Categories
Online Marketing

Solving SEO with Headless Chrome (Polymer Summit 2017)

If you manage to pick up on my accent in the last five words, I am indeed Australian and it’s honored to be followed up by Trey, my fellow Aussie, as well prior to joining this team. I’d worked on the beloved chrome dev tools, one of my smallest, but maybe my greatest contribution was adding the ability to rearrange tabs in dev tools, there’s probably the greatest five lines I’ve ever written.

I did work another five other features. So if you find me afterwards feel free to ask me about them, and I might share dev tools trickle to more recently, I’ve had the humbling experience of building web components at all and witnessing all the incredible components that all of you have built and published. For example, the one and only Pokemon selector and if you’re the person who says but there’s a Hanyu, only 151 pokemon in the original set well there’s even an option that lets you set that too, so all kudos to Sammy.

For this, it was, however, in the process of building web components at all, which brings us to what we’re here to talk about today. So, first I’m going to cover my story of how I came to encounter this SEO problem while building web components. Our dog will then look at how I used have less chrome to solve this before diving into all the details of how that actually works and how you can use it. So I’m going to take a step back for a moment and talk about what I learnt in the process of building web components.

A talk. The first thing I learned was how the platform supports encapsulation through the use of web components with this encapsulation comes with inherent code reuse, which leads to a specific architecture. I also learnt about progressive web apps and how they can provide us with fast engaging experiences. I learned how the platform provides api’s such as service workers, to help enable those experiences, as I learned how to compose web components, to build a progressive web.

App we’ve heard from Kevin yesterday about the purple pattern: push render precache lazy load as a method of optimizing delivery of this application to the user and one of the architectures which enables us to utilize. The purple panel is the app shell model. It provides us with instant, reliable performance by using an aggressively cached app shell. You can see that for all the requests which hit our server, we serve the entry point file which we serve regardless of the route.

The client then requests the app shell, which is similar, but because the same URL across the application, we can combine that with a serviceworker to achieve near-instant loading on repeated visits. The shell is then responsible for looking at the actual route that was requested and then request. The necessary resources to render that route. So this point I’d learned how to build a progressive web app using client-side technologies like web components in polymer and how to use patterns such as the purple pan to deliver this application quickly to the user.

Then there’s the elephant in the room SEO for some of these BOTS they’re, basically just running curl with that URL and stop right there, no rendering no JavaScript. So what are we left with with this PWA that we built using the app shell model? We’re left with just your entry point file, which has no information in it at all, and in fact it’s the same generic entry point file that you serve across your entire application.

So this is particularly problematic for web components, which require JavaScript to be executed for them to be useful. This issue applies to all search engine indexes that don’t render JavaScript, but it also applies to the plethora of link rendering BOTS out there. There’s a social BOTS like Facebook and to but don’t forget the enormous number of link renting BOTS such as slack hangouts Gmail, you name it.

So what is it about the app shell model that I’d really like to keep well? For me, this approach pushes our application complexity out to the client. You can see that the server has no understanding of routes. It just serves the entry point file and he has no real understanding of what the user is actually trying to achieve. This allows our server to be significantly decoupled from the front end application, since it now only needs to expose a simple API to read and manipulate data.

The client that we pushed out to the application that we pushed out to the client is then responsible for servicing. This data to the user and mediating user user interactions to manipulate this data, so I asked: can we keep this simple architecture that we know and we love and also solve this SEO use case with zero performance cost? So then we thought what, if we just use headless chrome to render on our behalf, so here’s a breakdown of how that would work.

We have our regular users who are making a request and they would like a cat picture because who wouldn’t and as part of this approach, we ask our robot and to answer this, we look at the user agent string and check if it’s an own bot that Doesn’t render in this case the user can render so we serve the page as we normally would. The server responds with the fetch cat picture function and then the client can go and execute that function to get the rendered result by the way.

This is one of my kittens, which I fostered recently, which is super adorable. Now, when we encounter a boss, we can look at a user agent string and determine that they don’t render, and instead of serving that fetch cat picture function, we fire for a quest to headless Chrome to render this page on our behalf, and then we send the Serialized rendered response back to the bar, so they can see the full contents of the page.

So I built a proof-of-concept of this approach for web components rock and it worked. I wrote a medium post about it, and people really interested in this approach and want to see more of it. So, based on this response, I eventually decided that instead of my hacky solution that I would build it properly but then came the most challenging part of any project and I know you’ve all experienced it as well naming.

So I asked on our team chat for some suggestions and I got a tongue, so these are some of our top ones. There’s some great ones in their power renders use the platform as a renderer. However, today I’m very pleased to introduce render Tron. Let me render that, for you. Brenda Tron is a doc arised, headless, chrome, rendering solution. So that’s a mouthful, so let’s break it down. First off what is docker and why did I use it? Well, no one knows what it means, but it’s provocative in all seriousness.

Docker containers allow you to create lightweight images and standalone executable packages which isolate software from its surrounding environment in render Tron. We have headless chrome packaged up in this container so that you can easily clone and deploy this to wherever you like. So what about headless chrome? It was introduced in chrome, 59 for Linux and Mac chrome 60 for Windows, and it allows chrome to be run in environments which don’t have a UI interface such as a server.

This means that you can now use Chrome as part of any any part of your tool chain. You can use it for automated testing. You can use it for measuring the performance of your application, generating PDFs amongst many other things. Headless chrome itself exposes a really basic JSON API for managing tabs with most of the power coming from the dev tools protocol. All of dev tools is built on top of this protocol.

