 Hey everybody, thanks for joining us for this panel discussion on making Drupal more accessible. But we're also going to cover a range of topics. We're going to talk about is Drupal really the most accessible CMS out there? How would we know? What would that mean? What are you excited about in WCAG 2.1, 2.2 and Silver? What should developers know going ahead? How quickly will it change? Three, what is this Bill C81 thing and how does it apply to me? What effects do you hope it will have for creating more accessible Canada? And so if you don't know what Bill C81 is, you're about to find out. And AI, how much is hype? Where can we realistically see AI being used in our web projects in the near future? So yeah, so let's just introduce ourselves. So first, Denny Goudreau. Hello. Do you want to introduce yourself? Oh sure, yeah. So my name is Denny. I live here in Montreal. I've been working in web accessibility for 19 years or so. I'm currently working mostly with a company in the US called Looking Systems. Can you hear me okay? Yeah. I don't think the mic's on. Yeah, that's why I'm wondering. Yeah, I'll just turn it off. So today we're working for this American company called GQ Systems. You may have heard of them. You may have heard of a tool that we developed called AX for accessibility climate testing. So that's one of the things that we do. And a lot of assessments, training and stuff like that. I'm involved with the W3C in developing the accessibility standards, looking into the future of those standards as well. And a bunch of other things related to how we can integrate accessibility in the life cycle in general. That's kind of what I do. And not a brutal user, which is kind of my rank. But I'm here to maybe give a different perspective on some of the things that Mike will be super excited about. And Mike thinks that then you may have the most accessibility-related tattoos that anybody in the entire world. No, you need an iron forearm. Same philosophy. Okay, Mike. So yeah, I'm a triple core accessibility container, along with Andrew Biferson. I'm also the CEO and founder of Open Cuts Consulting. I sit on things like the Open Source Prize Reward for Trigemore, Secretary of Canada. And generally, I'm involved in the accessibility community with the conferences. The first conference that I organized was supported very much by me and Dennis and then came to Ottawa. And I'm a supporter of people learning and sharing information about accessibility through events like the Toronto Accessibility Camp. Thank you. My name is Laura Johnson. I'm a software developer at MyPlanet. I've been a Drupal developer for over 10 years. And MyPlanet does a lot of Drupal. We started as a Drupal shop. Now it's about 50% Drupal projects and 50% sort of work full stack JavaScript. I've written a couple of accessibility-related articles. The last one I introduced any for, that was an article on web accessibility and machine learning. And if you want to read that, it's online. You can find it probably just by Googling it or by finding it on Twitter. I've got a Twitter handle to send out afterwards. And I'm an organizer of Drupal. Okay. All right. So the first topic. Is Drupal really the most accessible CMS out there? How do we know? What would that mean? I've said it many times and nobody has contributed to me and said that there's more accessible CMS. So therefore it must be the most accessible CMS out there. Done. Mic drop. So is that a new one in the question? Well, it is actually interesting. I think that, I think that Colin Burroughs more often released something looking at CMSs and accessibility and CMSs in the wild. And in his evaluation, actually WordPress did better than Drupal on the implementation side of public-facing websites. So that's definitely something that gave us some thought and things to consider about it. But as we know, WordPress completely fell on their face with Gutenberg and none of the back-end of WordPress is now accessible thanks to the Gutenberg project. So Drupal has had a commitment to have the front-end and back-end accessible since Drupal 7 thanks to Everett Zufeld, who was at one point open-causance staff member. It's now in my plan. Yeah, that's great. So yeah, and Drupal also has a couple of cool items that I'm not sure everyone will know about, which is the URL list that Drupal announced is something that's incorporated into Core now, right? Yep, absolutely. So we need to introduce a bunch of stuff in Drupal that helps make it easier for 3D users to be able to create accessible websites. In Drupal 7, we introduced sort of alternatives to CSS display none. And so visually in, this is how it's structured in Drupal 8, we started taking the HTML5 version of the code and brought that to the HTML5 lower-place language and brought that into Drupal to try to broaden the, have a more broad consensus in terms of terminology and make it easier for people to adapt it. But the Drupal Announce function helps you implement area-live functionality. So if you have any dynamic content in your website, your web page changes over time, you need to have something that alerts the screen reader to say that the content has changed. And Drupal Announce is an easy JavaScript helper function that allows you to do that in a clear, consistent way. And again, if there's any, because it's been handled in the centralized location, if the best practice for handling that dynamic content changes, you can fix it in one spot, probably the Drupal community will fix it in one spot, and you will just inherit the best practice with the next step, the next security update with Drupal Core. Right. And that's just it. So if a region of the page changes, then a person using a screen reader is not going to be aware of that change. So you can choose to announce it to them. And I think that you can announce it in passive mode or aggressive mode, in passive, basically, wait until... Not passive aggressive mode. Not passive aggressive mode. No. We haven't even used that yet. Right. So, go ahead. The question I want to ask you is, as someone who's not using Drupal, but is using CMS in general, or especially assessing