 Just early on. This is not me. I do have some vague archery skills, but I'm probably not as good as he is My wife thinks I actually look a lot more like this guy He restores things in the UK apparently some people think I'm him and I am I've been working for a global insurance company JLT for the last 12 years My last day with him is on Monday actually, which is a bit scary in doing that, I've done a lot of different roles and one of them has been Putting together a UI style guide, which is something I've been working on more recently And that's the basis of why I'm here in front of you today Before we get into that though This thing here is a stair ramp combination, which a lot of people call a strap This is going back Rico took us back to the early 90s before I'm going back to the 1960s If anyone was the live back then then they're a lot older than I am Arthur Erikson designed this as part of Robson Square in Vancouver and the concept of this was universal design so that The requirements to to meet everybody's needs were built into the basic architecture Rather than something that's added on and as an exception This is a very famous Design the strap and it's there's lots of replicas around the world This is the first time that was built in the early 70s This is a really complicated poster that explains what universal design is and If I didn't like you I'd stick this on there, but I'll make it a lot simpler The principles are Designs need to be equitable. They need to be able to they need to be used by people with diverse abilities So no matter what? Disadvantages you may have coming in You'll still be able to interact with something Design should be flexible It should be able to be used in a variety of ways according to preference If I'm a left-hander should I not be able to use scissors or maybe you can design some scissors that work with both Should be simple and intuitive The design should make sense without requiring a manual Now this this is something that gets raised quite often we always talk about making designs intuitive in Practice it's not that straightforward when we talk about touch designs We say touch you I was really intuitive you can pick it up and it just works everyone knows what to do But it's not the case the first iPhone shipped with a manual explaining how to pick up the phone and tap it because people Didn't really know what to do. There is a lot of research behind it But a lot of our behaviors that we think are intuitive are really just us adjusting to our environment So always needs to be taken into account Your intuitive design for ten years from now won't be the same for the one that's around right now Design should be perceptible If you can't see it, it doesn't exist With young babies it takes them eight months before they achieve off object permanence, which is when you go like this They realize that you're still the same person there As adults and when we're interacting with something if we can't see that it's there We don't know it's there so Don't hide anything from from anyone and perceptible of course varies according to the different Requirements we've covered before being equitable and flexible If someone is Unable to use their legs then you need to allow them to use a wheelchair or something along those lines Design should be tolerant. Now. I've seen videos of people crying when they get an error message in a web browser you need to be aware of that when you're actually creating something that your Decisions as a programmer or a designer actually have a really direct impact on people's emotions and what we need to do is try to anticipate where people will fail and Ease them through the process so that they don't feel like they're doing a bad job when Half the time it's probably just a bad design in the first place Design should be low efforts. You shouldn't have to do gymnastics just to use something any Photoshop users out there if you've Got used to exporting images for save for web that keyboard combination is a nightmare The the guy who actually came up with that Did it almost as an aside because it's like we need to do something there aren't that many keys left no one's really going to use this thing anyway because he cares about the web and Every Photoshop user ever since has just hated the guy he wrote a blog post about it a while ago But forgive him because he didn't mean to Design should be Within the right size and space so appropriate and available regardless of individual abilities If I'm trying to access a train station and I've got a pram I should be able to get there just as easily as if I'm walking Moving on a few years the web came along Accessibility was a bit of an afterthought unfortunately But we realize now that accessibility is integral to what the web needs to do everything we do needs to be accessible at some point or other Joe Landman says he's one of the the members of the government digital services team in the UK Accessibility is separate from user needs because you might survey your clients and say, you know Do you need it to be mobile friendly? Do you need it to? To work on low bandwidth or anything else like that. Do you need it to be accessible? Don't ask that question because it needs to be accessible They may not know they probably won't know But then they'll hire someone who's got a vision impairment and are they able to use their products? Maybe not and you could affect someone's livelihood based on your decisions And the essence of what Joe says here