So it’s a pretty powerful API, and one of the key reasons that headless chrome is great. Is that now we’re bringing the latest and greatest from chrome to ensure that all the latest web platform features are supported with render Truong? This means that net your SEO can now be a first-class environment which is no different, the rest of your users. So just a quick shout out. This all sounds really interesting to you and you would like to include headless chrome in some other way in your to a chain.

There’s a brand new library, node library that was published just last week, that exposes a high level API to control chrome, while also bundling all of chrome inside that node package. So you can check it out on github at google chrome, slash puppeteer, so we’ve looked at the high level of how headless chrome can fit into your application to fulfill your SEO needs now it’s time to dive to how it works.

But I’ve been talking a lot. So, who wants to see render tron in action alright. So this is the hacker news PWA created by some of my awesome colleagues and it’s built using polymer and web components. It loads really fast and all-round performs pretty well. We can see that there’s a separate network requests which loads the main content that we see and we can guess that it’s affected by this SEO problem, since it uses web components which require JavaScript and it pulls the in data asynchronously.

So one quick way to verify this is by disabling JavaScript and refreshing the page, and once we do that, we can see that we still get the app header, since that was in the initial request. But we lose the main content of the page which isn’t good. So we jump over to render Truong the headless chrome service that is meant to render and serialize this for you. So I wrote this UI as a quick way to put in a URL and test the output from render Tron so first off.

What are we hoping to see because these bots only perform one request? We want to see that whole page come back in that one network request. We also want to see that it doesn’t need any JavaScript to do this. So take a look, I’m going to put in the hacker news URL and tell render Tron to render and serialize this and that using web components, and it renders correctly I’m going to disable JavaScript and verify that it still works.

So you can see it’s still there and it all comes back in that single network requests render tron automatically detects. When your PWA has completed loading. It looks at the page load event and ensures that it has fired. But we know that’s a really poor indication of when the page is actually completed. Loading, so Rena Tron also ensures that any async work has been completed and it also looks at your network requests to make sure they’re finished as well.

In total, you have a ten-second rendering budget. This doesn’t mean that it waits 10 seconds, though it’ll finish as soon as your rendering is complete. If this is insufficient for you, you can also fire a custom event which signals to rent Ron that your PWA has completed. Loading serializing web components is tricky because of shadow Dom which it straps away part of the dom tree so to keep things simple.

Rennet ron uses shady Dom, which polyfills shadow Dom this allows render tron to effectively serialize the dom tree so that it can be preserved. In the output, so let’s take a look at the news PWA, which you’ve all seen – and it’s also built by some of my other colleagues and we’ll plug that in to render tron will then ask render tron to render this as well and that I’m also using Web components, and then we have it.

So what do you need to do to enable this behavior with polymer 1? This is super easy and render tron doesn’t actually need to do anything simply append D’Amico’s shady to the URLs that you pass to render Tron and polymer 1 will ensure that shady Dom is used with polymer 2 and with web web components. V1. It’s recommended you use web components, loader jeaious, which pulls in all the right polyfills on different browsers.

You then set a flag to render tron tell it that telling it that you’re using web components, and it will ensure that the necessary polyfills that it needs for serialization get enabled so another feature of render Tron is that it lets you set HTTP status codes, these Status codes are used by indexes as important signals, for example, if he comes across a 404, it’s not going to link to that page, because that will be a really poor search result.

Now server, though, it’s still returning that entry point bar with a status code of 200. Okay, so it looks like every URL exists. Rena-Chan lets you configure that status code from within your PW, a which understands when a page is invalid, simply add meta tags. Dynamically is fine to signal to render on what the status code should be render. Tron will then pick these up and return that status code to the bot, so this approach isn’t specific to polymer or even web components, let’s plug in Fahnestock google.

Com and sees what happens when we serialize it. So that looks pretty good. Who can guess what javascript library was used to build? Google fonts angular render Trond works with any and all client-side technologies that work in Chrome and whose Dom tree can be serialized. The render tron endpoint also features screenshot capabilities, so that you can check that headless, chrome and the load detecting function are performing as you expect.

Unfortunately, this service is not fast for each URL that we render we spin up headless Chrome to render that entire page, so performance is strictly tied to the performance of your PWA. Renat Ron does, however, implement a perfect cache. This means that if we have rendered the same page within a certain cache freshness threshold will serve the cached response instead of rear-ending it again. So how can you get your hands on this today and how do you use it? Well, first, you need to deploy the random tron service to an end.

You’ll need to clone the github repo at Google, Chrome, slash, magnetron, and it’s built primarily for Google cloud. So it’s easy to deploy there. But if you remember this is a darker container, so you can deploy this to anywhere, which supports a docker image. So to make things simple for you to test our. We have the demo service endpoint, which you can hit at render Tron appspot.Com and that’s the one with the UI that we saw earlier.

It is not intended to be used as a production endpoint. However, you are welcome to use it, but we make no guarantees on uptime. Having this as a ready to use service is something we might consider based on the interest receive. So, just in case you’re wondering my boss’s twitter handle is at met, Matt s McNulty, just in case. You want to tell him how awesome I am so once we have that endpoint up you’re going to need to install some middleware in your application to do the user agent splitting that I was talking about earlier.