CMS in general, and what they produce. So you keep talking about accessibility. You refer to screen readers. We hear that a lot. You like to think it's the most accessible CMS out there. So beyond screen readers, beyond maybe even a particular screen reader, maybe using with your computer, what kind of work has it been done to make that client, to be able to do that one? Keyboard access, voice recognition software, other types of the specific knowledge is beyond screen readers. What kind of work has it been done so far? Well, not enough. But I'm confident that we've done more work than anyone else has, which is just sort of a sad state of affairs, is that they're like dragging actually speaking. I would love to go off and find out what some best practices are for implementing dragging actually speaking that we can incorporate into core. I can't find anyone who's got a good case study of what are the best practices for implementing the dragging actually speaking that we can learn from. There should be some. But we're not going to get any help from dragging actually speaking because they don't believe that this is actually an assistive technology device. And the assistive technology users, I think that the dragging users, I don't think are a community that have gathered together and are actively looking at ways to go off and to work as a community. So if we had a dragging actually speaking user in the DRIPPLE community that wanted to give some advice and direction, then we definitely would be able to act on that. But we don't have any. So we're very much limited to who we have access to in the community and who's willing to step up and say, here's the problems that we face. But DRIPPLE has such a large footprint. It's over a million websites around the world. It's 3% of the web. It's driving websites from the CNIB that we built to the RIV to AbilityNet. There's enough organizations out there in the disability sector, in the government sector, in the education sector that have tested and evaluated this tool for their own purposes. We haven't done enough, but I don't think anyone else has done more. And would you feel that, like if you were to try to evaluate, maybe you did. Maybe maybe your team did. Try to evaluate just how much accessibility comes out of the box with DRIPPLE and how much is then required by developers to sort of go to distance to make it accessible. Well, yeah. It's hard to be, it comes down to any software that is open source, you can customize it to meet whatever your requirements are. So just swapping out one theme from another can very easily throw off the accessibility of your site. But what we're trying to do in Core is to build in those accessibility best practices by default. So that what is coming, we can get fun with Core, whether it's in the humanity theme, whether it's in the new admin theme that is being tested for accessibility. Not that it's being fully tested and there are going to be some errors, but that accessibility issues are clearly a bug and things can't get into Core as a stable release until we've addressed that. And that does include things like the layout builder module. A lot of work has been done with the layout builder module thanks to Tim and others to try and to make sure that these new initiatives are building in some best practices for everyone, including keyboard only users and screen users. And there are sort of good examples that you can look to in the Core if you're developing or want to design something well, there are generally examples of accessible ways to do that in the code. Because we do believe that generally developers are lazy, sorry, the developers here, and that we will always look for patterns in the code that we're using to replicate in order to go off and do what we're trying to do as quickly as effectively as to fit in the manner of the patterns we're trying to build. So if we build in good patterns that are patterns that can be used by, those patterns are going to be reused and reused to be used. If you take the approach that so many software companies have done and said we're going to go off and create an accessibility plugin that will bring out your code base from what's in core to an accessible level. Like well, only the accessibility geeks are going to notice and care about that as opposed to everyone and trying to set that default so that the assumptions of the community and the culture is that accessibility is something that is of core importance. So you mentioned earlier about how to do Uber, WordPress 5 and how they come up with their face with that. Is there any effort going on or any idea around comparing the latest version of the WordPress version of the Drupal with accessibility? There's more communication now between the WordPress community and the Drupal community which is good both in terms of security, accessibility and privacy. From a privacy perspective, WordPress is way ahead of us. But in terms of comparison, there really isn't that much even just in terms of reuse of patterns. We're getting to the point where there's a greater level of conversation between our communities and there is a Gutenberg Drupal module. Is anyone playing the Gutenberg Drupal module? No. One person is sort of making a little bit of a... Right. So it's there. And there's also an admin UI team that's looking at building a React implementation of the Drupal admin interface which is facing the same problem that Gutenberg had. That's the challenge they were trying to meet. But again, that core team has a commitment to accessibility and is building that testing infrastructure that didn't work that they're doing. I think leverage and axe core. So there's at least a basic understanding of at least all of the testing. Cool. Not that we're leaving the talk to Drupal, but let's move on and we can refer back to Drupal in the next topic too. So just a quick thing. So I think we're pressing the second most accessible CMS. And I don't know if either third or fourth would be. Do you have any thoughts on that? I don't know if we should be hard-pressed to say. Based on experience, most of the assessment we are seeing or assessing are given