thank you.
Peter Rushforth: Thank you.
Amelia Bellamy-Royds: Yeah, those are good points. And it's kind of
Amelia Bellamy-Royds: There are sort of two sides to this story. Um, we had a good discussion today now about rendering issues and low level issues and trying to improve things from that end. But then there's the other end of the people who are
Amelia Bellamy-Royds: Content creators who want to post stuff on the web. And so they're coming from that side of what's the easiest API for them to express their content as something that they want to see on the map and
Amelia Bellamy-Royds: I suspect these things will have to move forward in parallel. So
Amelia Bellamy-Royds: They're already, the map viewer widgets like
Amelia Bellamy-Royds: Leaflet and OpenLayers and the commercial ones as well, but they don't necessarily have the easiest APIs for content creator. And so something like
Amelia Bellamy-Royds: Peter's map element polyfill is basically just, it's a currently done as a wrapper around Leaflet but it's a different API for sending your data to Leaflet and there are other
Amelia Bellamy-Royds: Ways that could be done with Web Components and
Amelia Bellamy-Royds: Just playing around with what the markup could look like. While we're using the existing mapping libraries as the rendering layer, if that's something that could be done as a proof of concept and shipping to people like
Amelia Bellamy-Royds: The, Bryan's museum
Amelia Bellamy-Royds: Work where they don't care about the rendering details so long as it's acceptably decent rendering, but they care about, how do you describe your content and
Amelia Bellamy-Royds: So that's one end and then we also have the other end were trying to get the map frameworks developers and the browser developers and web standards developers talking about improving the rendering layer and improving those lower level features.
Amelia Bellamy-Royds: Sort of two streams of attack I think, that will one day maybe merge in a standardized map
Amelia Bellamy-Royds: Viewer in the browser.
Peter Rushforth: So like, I'm just to hearken back to the open source discussion and the open source to community. We tried to seed the idea in the real world, by creating MapML implementations in GeoServer, MapServer, GDAL, OGR because these are
Peter Rushforth: The, I call them geospatial content management systems because they're clearly a server side component and I'm
Peter Rushforth: You know, I don't think anything is really changed on the server side like GIS is not threatened by maps in HTML. It's just another format, you know, a more natural format like with less friction
Peter Rushforth: In theory, so. Um so yeah, I mean, this is why the ESRIs and the MapBox, I don't believe, should be threatened by maps in HTML, because it's just another
Peter Rushforth: serialization format.
Peter Rushforth: And just the most important format on Earth.
Peter Rushforth: In history, of earth.
Doug Schepers: You mean the most important
Doug Schepers: Feature of on Earth, you mean regarding earth right
Peter Rushforth: No, no, HTML, talking about HTML.
Peter Rushforth: You knew that
Peter Rushforth: You're teasing.
Doug Schepers: Well, I'm saying, you know, it is
Doug Schepers: At least, for a lot of purposes for maps, it is about Earth, is the most
Doug Schepers: Earth related element that could, that we could have
Peter Rushforth: Yeah.
Ted Guild: For sure nothing on Mars.
Ted Guild: Moon to
Ted Guild: For future callings
Doug Schepers: So a Mars, Venus. Yes, yes.
Ted Guild: I'm looking at the various other moons of Saturn myself, as far as to relocate after the election results, but anyway.
Ted Guild: The
Ted Guild: I want to hear more from others. You know how many will appear acknowledge that, you know, there's some sort of it in vocal in in these proceedings, sort of oppose this to the proposal.
Ted Guild: For various people agreeing. That is absolutely necessary.
Ted Guild: And then sort of in between. Some people that want to see.
Ted Guild: Some differences and how it's
Ted Guild: Was using that trying to overload existing
Ted Guild: Elements, for example, of a HTML being being one with, what do others think
Ted Guild: People who are here agree that this is worth pursuing. So we can sort of figure out how and next steps. I think
Fred Esch: It. This is Fred and I've been listening for a while and my background is data visualization charting engine. So in about 15 years and a
Fred Esch: Mapping, charting and Geodyssey community, but my most recent life has been in accessibility and one of the things that I think that you're totally ignoring is
Fred Esch: The semantics, interactivity and accessibility APIs that are needed for any native HTML element. And I just did not hear anything about that.