So this middleware needs to look at the user agent figure out whether or not they can render and if not proxy, the requests through the render tron endpoint, if you’re using purple server, which is a node server designed to serve production applications using purple. You simply need to specify the bot proxy option and provide it with your rennet on endpoint, if you’re using Express, there’s a middleware that you can include directly by saying app, don’t use render on top make middleware with the proxy endpoint and whether or not you’re using Web components, if you’re not using either of these check the docs for a list of community maintained bit aware, there’s a firebase function there, as well as a list of existing middleware that render China is compatible with.

If it’s not listed, it’s also fairly simple to roll. Your own middleware by simply proxying based on the user agent string, and that’s it, that’s all the changes you need to make to use, render tron today and all these bots can now be happy. Brenda Tron is available to use today compatible with any client-side technologies, including both polymer 1 and polymer 2. Thank you.


 

Categories
Online Marketing

Web Fonts – The State of the Web

”, My guest is Dave. Crossland He’s the program manager For Google Fonts And today we’re Going to be talking about the state of web fonts –, what are they, how to use them? Effectively and what’s new, Let’s get started, [ MUSIC PLAYING ], So Dave. Thank you. For being here, My first question is: About why web fonts, What do they bring to a website? Beyond the standard fonts like Helvetica DAVE, CROSSLAND Well, Web fonts really express a certain kind of Feeling for organizations They express a brand And you can have a web Page without a article, but you can’t have a Web page without text You have to have fonts And so a brand at its core Would be like a logo, a color and a typeface or a font, And so web fonts bring The kind of rich design that we have in print Media to the web, RICK VISCOMI And according To the HTTP Archive nearly one third of Websites use a font from the Google Fonts API, So why are developers turning To the Google Fonts API DAVE CROSSLAND, I would say That Google Fonts is fast, easy and free And so on.

Our Analytics page we’ve clocked up over 22 trillion. Font views in total since the service Launched in 2010 – And I think that being on Google’s content, distribution networks, we benefit From cross-site caching, So when you visit the First website that uses a font like Roboto, it’s downloaded and you may see Some latency there, But then on all Subsequent websites, which use the font From Google Fonts, then it’s in a cache And loads instantly across the different websites, We also try and Make it really easy So the font’s API Abstracts, a lot of the complexities of web Font technology from you, So we serve different formats.

To different browsers, For example, with better Compression formats, like WOFF2, only the newer Browsers support those, And so we serve WOFF2 files. To those newer browsers And we serve other Formats to older browsers And then finally We make things free and we have a directory of Hundreds of choices which everybody can choose from Now, of course, if you Want a particular typeface, then it may not be Available in Google Fonts and you would go and license That font for your usage, But not everybody, has the Sophistication in design or the resources To license fonts And I think it’s important That everyone in the world is able to do typography, RICK VISCOMI, So I don’t Know if developers truly appreciate how complicated Web fonts are under the hood.

I got a taste of this when I Was at YouTube a few years ago, I helped change the Default font to Roboto, and it was not as easy as just Changing the font-family CSS style, There’s a lot you need to do to Make sure that it goes smoothly for the users and they Have a good experience, For example, like YouTube users, Are from all around the world, They have different languages. Different alphabets, What are some of the Things that developers need to be concerned about For an international audience, DAVE CROSSLAND International Users face a challenge because the file sizes Of fonts, for them can be larger than just For European languages Traditionally Google Fonts has done a kind of slicing of Fonts into language or writing system sets, So we might have, for example, Latin Latin Extended Cyrillic Cyrillic Extended Greek Greek, Extended and Vietnamese: That’s your current support for Roboto that’s used on YouTube.

We also support Other languages — Hebrew, Arabic, Thai, Many different Indian writing systems And the biggest Challenge has been for Chinese Japanese And Korean fonts, A typical font for Indian languages can maybe be two or three times Larger than a European font, But for East Asia it can Be a hundred times bigger And so we’ve been able to Use a number of technologies, for example WOFF2 Compression which is now a W3C standard this year, And also the @ font-face Css has a new aspect called unicode-range.

Unicode-Range allows us to Slice, the fonts into pieces, dynamically And the browser will Only download the pieces that it needs, So that means that We were able to slice a Chinese, Japanese or Korean Font into over a hundred slices And therefore the Latency of each slice is similar to your European font, This means that the experience Latency is much better And because the Slices are cached across different Domains then the font gets faster and faster To load over time, RICK VISCOMI Custom fonts have Also been used for icon fonts to show images And more recently, they’ve Been used for emoji as well, So we’re moving beyond just Text and on to these other ways that we’re using to communicate, But it’s not without its Own challenges, right, DAVE, CROSSLAND, That’s right! Font technologies are always Improving and evolving And the use of emoji as a kind Of special case of icon fonts is particularly interesting.

I think that there’s a Debate in the web development community about how To best approach this Using images for icons, whether That’s PNG or SVG vector images is –. There are some advantages there. One of the advantages To using icon fonts is that aligning icons with Text in labels is often is a common use case And getting the alignment Onto the baseline of text can be tricky when You’re dealing with two elements: — a text Element and a image element And so icon fonts can Play a good role there.

They also have good legacy. Support because obviously text systems work everywhere. Unfortunately, for Emoji and color fonts that’s a little bit. More complicated because there are Different color formats for different platforms, And so one font file needs To have a lot of data to support all of the Platforms at once And they can look different On different platforms, So yeah emoji as web fonts Is still I think, kind of — is a cutting-edge thing, But it can add.