by CMS or other people who are aware of this by far. We see Juma every now and then, but there's nothing really worth mentioning there. Right. It's been a while since I've noticed any other patterns that were worth mentioning there. Excessively per se it's not a topic that we often find in product studies, except for Drupal and what they're saying. I've known Mike for about 10 years I think. Since I've known him Drupal was the best one. Version 6 was the best one. Version 8 was the best one. I might have heard that often but I've never seen a benchmark that compares against very specific points to say here's why we can claim this with Gutenberg and what used to happen if the article was conducted by Tenon. They said they're reporting a 329-page document of issues found in that version. There's a lot of problems with it for sure. Someone had a question. We're talking here also a lot of it in their faces, but how about this Drupal community pushing hard on the headless and I'm wondering what you're seeing there because already I went to the headless conference on Wednesday and people were kind of officially admitting that they had to do everything by hand and from scratch. Yeah, so the question is what about headless Drupal which is obviously becoming another CMS which is becoming more and more of a thing and how do we sort of extend accessibility to headless applications. So, I would say that one of the big advantages of Drupal is that there's all these default patterns that have been tested from the front-end and if you're just doing a native React implementation, you're going to have to go off and remake every single pattern in order to go off and do that and React is particularly set up to give you good defaults so you can't necessarily trust the implementation that you're copying to go off and build your prototype is going to be accessible for all kinds of reasons. But the Gatsby implementation I am fairly confident will be giving you a better head start so because Marcy Sutton is there and the rest of the so, there are people who understand accessibility and are committed to that as part of their vision of how Gatsby will evolve and are trying to go off and build in best practices in the documentation, in the code examples, and in the software that they're building themselves to see that there are best practices in place so that developers are here at least following the best practices of Gatsby that you'll be able to implement a reasonably accessible headless front-end for Drupal. Yeah, it would be interesting to follow React as well because React is pretty big in terms of headless Drupal so yeah, it would be great to see if there are good accessible defaults brought into React as well. No more ideas for everything. Use cement in HTML HTML's and build for others. Okay, let's move on to the next topic, which is what are you excited about and look at 2.1, 2.2 and silver. What do developers know going ahead? How quickly will it change? So yeah, so I had a chance to look back at 2.1 a little bit today and I was trying to think of what are the 2.1 the new items that could be potentially addressed in the CMS, for example. Things like label in name so this is where if you have a submit button the label the text that you have in the label needs to match the text that you have in the name attribute so that some of you speaking the I think this is perfect but so someone that speaks, submits and submits, you know, by voice it has to match the actual label or else they won't be saying it right. The name attribute or the accessible name of that object for instance, if you had an ARIA label it would be the equivalent of that. Having a match between what you see visually and what is actually in the code underneath so that if you're speaking to trigger that one button, for instance you call that button by its name that you visually see it refers to something underneath in the source code which there can be a match. Often times we're talking about a match but the situations that arise that were not covered in previous versions of WCAG was that if you had a mismatch or a disconnect between what you were visually seeing and what was given as the accessible name for that button you would try to trigger that button by saying click purchase for instance and purchase would not resonate because underneath maybe the name for the button or the object is by or something like that for instance and therefore there isn't a match and therefore you can't trigger that thing because it doesn't find what you're looking for. With that success criterion we're basically ensuring that there is a match between both representations so that when you speak that particular command it resonates to something it match to something that exists in the DOM. Right. That's an example for sure where you can standardize a lot of that in core or in areas of Drupal. Yeah and in that case I don't know if that's something that we could put an example of into Drupal core So there's for people who are not familiar with the Drupal issue queue there's an accessibility tag that's got all kinds of issues in there you just look at the accessibility issue and you think that Drupal is just like the worst for the Drupal period because there's so many errors that are listed there things that we're working on things that we're aware of things that people have reported and need to be verified We also have a Devacate 2.1 tag that Andrew McPherson has been pushing and they're doing some work to try to make that relevant so we're trying to identify issues that are Devacate 2.1 issues as yet there are no Devacate 2.2 is not yet but at some point when that is released or probably there's some issues that we can identify in Devacate 2.2 there's also an A tag issue queue as well so Drupal's done more than any other CMS I think to go off and deal with A tag which is the author of the accessibility guide on part A and B so that also is something that sets us apart from everyone else and again there's that issue queue for identifying those issues but a term that was I think that the cognitive stuff is really quite exciting with 2.1 I think that has a lot