is design or services should be accessible to all this is the crux of Universal design what we do should be available to everyone the internet is exploding around the world, especially as The next one billion users jump on With the mobile internet and it's really important that we actually make our services available to everyone who's there Now through my presentation. I will link to a number of different sites To save you a lot of effort. You can just go to my blog and I've got all the links in there If not, I'll be posting the URLs as I go anyway That's assuming you do actually want to follow along Now this is a UI style guide that I wrote working at JLT quite recently I've spent the last year or so working on this and the reason that we came out with with it was that There are a lot of different products across the company as I've said before JLT is a global insurance company Global means you have software development teams all around the world And each of them has their own idea of what the colors should be within the company logo and things like that Usually that gets handled fairly well when you have a good brand style guide The concept of a style guide and Rico touched on this before with the atomic design Is to actually have a universal reference For all your code all your behavior so anyone anywhere in the world can pick it up and just implement it and get it right You don't have to think about it one of the The keys when we came up with this was to actually make this as good as it could possibly be That's I think that's that's me to an extent. I don't like doing things half-heartedly So we set out a whole list of rules at the start of things that we want to do and They essentially reflect universal design even though I didn't realize it at the time The first rule is no dependencies when you're starting out a new library It's fairly easy to say this because what else have you got behind you? but No dependencies is a lot more complex than that because what I'm really talking about is Things like this bootstrap jQuery sent your touch jQuery UI you name it if it's a front-end library It's going to affect things at some point or other So we immediately got rid of that Ditching jQuery. I found was really hard. This basically what we did I Thought it was going to be bad. I've been writing web things since the 90s So when JavaScript came out it was exciting. So I've been writing it pretty much since language was first put out there JQuery came along and simplified a lot of things because there are a lot of compatibility issues at the time But they're not really there anymore and it was a big revelation for me to get rid of jQuery And actually move on to jQuery list code again. It's a whole lot lighter You don't have to load 67k worth of a library on someone's browser just to execute something and in the end All your code is really fast as well. So you'll never fail a performance benchmark because you just have no baggage. It's fantastic The essence of this goes down to what Marche here in board says Programmers credo. We do things not because they are easy, but because we think they were going to be easy There were hiccups I had to basically learn how to write JavaScript again because If you've ever done any hiring and so on says they know how to write JavaScript usually they mean they know jQuery and the they're not the same The dom has come a long way since jQuery first came in and you're missing out on a few things if you don't go native Practically speaking element query selector and query selector all take care of most of what jQuery does for you anyway The next and the big thing was adopting aria roles I'd always tried to make my code accessible over the years because I knew accessibility was an important thing But I'd never really tackled a UI library like this before actually getting into complex components So I thought it was really important to Embrace aria, which is the w3 standard on accessibility and make sure that we implemented those roles as well as possible Now aria roles and properties. They're basically an extension on top of html That glue everything together So if we saw with Ruko's talk about how having really smartly named classes Structures your CSS well aria roles actually extend html semantics to set a little bit more Just to give have that meaning that wouldn't otherwise be there if you've got a tab control or a Auto select or something along those lines The key thing with this and the reason that we have the no dependencies first is as Hayden Pickering says You can't have CSS or JavaScript without how html so you have to get your html right first If you're using something that's generated this can be hard But that's we're having a really solid CSS naming scheme helps you immensely I've integrated this library into old SharePoint and it worked So if you're having trouble with it, then don't be afraid if you can get SharePoint to do something you can do pretty much anything This is how I used to write Modifiers so this is using block element modifier or Ben We've got our global tab and the actual tab element Then down below is the selected state for that element. That's what we used to do Getting into aria roles. It actually changed things a bit because you can actually select on the aria states It actually that you're doing something directly with the element not on the classes This strictly speaking is no longer them and the way that I use them is to take the essence of it and then extend it Just