Consistency – and I hope we see more developments – Of that in the future RICK VISCOMI And going Back to the Roboto at YouTube example: one Of the things I remembered that was kind of tricky Was when we would have font-weight bold in our styles, That would default to Weight 700 by the browser, But our designers decided that It looked best as weight 500, So we actually had to go back. And change all of our styles from font-weight, bold To font-weight 500 And it became kind Of a new way that we had to ingrain into Our style development, But there’s something new.

That’s Coming out called variable fonts, How would they help Address the situation, DAVE, CROSSLAND, Yeah Variable fonts can help a lot. It’s a very exciting. New technology, It’s part of the Opentype standard, which is the font format that that’s Widely supported in pretty much all platforms today And variations allows you To do runtime interpolation between different sort of styles, Or faces within a font family, So traditionally you would Have like a thin weight, a regular weight, a bold Weight and extra bold weight And in CSS you’ve only Had up to nine weights — 100 through 900 With variations, then you are Able to specify weight, 154 and dial in a very specific And dynamic weight You can animate these weight.

Changes using CSS animations And in CSS4 there’s more Direct support for this RICK VISCOMI So does that Mean that every font is now going to be able to be Completely customizable, Or are only a few fonts going To be eligible for this DAVE CROSSLAND Well, it is something that font developers Need to add to fonts, And so in that Way it breaks down the traditional wall between the Font maker and the font user And so variable Fonts create a kind of dialogue between the two, So as a font user, you Can customize the font, but only in ways which the Font maker has provided for, And so that means that you don’t Need to become a type designer yourself, but it means that you Have that flexibility that you didn’t have before And the variations are Not only for font weight, There’s, also font width, There’s slanting And there’s also optical size, And those are all part of The OpenType standard today Optical sizing, means that When you change your font size from 10 point to 70 Point then, the letter forms will actually react and Respond to that change, And so, as your font Size gets larger.

The letter forms will Become more elegant And as it gets smaller They can become more legible, more readable And there’s also all other kinds. Of variations, you can imagine, which aren’t part Of the standard and are specific to each font, Things like rounding and many creative options. Google Fonts is commissioned. To sort of experimental trial fonts from type designer David Berlow at Type Network.

The first is Decovar, which Has a lot of variations which are decorative so rounding Different kinds of serifs different kinds of Stroke patterns And this can be used as a Kind of graphical device, Because variations Can be animated, I think there’s a lot. Of potential there, The other typeface is Amstelvar And Amstelvar is A text typeface and it has a set Of parametric axes which go far beyond Just weight and width and into things like The ascender length descender length and A lot of variations which can be used together, To create more readable text, RICK VISCOMI, I’m Especially interested about variable fonts, We’re going to have to Have you back on the show once they’re a little Bit more established, Then we can talk about The state of them, But where could Developers go if they want to learn more about Any of these technologies DAVE CROSSLAND Microsoft Edge has on their developer site a Really good variable, fonts demo site That’s a great place to learn.

More about variable fonts, There’s also the Design.Google.Com/Fonts articles website, where the Google Fonts team publishes articles about type and Typography in collaboration with the Google Design team And then there’s Also material.Io, where you can get the Material Design, icons, font and learn more about Material Design guidelines, RICK VISCOMI, Well, there you go, The links are in The description so go check them out.

Share your web fonts stories. In the comments below Don’t forget to Like and subscribe so you can tune in For another episode of “, The State of the Web” Every other Wednesday, Thanks for reading and We’ll see you next time, [ MUSIC PLAYING ]


 

Categories
Online Marketing

How We Designed Chrome – Designer vs. Developer #20

What you want? It’s always they want more and more, but they never use those things, and it’s really hard and quite brave thinks they know we’re going to script stuff away. There was internal talk about how chrome is built, and I think back to that time, where Internet Explorer was the dominant browser, Firefox was just was fighting and like the developer, tools were becoming quite prevalent and Safari was was, was just released.

I believe and Google designs to build a browser. So how do you start in that environment? Where there’s so much competition chrome was released in 2008 yeah, but actually we started on it in 2006, oh wow and the team at Google that started on Chrome was actually we were all working on Firefox. When I first joined Google, the beginning of 2005. The idea was to work on making the web better. One way to do that is work on making browsers better.

So we started out as a team working on making Firefox better a year and a half into it. We made the switch to actually building our own browser, and that was a big, big, complicated decision right, because you know we had already. We had been going down a certain path right. So looking back, I think or a number of factors right. First off we thought we could do a really good job, so that had to be true yeah, but also you know there were a lot of things about browsers in those days that I think created, frustrating user experience.

Yeah you got to go about going back 2006. You know applications like Gmail, yeah Maps and YouTube, and so on. These things were becoming popular and other folks were building complicated web applications like this and your typical browser. In that day, if you were to leave Gmail running overnight, you come back the next day and your browser to feel pretty sluggish and bogged down because of just the weight of these applications and so way back then we we had the idea.

That would be really nice to split up the browser into multiple processes. Right operating systems had gone through a revolution from the days of windows, 3.11 to Windows, NT and so on. Yeah we’re pre-emptive multitasking was the thing OS 9 to OS 10. Could we use pre-emptive multitasking? Could we take advantage of actually multiple processes on these systems for web browsing and seemed pretty pretty pretty like, actually seem possible? If you are thinking about a browser from scratch? Yeah I mean in terms of like the UX of again is like going back to the beginning of like browsers or the browser’s of that time.