of interesting potentials because I think there's a lot of people in the world that have cognitive disabilities or learning disabilities or language challenges so bringing that into focus is wonderful So your question is what are we excited about and maybe if you don't know what you just said so one of the biggest issues that we had building 2.0 and then 2.1 and now 2.2 and also similar like all these different things is that back in 2008 when 2.0 was released it was the work of the past 7 years building up to a single version of what came about and by 2008 in the couple of years that preceded that there wasn't no such thing as mobile experiences so there was a big lack in what we could cover or predict in terms of the accessibility needs for any mobile experience and then what kept you in this virtual arena where you're trying to get your issues across and you're competing with a bunch of other people that also have their own issues that they want to get see addressed in the standard who ended up winning or basically the bigger lobbies so the American Federation of the Blind the American Federation of the Deaf National Federation of the Blind like the bigger lobbies that were more represented or had more ability to support the effort were still standing in the end and if you've ever worked in centralization you know that it doesn't really matter if you fight really hard at the beginning the person who wins is the person who's still standing at the very end of that process so if I can hate your process then I will drop at some point and whatever issue that they had were never really either brought into the standard or did not make it to the level A or AA which are basically the ones that really matter because everybody pretty much sets their standard to your scope to AA so everyone or most things that were relevant to someone who had a low vision issue as opposed to being blind or someone with a cognitive issue for instance and could not get represented blind people that would do for instance they all ended up either not being addressed or ended up being at the third level or triple A level so for the last 10 years working with WCAG there really wasn't much that was being done more than we could do if we wanted to create an accessible website and yet help people get a low vision or get a cognitive issue because it wasn't part of our scope so what I'm excited about what I'm excited about in 2.1 is the idea that the reason why it exists in the first place is to bridge those gaps introduce new success criteria for mobile experiences for low vision issues and this one success criterion for cognitive because there's only one that ended up making it to this part so not as much as we'd like to but there's a whole backlog of potential success criterion that did not make it to 2.1 that will maybe make it to 2.2 or raise as needs in 2.1 for low vision and mobile app in the press so there's hope that we can cover more of this in 2.2 or 2.x because when we think about the future of WCAG and I'm not getting too silver just yet but the future of WCAG we regularly joke about when we get to WCAG 2.17 we'll still find a couple of things in AR, VR and all these other things eventually we'll need to be addressed as part of this paradigm that is WCAG as we know it today the web content accessibility I don't know if we'll see it today so that's what I'm excited about is not so much whether or not it gets applied more broadly because I think we are naturally we are progressively winning that fight every year there's more people that abide by this and want to do this but this idea that we should have rambled the technologies and bring them in so that eventually they are also covered right on this guy just do people know what silver is no or 2.2 for that matter like I heard it too much but as googling I was like I might need to talk more but it's probably for like a quarter of a minute you're the best person to talk about but let's see it's the fifth of last year so it's just a really year like a week ago and 2.2 is the next iteration of that standard so if you're googling the stuff you might find references to say google 2.x or google WCAG 2.x or WCAG.next for instance it's because we're keeping it we're keeping our horizons open if that makes any sense where we recognize that we could have multiple versions of this standard now what's important to understand when it comes to WCAG is that if today you are debating whether or not you want to do 2.1 or 2.0 they're the exact same thing 2.1 is exactly like 2.0 except it has 17 new specific criteria so it's really not a matter of choosing it's a matter of whether or not you want to introduce the new or since this criteria existed today so 2.2 is an exercise that just started a couple months ago where we're looking at that next iteration and if you're familiar with the HTML5 standard for instance you know that it's now a living standard it doesn't have a version anymore and the idea is that it's a living document so we're leaning towards that mentality as well where we want to have 18 month cycles where we release whatever is already and then we just keep going and going and going so if there's still issues that are raised by people with disabilities when they use web content 5 years from now maybe that will make it into WCAG 2.4 for instance or 2.5 so we don't know how long that's going to last but as long as there's a need to create new success criteria we will, that's basically the idea and the more technologies evolve the more ubiquitous ubiquitous the web becomes as we're using it on a bunch of different devices that are not computers or mobile devices for instance the more issues my are likely to raise as a result of that so the idea will be to address those things again virtual reality, augmented reality there's our own examples of situations where it will bring new challenges that we have not thought about yet the voice interface is the same thing it brings challenges that we have not yet thought about much because up until recently speech impairments weren't really a problem when you're using the web but if all of a sudden if your interface is now a