to ease off the class congestion that can happen with them So strictly speaking if you are using them, you probably wouldn't do this But this is actually better. I find because it makes a lot more sense Even your JavaScript is a lot more readable as well Here you can see without knowing too much that I'm sending the selected state when I first got into this I got really over zealous and decided that These are your states are awesome. I can just write global selectors for them and Global in CSS is a bad thing. So writing Hidden true display none means that every time you've got something with our hidden true doesn't show What could be so bad about that? The problem is when you want to animate something then you can't just hide it immediately So that didn't really work So I quickly learn the errors my ways and Had a more sensible scoping naming scheme Once you get into Documentation, it's actually not too bad Reading w3c documentation is frightening at the best of times, but the Aria authoring practices guide is actually really helpful because it gets straight into Examples of what you can actually do which is really rare for a w3c thing So you can look directly on how they implement different roles and get an idea of how it's actually meant to work Unfortunately things aren't perfect. So there are other people's work out there hate and pickering. It's got a whole lot of our examples as well That can be a lot clearer than w3c ones. I Did find some of the examples in the area But linked from the area site didn't actually implement the roles correctly, which is a bit strange So I don't know who checked them But just do a bit of searching around you can get some really good Groundwork for what you're trying to do Speaking of which when I first had had to do a hierarchical tree picker I Found Leah veruse awesome, please which is as awesome as it suggests It doesn't quite do what I needed to do But it gave me a lot of the groundwork I needed to get started It's what we had was a really complex hierarchical structure and we needed a quick way of people to select a Particular node within that structure It's a UI nightmare. We've had tons of variations of this over the years until we came up with this I'll just jump to a demo to show you how it works The big problem with this though is there are no area rules for this There's a tree navigation view and there's an auto complete But there isn't something that does both so you need to read within the lines and Once you get into the rhythm of it the area roles make a lot of sense. I'll just show you what this actually does This is a live code demo, which makes me extremely nervous We don't need to see the code Everything's fine This is a really early prototype So when you're looking at how ugly it looks then forgive me if I'm typing in normally I can select things based on My what I've entered I can also use my arrow keys and select things I've not got that working right. I haven't Like I said, this is a crude demo I'll get the keyboard now later on but Combining the two made it really easy to work out what we needed to do one of the strange things with this is you still got focus on the Text input as you're able to navigate through the tree and as you type it filters accordingly But there is there are no guidelines for this. So you really need to go on your own to work it out Skip through those. Okay The next rule is trying to make things obvious and this is a lot of universal design right here Needs to be simple and intuitive perceptible low effort right size and space Text sizing is the first one if you can read this then you're doing really really well and There's not really in that much to say about text sizing Alan It's usually not as good as it could be. I can make it a bit better. It's like you can read that alright, but They bump it up a bit more than it makes a lot more sense and the problem that we have here is We sit at these beautiful screens all day Designing this stuff the people who are using it are not us They're a cricket umpires who use our things and we know they can't see anything So you need to actually accommodate for every user that's out there In essence what you really need to do is bump your font sizes up a couple of points Maybe just 16 point as your baseline you will get some Hostility if you try to do this on a transactional site But it actually works quite well most of what I've found is that you get just change resistance Everyone's used to seeing something in 12 point. They'll complain if it's 10. I'll complain if it's 14 Once you change it to what it is they'll get used to it in a couple of weeks and I'll be fine With some transactional sections you do need to adjust the font sizes according to what data you need to display on the screen That's okay, but just make your baseline that little bit better and your users will thank you for it in the past We used to have Really small interaction elements because when you're using a mouse you can hit a really small target on the screen This study shows that object size can have a very direct influence on the perceived weight of things It's talking about physical things, but it still applies to virtual screens that are designed on the screen How big an element is Determines how likely someone is to interact with it if you have interactable elements such as buttons You need to make them big enough for someone to actually interact with them So the basic rule of this with mouse and keyboard was there for a long time