It reminds me a bit like search before Google’s like search was basically portal sites and the search input field was like almost the most least important thing, but then Google came along. It’s like Nana, that’s the wrong user experience with when chrome came about. It was quite radically different because I know remember this phrase is a Content, not crime, yeah um, so just making that kind of UX decision of like you know, because it was all toolbars and remember when you install anything everything.

It’s all fact you back, then, is very common to find a user with internet explorer and they had installed multiple toolbars. So it’s not just one tool: Bartlett, multiple tours and there’s it’s great absurd. Screenshots of people was, you know those browsers had like five toolbars and it’s not a lot of room for the content right. So one of the things with Chrome’s content now chrome idea was to really remember that the whole point is people want to engage with the web application of the website.

The web content and the browser site try to get out of the way just facilitate helping. You use the web, and so even when we designed the extension system, we resisted the idea of having a first-class way or proper way to do toolbars or sidebars. We really didn’t want extensions over you really. You know using up screen space when that screen space to users really want that for the content. So we designed things, like extension buttons.

That would be the primary access point tried to guide things in a way that would um preserve that notion and even the UI of chrome itself. We tried to keep it very minimalistic. We you know we spent a lot of time in the early days. Thinking, if we’re going to introduce another browser, it’s got to be so awesome right, it’s got it and what does that mean? It’s got to have like the most amazing features.

It’s got it like have a whole new take on browsers. It’s got to be radically different. Ui, surely that would be the reason why we’re doing this right yeah, but in and we tried many different things: putting tabs on the side. You know fancy user gesture kinds of things, Mouse gesture types of things I mean none of that really felt right, and we can do that process. We came to realize what what actually we were doing and what really would set chrome apart is that as a browser, just works better yeah like creating software.

That’s not frustrating is actually hard to do yeah, and I think users appreciate it and so started to think about it, and what does that really mean for us? It was like all products should be pretty simple right should try to try to come up with elegant UI choices. Keep it simple: it should be performant but, like I said, browsers, browsers, have a history of being janky and not well-behaved, and and and you, the user has an expectation when they click on something, especially when it’s the chrome of the UI and when it’s the the Browser UI, they click on it.

They say close this tab. It should close right away yeah, you know par for the course. Those days was. You click close that click to close that tab, and you see you might see a beach ball on Mac, os10, yeah or nothing happens on Windows. You start to see the application not responsive problem right, but in chrome, because we went with this multi-process architecture. We were able to guarantee that if you click close on the tab, it’s gone yeah and those are examples of like responsive UI that you know.

Sometimes, when we talk about performance speed, we mean like how did welded perform on a benchmark, but a lot of times. It comes down to like was experience, smooth, responsive to the user input. Did it actually do what the user wanted it when the user wanted it, that kind of thing, so, simplicity, speed. We also put a big focus on security and stability, so we had these four s’s yeah, and that was the thing that we just repeated to ourselves: if you’re not sure what to work on work on one of those things.

Yes, work on making a simpler design work on making a more performant work on making it. You know more secure so and really with security we mean making it so users feel safe on the web. I feel in control of their privacy. They understand what’s going on, but also that it’s the system is protecting you from malware and so on and again our multi-process architecture not only helped us make something more performant, but also something more secure, a browser more secure and, finally, it helped a lot with stability.

We knew that starting from scratch, with a browser that might actually be the biggest concern, is it going to just crash? Is how do you? How do you exercise enough of the browser in your testing to know that you’ve got it right? We based the browser on not on Firefox, and we based it on WebKit, which is what, at those days that was Safari 2.0 Safari, 3 had just come out and WebKit Safari was known to not necessarily be the most compatible with the web right.

Modern web standards, driven by Firefox, were just becoming a thing. Internet Explorer has had a lot of quirks about it. Internet Explorer 6.0. A lot of quirks, especially thinking about like flowed yeah, that with the flow, though we had a box model. All these things were very impactful to like how web pages were built. If a developer was testing a lot with Internet Explorer, there would be the quirks that they would code to if they were testing.

A lot was Firefox we’d, see that and the Safari it was like. Well, probably, they weren’t testing with Safari, and so it was a big challenge in a big fear. When we launched Chrome, is it going to just crash all the time yeah? How are we going to? How are we going to manage that? So we put a lot of effort and in fact that same issue in forms like our choice of the user agent string. If anybody’s seen the user agent string of Chrome, it’s kind of hilarious because it mentions ever every browser ever since chrome came along.

And that was part of navigating this whole like does it work conundrum we always taught in software development and UX, add more features, because more features means more value, so I mean: was there ever pushback or was there like a fair, maybe we’re taking away too much From the browsing UI, we certainly ugly launched, and it originally chrome, without an extension support, and even the bookmark manager when was was, was revised quite a bit.

I’m going to post the initial beta things like this, so we we intentionally went with a very minimal approach, but we also really encouraged the team to try a lot of things with the idea that, knowing going into it that we would probably throw away things that Aren’t good yeah, that was the I don’t know the mantra if you will like. Let’s just try a lot of stuff and if it doesn’t work it’s okay, we just throw died, it’s not the end of the world.