invisible interface and you're only communicating with Alexa for instance, that's still the web but there's an issue with how you can communicate so what are your options I don't know what that is I think instead of that this is all web content and silver is the next generation of accessibility standards to go beyond the web to just talk about accessibility in the digital space much more broadly and it's called silver simply because so what can't expense for web content accessibility guidelines the what can't working group was being named the AG working group last year or the year before so it gives accessibility to the working group because we want people to be on the web and geeks being geeks AG is the chemical in the table of creative table of elements, stands for silver so AG became silver so it became the code name for the standard alternate name is going to be named something else it's not going to be silver for sure but it was enough of an idea that one of the one of the goals in silver is that we go beyond this idea just to check it out because one of the people that I like myself like Mike, like Laura and others who work in accessibility will say regularly is that accessibility is not just a checklist you can just check off things and then you're done you have the whole elistic approach to it that's also important getting them involved and getting them back on that stuff so silver has this goal to go beyond just a set of requirements to also include like an actual exercise in making your content more usable to people with disabilities because you've involved them in that process so getting their own feedback and using that to make yourself better so what we're working with instead of adding like a level A double and triple A as in with AG what we're currently working with we have like a bronze level silver level of gold level and then if you do the equivalent of a WCAG compliant site so it would be like your entry level and then if you started doing other things like involving users in your testing you would eventually go to silver and gold so it was like ISO standard like oh is your process made like some standard process than you guys know so it's broadening it's broadening the scope so it's not just a question of checking off a bunch of items on the checklist but rather creating an experience that will be as usable as possible to people with disabilities so like hypothetically let's say there's a site and you test it with people only as the very last step they would fail right? what they most likely would fail because the issues that you would not be able to predict would come up at that point as opposed to if you introduced people with the process earlier they could find those things earlier and as you know the later the more time it takes for you to find a bug in your system the more costly and complicated it would be to fix it so it's always this idea of shifting left and bringing as early on as possible in that process it's true for accessibility in general if you integrate accessibility in your process at the design level you will avoid a bunch of issues if you only think about it at the end if the process is early enough and they will help you avoid some of the common but it follows the same way so yeah in that sense it would be easier if you did that right there that's great if people actually do that because they won't ever have to I mean much in the same way that when when WCAQ was introduced with this idea level of conformance people were expecting that not everybody would want to go to people pay they sort of expected that W would become the scope for most people or the baseline for most people so it could very well be that in silver if it stays bronze, silver and gold like that most people are going to say you know what bronze is good enough let's just aim for that like the government of Canada for instance with the new law that just came in currently sets up the bar application they could very well at some point say you know what we're going to shift to silver or whatever it's called so you would still establish a baseline and then not a lot of people would go beyond that so same idea has a lot of people most people never bothered looking at the fully satisfied period in WCAQ so it's a double-edged sword I guess but the possibility to do it exists with what I mean will exist and I'm talking probably five years before it's actually there even though some of us are like to say it's going to be there in three years like being part of those meetings every week but five years is probably reasonable as you said being involved in the Drupal core community it takes a lot of patience and perseverance to do that not quite as much as it takes to be involved in the this changes around the Drupal dot org that takes a whole other level of patience but being involved in the WCAQ community and the mailing lists and the discussions that takes a whole other layer of patience and determination especially to have the determination to have sat through those meetings and got to know the people and deal with the struggles for as long as people like yourself have to exist Thank you It just makes a very clear mindset I think and having a boss that pays part of your salary to do that also helps quite a bit because I mean seriously I mean the most important contributors are the ones that are backed by their employers for X percentage of hours a week it's much easier for you to keep fighting you can fight if you're being paid to do it and do that on your own time so and like what I was saying earlier that's what I think for most people that they're fighting for this really important thing that would make a huge difference in someone else's lives in other people's lives but they're doing that on their free time when they have some so they can't keep up and then they end up having the problems at some point so thank you but you know I'm also blessed to have someone who supports me though Yeah, totally Okay, we have too much time left let's just briefly touch on what is the Bill C81 thing and how does it apply to me what effects do you hope we'll have for creating a more accessible Canada So I