That something needs to be a big enough click target that you can actually reach it 10 years ago pocket super computers became mainstream and The rules changed a bit because the difference between a fine cursor and a finger is quite significant So we had to change we actually had concrete targets here that gave us a better idea of what we needed to aim for and We know what humans are like because we can measure ourselves every human fingers pad size is 10 millimeters So when you're tapping on something you've got 10 millimeters that you can actually work with That's a whole lot bigger than a mouse cursor We can generally hit targets as small as seven millimeters Once you actually take into account screen resolutions and the like You're basically looking at a minimum of 24 pixels By 24 of any given thing that you want someone to interact with Google's old standard is 22 by 22 they've increased that to 24 But this is the minimum the ideal is probably around 36 pixels In order to actually make sure people can tap it at all The thing I found with this though is even after you've done all your measurements You need to test things in a browser. You'll notice with material guidelines for spacing There are clear spaces between each element as well That's just as important as making sure something's big enough to interact with Because when you get onto a real device and you're there with your your clumsy thumb or something like that then How are you using it's going to be a bit different from when you're designing this in a mouse and keyboard situation So testing on devices really useful for this if you're talking about color contrast then Something like it in a Shahara plate like this would come into a lot of people's minds This is to test for color color blindness and the key part of an issue higher plate is that The contrast between the green and the orange hues here is not great Which makes it really hard to see if you're colorblind you can't really work out what's going on If they adjust the contrast color blindness I know you'd still be able to work out what's happening and that's really what you need to know with color blindness the W3C has put together a standard that you need to aim for and they measure the difference between the The contrast of between the text and the background of any given scenario So you might have light text on dark or dark on light or whatever it is But basically you need for something that's less than 20 points in size. You need minimum four to one Contrast ratio if it's greater than 20 you need to aim for a minimum of three to one How you do this is the hard thing I've looked thoroughly in the past and I haven't yet found anything I can give to designers to say hey Just run your designs through this and you'll make sure the colors are going to be contrast friendly Some tools that we're out there in the past don't work anymore. It's a bit of a mess Fortunately on the browser in the things. It's a bit better This is a plug-in for Firefox, which immediately luckily the W3C site actually passes all this that's good But it shows you each element that's on the screen and shows you whether it passes the The contrast tests are not relevant to the size There are similar plug-ins in every other browser that you can think about as well I highly recommend doing this as soon as you actually get someone else's design and have a look at it Run it through this color contrast analyzer and you'll see right away if something needs to be fixed This can be really bad When you have to tell a client that the color of their main logo isn't actually accessible and You need to change it. It's painful But yeah, you have to deal with it sooner or later if you don't then people just can't read what you're doing so make the choice one of the complex things with this is when you have complex background images and This is a site that actually lets you test that out I've found over the years that you get used to it basically so if I'm looking at someone's design I can usually tell if it's going to throw an error or not in the contrast analyzer Until you get to that point you need to use tools like this and even still I don't trust myself because human perceptions are not exactly stable So this shows that the dark blue over on the left and the white is okay But when you're moving over the text doesn't actually work anymore Once again, it's down to the programming end to test this out You could throw this to a designer to get them to test their designs But it's quite ugly at the moment. You still need to have this conversation for its work one of the key things to remember with color contrast is It's not really about People who have vision problems necessarily It's about especially these days with mobile devices using that device in a particular situation If I'm out on my my phone in bright sunlight my contrast is not going to be as good Even here with a projector projectors have notoriously bad color So everything we've got off on the screen is a lot harder for you to perceive than it is on my really nice display here There's never a situation where you can assume that there's not a contrast problem. It's going to pop up sooner or later Because we're using We're using websites and applications everywhere these days It's not just in a locked room in the basement anymore because we're human We don't do anything right Having really strict contrast ratios don't exactly work Even though this is the recommendation from the W3C. It's a good