Fred Esch: And I think widgets are fine. I can't see you convincing, having been on the ARIA Working Group and the Accessible Platforms Working Group. It's really painful to see
Fred Esch: API's added or changed and bringing in a new element seems like a humongous task. I think it really should focus on widgets that do things useful.
Fred Esch: And worry about native element ater you know what all your semantics, how to support interactivity for people that only use keyboards and I mean interactivity not zoom and pan, but
Fred Esch: A investigate the features on the map. I mean, people don't just want to see a map. They don't want to see tiles. They want to be able to interrogate, do you want to do line a site, do you want to do.
Fred Esch: You know, what is the thing next to this in the Northeast direction they want to know spatial relationships and people need to be able to invesigate that like maybe when mouse over
Fred Esch: That you need to do that with keyboards. It's all possible. But I don't think that any, any of the things I've heard, and I did miss the first hour on Wednesday, which is the one I want to
Fred Esch: Just think there's there's big gaps and you're just missing the whole you know
Fred Esch: whole need
Amelia Bellamy-Royds: That's some very good points spread. And of course, we've worked together in the past on
Amelia Bellamy-Royds: Trying to describe the semantics of data visualization and that never quite got into the browsers in any standardized way, but that is something else that needs to happen is, how can we describe
Amelia Bellamy-Royds: Geographic content in a standardized way and how can we describe all the interaction patterns in a standardized way. And I think this is then a third path forward that should probably be happening in parallel, because all the existing map viewers need that.
Amelia Bellamy-Royds: The work that has happened
Amelia Bellamy-Royds: In the discussions about accessibility has made it clear that there is lots of potential and there really isn't the
Amelia Bellamy-Royds: Standard right now that maps on the web as they currently are serving that need, so
Amelia Bellamy-Royds: Whether or not there becomes a standard
Amelia Bellamy-Royds: Map element to create a map viewer widget.
Amelia Bellamy-Royds: Work needs to happen about making the existing map viewer widgets more accessible to more people.
Simon Pieters: I'd like to point out that we
Simon Pieters: Didn't really address accessibility.
Simon Pieters: Today, but this has been covered earlier in the workshop with the accessibility reviews of existing map widgets and also
Simon Pieters: I don't remember what their name was but demonstrating
Simon Pieters: What accessible, accessible maps can look like with the, if you're a blind user with the
Simon Pieters: Touch and audio or vibration which was very insightful, at least for me, and I think it's something that we should consider supporting
Simon Pieters: Somehow
Simon Pieters: By default, if we add maps to the web.
Fred Esch: And Nic did say most of the map widgets fail most of the WCAG to guidelines that she listed. It was pretty
Fred Esch: Terrible. The only other thing I saw, audio, I know that there was a panel discussion on Wednesday, which I missed that I really wanted to see
Fred Esch: But, but the main widget, you put up there. If you have a map element, you should let a keyboard user, go to the northeast to the next feature, if you have one feature and another feature and point features, they should be able to migrate around there, just like you would and see a pop up.
Fred Esch: You know, maybe there's a tooltip type functionality, a keyboard user should be able to do that as easily as a mouse user without making them go through the DOM order. That's just essential
Simon Pieters: Yeah. So I agree, and I think this it
Simon Pieters: ties back to the discussion on use cases and requirements. One of the requirements should be that maps should be accessible, and what falls out of that like
Simon Pieters: You need to have a discussion with the accessibility community, the people who have experience making maps accessible and try to apply that to
Simon Pieters: Whatever we designed for the web so that it's successful
Gobe Hobona (OGC): Hi, this is Gobe,
Gobe Hobona (OGC): I think with this workshop. We've done a good job of
Gobe Hobona (OGC): When I say we, I mean, all the particpants I've done a good job of discussing
Gobe Hobona (OGC): Need to get
Gobe Hobona (OGC): no,
Gobe Hobona (OGC): first class maps in the browser
Gobe Hobona (OGC): I think with future
Gobe Hobona (OGC): Events, we should also try to engage the hardware manufacturers. So the those that develop
Gobe Hobona (OGC): Devices that enable
Gobe Hobona (OGC): People to interact
Gobe Hobona (OGC): So not just like, mouse manufacturers, but all the other
Gobe Hobona (OGC): Devices that enable you know people to interact with applications.