can start Bill C81 is an act that was proposed by the Liberals that is in the next week or two will be going from Parliament to Rideau Hall and becoming law it had unanimous consent from the from Parliament there's a couple things were added by the Senate I mean it's harder and more difficult so which is good one of the things that we're going to see as part of this is a lot more silent which with official government communications so we'll have ASL, QSL more with the official communications just like we have in the mission branch right now but it's essential to go off and take the the current WCAG WCAG 2.0 AA standards for public-facing websites and say this has to apply to both the public-facing website and the internal-facing backend stuff so your mini interface can no longer be crappy because it has to be accessible you're going to need to build in processes to deal with accessibility and the digital front there's a lot of work on the built environment to try to make sure that that's accessible on transportation, on employment and to make sure that the government of Canada becomes ideally the most accessible public sector employee in the world and it's not just related to the government of Canada this also affects any family related to industry, social security and telecommunications transportation or finance then these regulations apply to you as well and you're going to have to take on many of the responsibilities for supporting people with disabilities both inside and outside your organization by law it seems like the government often buys off-the-shelf software I mean would you say that? So in that sense are they going to have to start looking more at the accessibility level of the software or just demanding that their providers are looking at accessibility or the interfaces I think right now there are a lot of problems So part of the coverage relates to procurement or the procurement and the delivery of products and services in general very similar to what we have in the US for something like this and if you're an agency who builds websites for instance for the federal government you are probably already expected to deliver an accessible website unless your work will come under the government that's completely inclusive of what's been going on they should have asked you for that but up until this point up until C81 it was not enforced in any way it was a standard for the government it was strongly suggested or recommended but that was it as of when the law comes into effect it becomes an actual law so you have to do it so it really brought in a little bit more than what you just did any organization that is that is federally regulated in the public and private sector now needs to abide by these rules the same WCAG 3.0 if you're familiar with the AODA in Ontario this is basically the same thing but extending to the entire country for anything that is regulated by the government basically so any products or anything like that will be part of what that means but banking, transportation borders, parks you perfectly build environment I think ICT and all in a way it is disappointing because I was sort of hoping that the law would say so what they did in Ontario is great they did that across the board for everything but it's really just federally everything in dispute but it's still a pretty good start because another interesting thing about this is that at least in the public sector there's a commitment to go out to hire 5,000 people with disabilities to try and make sure that the government is a leading employer with disabilities I don't think that that employment piece filters into the private sector but it certainly is going to be an impact for the government of Canada I think it's something that the government makes these decisions like the government recently offered and spent a billion dollars on the Microsoft office 360 recently and that's 10 years that the government of Canada had a Microsoft office rolled out across the government of Canada through the new office 360 that's not necessarily a bad thing I think they're probably more cost effective ways to deal with improving accessibility for a billion dollars but they did that because Microsoft doesn't generally make some interesting and useful improvements on accessibility particularly in their Microsoft office suite but I don't know whether or not they enable the accessibility evaluation tool by default so that all public sector employees are their first impression of Microsoft or the new Microsoft word comes with accessibility enabled by default and that the accessibility checker is to be explicitly turned off in order to go off and to do that I also don't think they're doing things like trying to make sure that we're aggregating that information to see that the accessibility you know how inaccessible documents are in each department, both by each staff person and to see what are you doing to try and make sure that the quality of the documents that are being produced by the government of Canada are getting more inaccessible over time so it's always two steps forward, three steps back and there's a lot of work to be done Okay, let's move along to our last topic AI where can we realistically see AI being used in our web projects in the near future so I guess AI and CMS I was going to talk about there's one module that I've seen that's already built called Automatic Alternative Text and so it's I've thought about different ways that this may be possible but it uses no pre-save in no pre-save when a node is saved it checks to see if there are any images in the node that don't have alternative text if they don't it sends the image to Microsoft as and then a caption gets generated automatically so you know obviously that's one way to do it I don't know like what do you think about that I'm not a fan of a lot of AI and I think there's a lot of good stuff in AI, a lot of potential but mostly people are lazy and people think about AI as a way of replacing people and if you're selling AI as a replacement for a human effort then we're not going to get further ahead it's only when we're selling AI as a way to augment or improve and