starting point, but it's not everything Now the team at Shopify the UX team actually looked into this because they had the designers complaining I've had this myself. Well, I'll say that it looks okay to me and Here you are telling me this tool says it's not that clearly the tools at fault The computers bad not me always the computers fault In this situation, it's actually true to an extent Humans don't perceive colors in any equal measure We see a lot more green spectrum than we do of any other color because that's how we've biologically evolved What it means in practice is that objectively the Color on the left the black on the red is a whole lot clearer there in the one on the right Actually, that's left and right. Sorry the white on the the red is Is meant that doesn't actually fail the contrast, but it's actually easier to read Pretty close depends on what screen you're looking at and the like but this is where the maths don't actually work anymore and The maths get a whole lot more complex, which is not so great when you're talking about design things The last thing you want to do is talk to a designer and say you need to learn this formula You'll probably get thrown out But all this really basically tells you is There's a lot more to it than basic contrast. This is from a 1992 book on highway designs on signs The signage on highways has to be legible. There's just no way that you can get away with it You can put something up on the web and no one will notice and it's alright If you get a sign that someone can't read and there's an accident someone dies. You're responsible for that. That's not so good and This formula actually shows that our perception is correct At least in so far as the text we think is more legible genuinely is once you take into account some other factors Putting the obvious bits together I'll talk about buttons for a bit because buttons get murdered often and deserve a bit more love When Apple introduced iOS 7 they scrapped all the The touchy-feely, you know make it look like a real-life objects concept and went through the flat design That's pretty much everywhere Microsoft introduced this first with Metro Which is now called something else because they lost the rights for that But Apple were the ones who got the device out. They made everyone actually pay attention Metro actually works really well because the borders between everything are really Really clear whereas in iOS 7 they're not I Don't know how big the tap targets are down the bottom. I have some idea. This is it's a bit of a gray area Because it does depend on where objects are on the screen and how you interact with them so you can Change the rules a little bit, but generally speaking. This is not a great idea and Google followed suit. They copied apples copying of Microsoft to copied it from someone else This is from the material guidelines The first button here the floating action button. It's got a nice shadow on it It's pretty clear to see that I might actually interact with that The raised button in the middle. That's got a slight shadow You can still looks like I'm going to do something the one on the end. What on earth is that? How How do I even know where my fingers meant to go or my mouse? How what's the border of it? What do I need to do? What's the hover state going to be like? I've got no idea. This is in the current standards and you will often get this where someone will say Well Google did it so it must be okay, but it's not it's really not This is an example from materials design guidelines, it's a coincidence that it's Australian I didn't read this in any way On the left is their example Left is their example of what is good On the right is their example of what is bad. I completely disagree with it They've made it quite garish that you could tone down the shadows a bit on the On the buttons that have them, but at least you can actually see what you meant to interact with This is a really bad UI trend. We do know better and we should do better than that what you really need to end up with and is Something along the lines of this This is from the US government design standards, which is currently in writing You can really clearly see how big each element is You know that it's going to be something significant because it has a stark contrast for the background The text is big enough to be legible. It has animated Interaction states so when I hover over if I focus it on it I can actually see them interacting with the element This is where you need to be anything less than this you're making it hard to use Apple have learned their ways with iOS 10 They're getting a lot closer to this kind of thing, but the backlash is there I mean, this is still really getting back to what Microsoft did with Metro all those years ago They've really had it right in the first place. Just when it got copied the the message got a bit lost along the way But the essence of this is actually providing appropriate feedback if I push against the door I know how heavy the door is just for an weight of it. If I interact with something on a computer I don't have that physical thing yet Haptic feedback is still a bit of a work in progress We need to give some kind of visual clue to people that something is going on and that's where this is absolutely crucial keyboard access This is one of those things where earlier on I said we do things because they we think they're going to be easy and Then just not Keyboard access is probably the most important part