We don’t have to ship everything we dry. I think that was really liberating and really helpful, because there were a lot of folks on the team who had different had had interesting ideas and and it’s empowering for people to try stuff. But it’s also, you know appropriate that we, we don’t just say because we built it, we should ship it looking back. What would you say were the best decisions you made and also for two part.

What would you regret in terms of like oh yeah, things that you did, that you’d wish you hadn’t? I mean you can also I’m an engineer. I was definitely an engineer at those days and I feel really good about some of the decisions we made. As an engineer from an engineering focus, you know we really put a lot of. We talked a lot about how important was that we were building a product, not building a platform.

I mean ultimately is a product that carries the web platform, but what I mean by not building a platform is that sometimes there’s a temptation as engineers to go off and build framework and and tools for creating the product that you’re actually there to create you. And we really resisted that a lot tried to make sure that we focused all as much of our energy on like actually building a browser which was very helpful to make sure that that that’s what we did so, for example, we said first we’re just building a Windows browser – and that meant, let’s just use win32 straightaway, all the Microsoft API is not looking for any cross-platform toolkit framework to build our UI.

Yes, one day we’ll bring this to Mac one day, we’ll bring it to Linux. You know, and so on, but like for now we’re just building a Windows application and when we went to finally build a Mac product a product for us 10, we told some of the engineers at Google. We said hey, you want to come work with us. We’d love for you to build the best browser for OS 10, and we want you to approach it. The same way that we approached building for Windows, which is all the UI, should be cocoa.

It should all be native, and we want you to have the freedom and flexibility to both embrace the native operating system primitives, but also move quickly as those primitives change, as the iOS evolves. So, let’s build a Mac focused product again with this idea that it’s we’re building product on a platform for building browsers, but what ends up happening as you do this and we did the same thing with Linux.

What ends up happening as you do this? Is you know we start to realize. We were coding same thing three times, yeah right and later on. Things like Android came along and iOS and Chrome OS, and so our world got a lot more complicated and what we ended up doing is, or is this arc from the singular I’m building a product to I’m starting to build platform things that helped me build that Product across and different platforms, yeah and that came afterwards – and I think that was actually somewhat healthy in a bit it.

To a certain extent. I kind of have some regrets that we built Chrome so much as a monolithic product. So while there is some code structure, that’s healthy and good, and – and there is somewhat of a layer cake, if you will there are – there – are some cuts that some some extra layers in the cake that should have been there. And now we have a lot of complexity because we didn’t make some of those cuts earlier.

We didn’t modularize necessarily as much as we should have. But again I think that came from that that focus some were just building this product and he does. I don’t need to be extra. We don’t need all that extra modular modularity, and now we find ourselves wishing he had a maybe done a little more, a little more forethought on that. What would you say, the decisions that were made that were actually really good to the success of a break, yeah yeah, so design examples in engineering examples.

There was this one one concept that was came up very early, which was – and we wrestled with this a bit. So the content area of a tab right, we started with the idea that there are some. We will actually have some browser UI that lives in the tab. So, for example, when you open a new tab, page there’s there’s some content shown to you, suggestions about things. You might want to do yeah. We started out building that natively and we started to find ourselves discovering an uncanny valley, because development users have this expectation that things inside the tab behave like web pages.

But building that not using web technologies meant that some things were subtly not right: yeah selection, behavior, wasn’t there context menus? Not there and the same. You know just things were subtly different, and so we scrapped that and we built the new tab page using web technology, and now it fit better everything we didn’t have all those little niggling little bugs you just felt natural. It felt natural it fit with the product.

On the flip side, we had some dialogues and some of those dialogues, mostly they were built natively, but a few of them were built using web and they never felt quite right, and so then we came to this. Discover that, like, let’s be opinionated about this, if it’s a dialog, it’s done natively and if it’s in the content area, it’s done with the tab, and then we avoid this sort of uncanny valley situation.

When chrome came out, there was a designing for best viewed in Internet Explorer, 6 yeah, and it’s interesting. You say like at the time. Webkit was not the priority of web developers. Now, we’ve shifted 10 years later, we’re seen best viewed in chrome or best viewed in WebKit browsers. So there’s this constant fear that we’re possibly entering back to the past, where, if, if, if development stops, then users and like the web technology becomes like a stagnant, oh yeah, that’s a great question.

I think that oh there’s a couple different things that happen with ie6 right, so, first off Microsoft stopped evolving the guys and we’re not stopping evolving api’s. We our mission, is to make the web better, and so it continued invest in that and the way we invest in that is, it’s very important to work with the standards community, the other browser vendors in particular and web developers, so that we get it right.

One of the dangers of shipping an API, if you’re the only one, only browser shipping it is that you might come to find that there’s a better way to do that. Api, yeah, a better design and then the end result is we’ll be tempted to ship. The new design as well the better design, but we won’t we’ll – have trouble leaving behind the old design so now we’ll ship, two ways to do something yeah or in the worst case three ways to do something.

If you look at CSS gradients, you will see. There’s multiple ways: yeah – and this comes from this – this this phenomenon. Where browser ships it early, then they learn that oh gee. I wish I’d done it differently and then they ship it that way too, and then oh yeah. I wish they would do it differently and they ship it that. Finally – and so you end up with a multitude of ways to do things in the web platform, gets really complicated and we don’t want the bad develop web developer to be thrashed by all of that.

Right, we want to keep it simple and make sure the api’s work well, so we want to do our do a good job, and that means spending time with other browser, vendors spending, time with web developers, learning understanding all the use cases and being very deliberate in The standards process, but we should still be able to ship something. Finally, and sometimes we do have to take some calculated risk yeah right.