Gobe Hobona (OGC): So making that
Gobe Hobona (OGC): Bridge, you know, from the physical world to the virtual world so
Gobe Hobona (OGC): So maybe with future workshops, we should try and speak to some of those manufacturers. I know in the modeling and simulation community there are a lot of devices that have come to the fore, but also in the research arena like
Gobe Hobona (OGC): Brain to machine, to machine interfaces that are being researched on. So I think there's a lot of options out there that could make for a very interesting discussion on this topic and in future workshops
Peter Rushforth:: Thank you Gobe.
Peter Rushforth: So I'm just going to paste the chat into the chat a Discourse item, where we can share any more of our thoughts about the conclusions and next steps and so on.
Peter Rushforth: On Discourse.
Peter Rushforth: And we're basically out of time, I know it's late for very late for people in Europe and
Peter Rushforth: May be dawning in Asia. But anyhow, so I would like to
Peter Rushforth: Give over to Ted from the W3C to maybe have a word about conclusions and
Amelia Bellamy-Royds: Looks like Wendy from the W3C's...
Peter Rushforth: OK, Wendy, Wendy.
Wendy Seltzer: Well, I
Wendy Seltzer: Don't want to take away from Ted and I'll hand it back to Ted
Wendy Seltzer: Just as strategy lead at W3C I'm excited to hear the discussion here and we'll welcome any follow up that you have for places that we can fill gaps and existing standards work or new work that you might like to see started at the W3C, that the team is here to help.
Peter Rushforth: Thank you very much Wendy.
Ted Guild: Program committee sort of met earlier and try and figure out how and where to structure
Ted Guild: Sort of what, what Peter mentioned that before we, we certainly have the community group in the meantime while we move towards the next step.
Ted Guild: So we don't have to wait until something is set up, we currently have a mechanism to do so. In addition to discourse, there's plenty of opportunities to sort of form these ideas.
Ted Guild: I my big takeaway that there is enough common interests that, this needs to advancement in some form.
Ted Guild: There's certainly gaps, accessibility being a rather significant one that we identify, we need to address better and figure out how to incorporate and
Ted Guild: I guess.
Ted Guild: You know, additional sort of tweaks and features, but that's that's what takes place and while you work on standards.
Ted Guild: And we have a great starting off point thanks to Peter and Amelia and everyone else who's contributed to MapML, and really thank you for putting that on the table, and thank you for
Ted Guild: Peer especially
Ted Guild: Amelia
Ted Guild: Certainly as well in the rest of the program committee, Gobe, Doug
Ted Guild: Ryan and so forth and missing lots of people
Ted Guild: But lots of, lots of people have contributed this, writing proof of concepts that are taking place as well, all the various speakers and there's there's clear interests.
Ted Guild: And we just have to
Ted Guild: refine the proposal.
Ted Guild: But I [dog barking]
Ted Guild: Where it can take place and that was the voice of agreement from my dog, she's giving a shout out to the Peter, she and Peter friends right out anyway, back to you. Here, let me go on you mute while I can.
Peter Rushforth: Yeah. Well, that's good.
Peter Rushforth: All right. Well, thank you all very much for making this a great workshop and, well, we'll see you on the chat forums and online.
Peter Rushforth: And with that. Good night.
Doug Schepers: Thanks, Peter. Bye all
Bryan Haberberger: By everyone will see you around.
Ted Guild: Okay, thanks.
Peter Rushforth: Bye.