of accessibility because Great, you've got a mouse to do a lot of things. But what happens if you don't? what if you're you're blind and need to use a brow keyboard or Countless other reasons that you may not be able to use a mouse Even for user convenience enabling keyboard access makes a big difference because using a mouse or time all the time can add a bit of a strain If you're not set up in a perfect economic way and very few people ever are It's not that bad. I thought it was going to be more of a nightmare than it is in practice There are only a few keys that you actually need to support and They are tab enter space and the arrow keys That's only a few keys. It's not that terrible to think about The spec for it is really good as well The W3 the specularia spec actually goes into the detail of what happens when you press each key If someone wrote this as a software spec for you when you received it You'd be over the moon for the detail that goes into it. It's really good So of all the ugly W3C things that are all standard specs generally anywhere This is one of the better ones that's out there It can get confusing trying to work out what's going on which is where examples make a lot of sense and they link to them Sometimes it just takes a bit of playing to work out exactly how The key is interacted with each other trying to remember the different focus states takes a bit of effort It's a bit like playing with animation the first time It's a lot easier than it sounds when you first get into it But once you get things right it makes a lot of sense Starting out with keyboard nav the first thing I actually got stuck into was tabs. I thought how hard can tabs be? Tabs are a lot harder than you would think Fortunately, I mean the first thing I do is basically put into Google tabs Accessible and that's what you should do. Just add area or accessible to your search and it will just Make everyone in the world a whole lot happier Remy sharps tabs came up quite prominently and he's done a lot of work on Looking through the history of tabs on the web how poorly they've been executed and how we can actually do things, right? He set down some ground rules of what he wanted to achieve and He got most of the things right at first he was using jQuery, but then he got rid of that because you don't need jQuery to do tabs There are a few other little bits and pieces, but I took from that and turned it into this Which is I've written a blog post about this as well This is basically a fully accessible working Demo time Tab and it's a tab so it's not going to be the most exciting thing you've ever seen before I'll try to make it interesting somehow Using normal mouse I can just interact you can this is free for concept again So don't pretend these colors are anywhere in production or near it Um That's great. I can select the different tabs if I start using Keyboard I can select them as well So here I can select my tabs based on what I'm doing on the keyboard One of the things that's often missing with tabs is state and a lot of things on the web is you we lack state You refresh the browser and it's like I've forgotten everything about you I'm just going to refresh that and it's remembered There's the storage API that's out there. It's been out there for years. It's really well supported All you need to do is throw a few values into it and you've got something that works You might question what that's got to do with accessibility, but accessibility is Really about design and making things work properly Irrespective of people's needs or requirements Making a tab that refreshes and remembers which tab you've got highlighted is something that should just happen Because it does in other environments on the web. We say that's too hard. It's not going to do it But it's not that hard So a little bit of effort we can actually make these things happen Since my demos work on in the skip through One barrier I ran into with keys was Trying to catch keyboards a keyboard events And I thought having written JavaScript for almost as long as the language has been around I thought I knew keyboard events really well, but I didn't There's the standard w3c way of capturing a key which is vent key and That returns something like up arrow returns a string description of what the thing is enter space something along those lines Unfortunately browser support isn't complete Chrome supports it as of version 51 We're up to version 53 now. I think so it's only recent I 11 and higher do the edge does as well, but 10 and lower don't Safari doesn't support this at all. Although since Chrome support it now, then I assume that Apple will deem it worth copying Basically you need to support the old way which is char code and that returns an ASCII key And If you want to get over an older you need to do this JavaScript is an awesome language because you can just do this in one line Which is fantastic if I have an event dot key Then I'll grab that value if I don't just give me the char code if I don't they give me the key code This captures everything use a switch statement and You're done You'll notice with the non-standard string in the middle This is IE's non-standard interpretation of how things are meant to be because they did theirs before the spec was written properly and Naturally Microsoft can't do anything right exactly the first time around Such credit for them are trying in the first place, but they haven't updated it unfortunately So at this point I need to show a kitten