Sometimes we are the first browser to ship an API, but we hope to do that in a way that stands the test of time, you’re looking for pain, points and you’re, trying to understand the why it is that people have these problems so that you understand their Mental model and you avoidable, designing in that way again.


 

Categories
Online Marketing

Design Systems with Brad Frost – The State of the Web

My guest is Brad, frost, web designer and author of atomic design and today we’re talking about Design Systems. Let’s get started so Brett thanks a lot for being here. Thanks for having me, I want to show off by asking you: has the metaphor of a web page exceeded its usefulness, yeah, it certainly has, as what designers we’ve been thinking about.

The web is in terms of pages for a long time right, it’s been with us since the web’s beginning right. We scope things out in terms of pages. If things don’t load in the browser says this web page hasn’t loaded and that’s had a really big impact on sort of how we structure our teams, how we scope our projects and how things are actually executed from from a web design and development standpoint. So, for instance, I work with a lot of large organizations and so they’ll have a team, that’s responsible for the home page and then they’ll have a team, that’s responsible for the product page and another team, that’s responsible for the checkout page and all of those teams Are doing things sort of independent of one another right, because they’re just focused on this notion of pages and as it happens, all of those pages are actually made of the same stuff right.

If we were to break things down, you have buttons, you have form fields. You have blocks and cards and heroes, and all these other things – and we end up with whenever you have these different teams working on different pages and thinking about things. In that way, you end up with you know one button looking similar but different than the next team, that’s working on the next page and so on and so forth, and you, you know, repeat that a number of times and span that out over a number of Years and you end up with a giant mess on your hands, it’s not to suggest that we should stop using the term.

It’s probably still useful for users. Yeah only see things as a flat page, but from a design and development perspective. It’s kind of updated yeah. Yeah, that’s right exactly it’s it’s! It still comes together as a cohesive whole and I think, that’s important, especially as people get into talking about design systems. A lot of people have a big misconception that oh design mean you just sort of isolate things at their component level and just designed the button and just design the sort of headings and just designed the card in isolation.

But that’s just not true. That’s you know. It’s important to sort of realize that yeah things do all of those components, do come together and form a cohesive page at the end of the day and that’s what the user sees and interacts with. So it’s important. It’s not an either-or thing, but we just have to be more considerate about how we make the parts of that page as the web and technology as a whole progresses forward.

How has that changed the way that web designers think about serving pages to users and the ways that the websites are accessed yeah? Well from like an access standpoint or from like a design and build process? The fact that a user could be I mean even these days like accessing the web from their refrigerator. You never know the form factor or anything about the user’s device. You can’t make any assumptions: yeah yeah, that’s right and again it’s gotten really complicated and that’s why I think design systems have become as popular as they’ve been because the devices haven’t slowed down right.

The device proliferation is still happening right. The number of contexts – and you know, screen sizes and form factors and, and you know, yeah native web embedded devices different screens. Different sort of you know. Mouse and keyboard touch inputs, and you know voice and, like all this. Other stuff is just the amount of things that users have or that that designers and developers have to consider as they’re, creating user interfaces and creating these experiences I’ve just sort of accelerated, and we can’t keep up right.

We can’t create bespoke pages, for you know: here’s our small screen view and here’s our tablet view and here’s our desktop view. It’s it’s so we’ve had to sort of pull back out a necessity just because we’re on the hook to deliver more features, more services to more users and more context using more devices in more ways than ever before, and it’s like unfortunate. Our budgets have been increased and our resources haven’t increased with that same sort of exponential curves.

So that’s what’s like sort of forced us to sort of step back and and reconsider how this all gets done, given that there are so many different viewport sizes and everything does that mean that the flat Photoshop file is no longer very useful as a means of Conveying the design, yeah, yeah and and still to this day, I’m working in if Photoshop might be a little long in the tooth when it comes to web design, but same thing happens in sketch in figma.

Just last week I got from the clients designers, you know a mobile version of the comp and a tablet version of a competent desktop version of a comp and and a lot of that’s just sort of wasted effort. Really because all three of those things in isolation are sort of one they’re already alive, because it’s a picture of a website not an actual website, but all those spaces in between is where things really fall down right.

You can sort of paint a picture, especially in a in a static design tool where there’s artboards and you could just sort of move things around in free space like that’s, not how things work in the actual browser right. There’s things like some order considerations and all that you can’t just sort of go on this side screen. I just want to move this from here to here, or this I’m just going to swap this around it’s it’s.

It’s really important to sort of make sure you’re. Considering the actual medium that this user interface is going to come alive and and do that much sooner in your process, I want to ask you about concept reviews before called design Det. What does that mean, and how do you avoid going design bankrupt? There’s no design debt and design bankruptcy. I’ve never actually heard design bankruptcy. Before I like that, I I think a lot of places could declare its design bankruptcy.

I think you know just when it comes to design debt. It’s you have. You know number of teams working on different things and just those we were saying you know working on different pages or different products right across a company and you sort of can can take a cross-section and sort of see a lot of discrepancies. Just in that. But that’s just one moment in time when you stretch out that process over time, especially products that have been around for a long time, the googles of the world or eBay or whatever it becomes like a little sort of Benjamin Button.