make it easier to create content and actually help coax the user ahead to produce better content that is useful if that automated captioning tool is engaged when images were outloaded to say hey can you approve upon this description of two people standing or three people sitting at a podium can you describe this in more detail to make the appropriate content here's a suggestion make it better than that would be useful but that's not what people are doing like this module goes off and adds alt text when a user doesn't so they're never going to go back and actually add relevant alt text they're going to take whatever crappy alt text the engine has produced and turn that into the default alt text for the image I'm disappointed because I was not in the green light paper then I am I was ready for a fight the same thing I mean, again not a Drupal user type not aware about that works in Drupal but when I hear you say this what I'm thinking is why aren't you just flipping down and then getting that information and then providing it ahead so that the user could eventually update or improve on the alt text whether we're talking about like an animated image recognition algorithm that gets pretty good at describing like every little piece of an image and then conveying that as a as a list basically because that's how it started I think Facebook was the first one doing that in 2016 where if you were loading an image and you didn't have alt text to it it wouldn't break it down into a bunch of things and tell you this would be something like this image may contain colon and then to people smiling blue sky the body of the water and that was it and their goal ultimately is to be able to feed that and work with the information that they know about you to say it's not just to people smiling it's actually your sister and her brother or your brother so feed that because they know stuff about you but again it's always after the fact so I agree with you on that part I happen to be pretty excited about AI in general because I see the potential to make accessibility work easier not so much because I want to be lazy about it but because I know people don't necessarily have that as their priority and if the tools allow you to do these things then you might learn to do that yourself but much like developers today who are thought to use a reactor manual or don't really know how to code manually I don't know they need to learn how to accessibility to create accessible content I don't think you should so if you have a tool that all of a sudden starts generating captions review because it relies on Google's neural machine translating system and then they use that to leverage a particular like you have you have each text and then you turn that into a caption and all that stuff is up and you can automate this caption yourself much like you can do it right now but with a more reliable I guess or accurate level of accuracy much more I mean it's less work real so I see a lot of potential and things like that and the exciting piece about AI is when you look at all of these little tasks this isn't what that is the essence of what artificial intelligence is and what we're doing is you have one very specific task which can be pretty simple but you have to speak it very complex because you don't have the ability to cover all the data that goes with it like image recognition in itself is very simple when you can recognize an image of a cat because you've already documented millions of images of a cat like the data makes this possible so whether we're talking about the recognition software to help you with making the caption on the video more accessible or you're leveraging the ability to transit content or you really recognize images or recognize spaces like in Drupal for instance another thing that can happen is instead of relying on passwords so much what if you rely on the camera you can recognize recognition much like the iPhone X is doing now just based on Apple's data when it comes to that when they introduced the Touch ID so this is Apple claiming something so you have to think about the software of course because they would tell you that the error rate with Touch ID was one out of 50,000 attempts you might think your term was coming regularly but it's one out of 50,000 and with face recognition they brought it down to one out of a million is not working so there's a lot of improvement there and you know you're using your phones to pay stuff or anything you don't have to remember taking your passwords that's convenient for us it's that much more convenient for someone who has difficulty typing for instance so all these things improve accessibility when you have a very very reliable image recognition algorithm for instance that you can introduce in another testing tool like in x4 for instance right now it's very easy for us to parse a page and extract all the images that don't have an alt attribute for instance but whether we send it to a cloud or whether we just we can just have our own bank of images that we draw from that would be open source or whatever I mean being able to use that in the same tool and say all of a sudden you have alt text but it sounds like accurate would you like to review this we could do that in the other tool I would not expect someone that long accessibility so much if the tools were getting them better towards the expected outcome the big area that I think that there's hope for me for our product and I mean there's a lot of things like plain language like the Hemingway app does a great job of trying to go out and produce plain language for English I don't know if there is an equivalent for French to the Hemingway app but it's a really interesting tool again like how do you deal with other languages whether it's Cree, Ojibwe, or Farsi English is a dominant trade language in tech right now so all these tools are being built for English but there's thousands of other languages out there that are not getting the attention and support that they need but I think they might have a lot of hope for looking at data so it's looking at bugs and code open source tools are used all over the internet and