slide every time present something Everything's perfect, which is great My last point which I've touched on already is go with the flow Accessibility is making sure you don't break things that are already there and This is one of the things if you ever work with Disney then The first thing they tell you is you do not touch the mouse Whatever you do you can change the colors in the background you do whatever else you want to do do not touch the mouse For us on the web the mouse is the browser We have our expectations of how we're going to interact with the browser the back button should work The forward button should work. I should be able to refresh it and this is going to happen. It shouldn't be that hard These are the things in the principles of universal design that we need to actually take into account I've already showed you that demo on that I'll finish on this point There's a company oxo that makes consumer products Pealers for example and they were founded in 1990 by Sam father when he saw his arthritic wife trying to use Apeela and she was having trouble So he spent quite a bit of time prototyping you can see all the different versions He went through to get to the final product to make a peeler that actually worked for her because he thought Why can't so on who's got arthritis use a peeler? The end result is far superior than the peelers that were there before and Oxo exists as a brand not because they make things that adhere to universal design, but because they're products good and There are a lot of sticks I could throw at you and say how you might get sued if you don't make something Accessible and so on I could do that, but I'm not going to the carrots are much more enticing If you make your products accessible then you make them better You make them better for everyone that uses them not just the ones who absolutely need them You will be able to find stuff Also, yeah, you know, thanks for a great talk sir You spoke a lot of methods to make things accessible my I have a slightly different question. What have you found any methods to make? Developers designers more, you know how create more empathy towards Users who need, you know, accessible interfaces because I see that like hugely missing Is there anything that you've come across to make that happen? yeah, I've done a lot of work on forms design in the past and If you show someone a video of someone crying when they get an error message, they will change their ways It takes education basically Videos are really good at this There's an Apple engineer who is working on the Apple watch and when you see she's blind and has been from birth when you see her Actually able to do things on her own. It's extraordinary They're countless videos out there You see someone who's got a hearing aid for the first time and if you're not crying by the end of it There's something wrong with you an accessibility That's that's almost the stick level of things that I didn't really get into But that's what accessibility is is about turning spent Personally, I think it's just good practice to do it anyway, but yeah, that's where it can really make a difference if someone's unmoved by Just doing things well then show them a crying child I'm asking a question from the person who's being the bill angle and If they ask you what's the percentage extra that you might have to invest in making it accessible Am I clear in my question? Yeah, so how much extra work is it to make it? Though it should be inclusive of the design process itself, but for someone who has not embraced the process said that yeah the automatic question Well, that's where the stick arguments come in Understand the emotion part but from a business angle when they start out They would probably looking at the money that's into it. Yeah. Well, the money that's into is not getting sued. That's really important The money that's into it is if you adopt these practices at the right stage, which is immediately I've found it actually made my coding practices better I've been a much more efficient programmer by adopting the ARIA standards That's something I touched on earlier where the the ARIA states make a lot of sense selected hidden, etc You don't get those interaction states on normal HTML elements, but you do with ARIA So it makes your code better. It makes you more efficient means that you'll get your product out Really doesn't make any difference. It doesn't add any overhead. Would you have numbers from your experience like this probably I mean You can say related to say at the end it would have cost us so much if you had not done it earlier But it's not a big number to think of say a 10% or a 20% extra effort It takes to be more conscious about these things But it shouldn't take any extra the risk is when you're adopting third-party components. I mentioned bootstrap. I hate bootstrap Bootstrap things if it works well, it would be awesome, but it just doesn't they They try but they just don't get things right and they're their goal is not the same as ours It is really tough when using third-party components relying on them and a lot of developers do I maintain and have done for a long time that it's much easier to write your own code rather than customize someone else's The problem is getting the talent to actually pull that off Yeah, unfortunately, we don't have more time. Please take all your questions to him offline. Thank you You can reach me on Twitter My blog and the slides will be up on slide share later on. Thank you