there's little effort to go off and take the bugs that are out there and to fix these problems at the source tools like SiteImprove and XCore could be collecting this information then pointing an AI to that data and saying okay where are the bugs are there any patterns that we can see here that we can find in the source libraries out there and can we use these tools to go off and submit tentative bugs on these open source projects so you're able to fix them or at least have somebody look at it and validate it like the XCore bot thinks this is an issue please go off and have a human take a look at this issue with this module because it's used by in according to ranks 10,000 websites around the world and is currently being trafficked by a million people a day go off and visit these websites so please address this if you can that would be a useful thing to go and generate which AI could do fairly easily I think because it is based on data and code and logic and could allow us to more quickly advance and identify some of these issues that's where I have the hope for a lot of AI alright I think none of it's hard and I think we have to wrap it up but let's just as no one's kicking his arches yet open it up for questions do you want to go back for the first part of your presentation when you were talking if you do okay I just want to go back to the first part of the panel where you were talking about if Drupal is the best CMS and it's accessible and then you were comparing two WordPress especially for a YouTuber if you want to compare two WordPress before a YouTuber how do you which one would be more accessible and I would suggest to if you want to compare Drupal 7 because Drupal 8 is definitely more accessible than Drupal 7 there's a lot of functionality added to Drupal so if you want to compare Drupal 7 versus WordPress which one would be more accessible in that way Carl's study basically said that there are more accessible WordPress implementation out there more accessible implementation and that's based on just a random sampling of tools that tend to be used to make it's still something that's really compelling to argue for that and one of the great things the WordPress community has done is they're looking at the WordPress themes they're evaluating accessibility they have a pattern that leads to the saying this theme has been evaluated for accessibility and the client comes up to go and say this one's good we don't have anything like that we don't have anything like that and I'm really impressed by the momentum of the WordPress community there's some really amazing sort of people there and they have regular meetings that they bring people in on a Slack they have discussions, they work through issues Drupal 7 has never had that level of organization that could be a reflection of how I organize things but they would say there's a lot of stuff that's being done there but still, if you're looking at WordPress before they made a commitment to the WordPress community made a commitment to go out and have everyone commit the WK2.0.8.1 but they failed in that, it was just PR and I think that the the Atomatic folks have not really gotten behind accessibility as much as Breeze and the other core engineers in the Drupal community have gotten behind accessibility well, they certainly haven't No, I was just going to say that that's something that I've been wondering is the way that the security team has their little seal of approval to module could we have an accessibility team that has their accessibility seal to themes Yeah, for sure, man So to your point about whether or not Atomatic has taken that seriously if we did not in one way dismiss that issue and it took a huge batch from the community for them to realize that this was actually something important and then we contracted a pen to get that out of it You also mentioned the other thing I want to talk about which is the accessibility themes or the accessibility rate themes in WordPress I, yes, those exist but I don't want it to start to compare, the next thing I don't want it to start to compare how many sites come out as accessible from WordPress or from Drupal Drupal is happening in WordPress how it's about a third of the sites on the web basically So in comparison, without your respect in comparison, Drupal is It's 3% 3% But in comparison it's not as much So when you if you take a sample of a hundred sites you are much more likely to find accessible sites in Drupal than you are that are probably by WordPress just because of the sheer amount of sites that are on that So the comparison is not that fair unless the comparison was made with sites from the same related sector using either one of those tools then you could like, you can go back to sites from like EAS, NRE, etc sites that are either run or made for people that have disabilities or representable disabilities So the transparency of the accessibility in each of those which one is best Because even in the accessibility websites or themes rather in WordPress, and if you use those you can still screw it up pretty bad if you don't know what you're doing So maybe your starting point is not bad but if you start, if you don't know what you're doing you can screw that up pretty quickly So it wouldn't matter if it's accessible it's not a guarantee it will be also a question you mentioned that WordPress with Gutenberg is not as accessible as it was before where you're talking about Gutenberg as an admin interface or that is whatever you update your content why is the code that's published necessary? I was talking about the interface itself I haven't heard any evaluations of how the output from Gutenberg is in terms of accessibility in my control From what I can tell it's still based on the things that you use if you still use accessibility right here you would do pretty well the backlash really was because all of a sudden people that were at accessibility were not able to use the interface the admin once enabled to version 5 So pretty big deal of course if you've been used to the autonomous doing that then all of a sudden you can't anymore Okay, thanks we have to wrap it up so thanks everybody