Like experience as you click through pages, you get further back in time in these older crustier user interfaces, you’re like how did I end up in 1999 and all of a sudden? So so I think that that’s sort of that sort of that visceral feeling of design debt where it’s like you have all of this sort of old stuff that was created. It’s you know once upon a time and that whenever that was launched, it was the new hotness and the new hotness becomes the old crusty experience.

You know pretty quickly these days right so so I think that the more sort of deliberate and the more sort of systematize you could sort of control and wrangle all of those those sort of user interfaces that are, you know out there in the wild. The better. Your chances are going to be as sort of like reducing that that sort of design debt and that’s again, I think, a big crux like that. The crux of design systems is to sort of help.

You know eliminate that debt to basically take those $ 19.99 designs and say: okay, we’re going to update them with a new design language, but we’re going to do it in a very sort of systematic way so that the next time we do a big redesign. We have actual hooks in there that we could actually sort of lift up the quality of in you know so to evolve that design language like flip the switch and roll that out to a bunch of places, sort of simultaneously or or in very short order.

Instead of like, oh, we have to do this big monolithic redesign, and we have to do that for each of our products again and again and again so the developer experience must be a lot better when you can have like a single source source of truth. For your design, but also the user experience as well, could you describe like what it might be like for a user to be on a site that has designed yet yeah? I mean it.

This happens all the time I mean so. The e-commerce example is a great one, just because I think that you know ecommerce sites, you know super sexy homepage or the super splashy super current right. It’s like it’s got the latest. You know shop fall trends, their shop Christmas like coming up or whatever. That’s like you look very campaign driven. So it’s often like a very modern experience. You sort of like click into like that.

Maybe a product detail page or a product category page that sort of feels modern ish. You know it’s like sort of a little bit more meat and potatoes like e-commerce stuff. So it’s like those templates sort of probably feel pretty good, but then, like you, might get to the shop card or if you like, actually log into your account, it’s like those things feel way different and and then you get to the checkout flow.

And then you know that might be sort of way long in the tooth or it might be sort of built by you know an external vendor or something because they’re processing, credit cards and stuff like that. So it might not actually be integrated with like the rest of the site at all. So what ends up happening for? Why that matters from a user experience standpoint? It’s not just about other things, look different like because who cares as long as that’s effective, then that consistency shouldn’t ever be like the number one goal of any of this, and I think that that’s when we talk about Design Systems, I think that’s another misconception as That, oh, we just want everything to look the same everywhere and that’s just really not true, because if your metrics are doing well and stuff – and you know the buttons look different on the checkout page then on the the product detail page, then that’s fine right! No harm no foul, but the problem is, is whenever you’re, a user and you encounter say a date picker or something – and this is a favorite one of mine just because those are hard to build so often times developers just sort of go and grab something.

You know a library they find on the internet somewhere and if you’re, you know say like at an airline or a hotel chain, and you have four different developers grabbing four different date – pickers across the site. Now, all of a sudden, every time the user needs to pick a date, they have to relearn that new library and that, even if it’s just fractions of a second or a second or two or the, where they’re like oh wait, I’m used to booking from the Homepage, but this is a different convention that slows down that process right and that has a negative hit on you know, certainly when you’re talking about you know booking flights or hotels or something that’s going to cause it dip.

So that’s sort of consistency from a user experience standpoint right that ability of like oh yeah. I’ve encountered this pattern before and I know how this works. So I could just sort of roll on and sort of fill things out a lot faster or interact with this thing faster like that’s. That’s what we’re after right, so that consistency for consistency sake not so much, but consistency from a you know, sort of mapping to what users are used to already like yeah.

That’s, that’s! That’s where it’s at one of the people, problems on a design and development team is not sharing the same vocabulary or calling the same components: yeah consistent yeah. So what are some of the problems of that? And how can designers and developers get on the same wavelength? Yeah, so that’s one of the biggest things that I encounter is as an one exercise that I like to do with design development teams whenever I’m working on design systems with them is right out of the gate, we conduct what I call an interface inventory, so we Basically go across their entire sort of suite of products, or you know, whatever user interfaces could be served by their design system and and sort of divvy things up is like okay, you go hunting for buttons, I’m going to go hunting for sort of.

You know input fields or whatever, and then we sort of do that as a group and then what we do is we get together and sort of present what we found to each other and that’s where it’s really fun, because, especially whenever you have designers in the Room developers in the room, QA engineers, business people in the room right like the product owners, like all these different disciplines and you actually sort of have to articulate what your UI is right.

So so somebody will get up and it’s like. Oh and here’s this admin bar and then somebody gets admin bar. We call that the utility bar right and then the developers are like. Oh, we we just mark that up as the gray bar right, and so it’s like. Okay, there we go right. You got everything out on the table right, these inconsistent names for the same thing, and of course that means you have to have again just like that sort of user experience you have to like slow down.

You have to have have a meeting to figure out what you’re going to call this thing like, and you know a lot – can get lost in translation in between design team or different disciplines, but also different teams in general right. If team one is calling it a certain name and team, two is calling it something else. That’s a big deal right, so so again, so bringing this all back to Design Systems. What that it, what a design system can do is sort of centralize your sort of UI patterns call them names right, give write guidelines around them, so that everyone is like, literally speaking, the same language right.

They know what you mean when you say utility bar, and you know how to use it where it’s useful, but also crucial. One of the other things that we found really valuable in in creating design systems for clients is here’s. What this thing is: here’s where it’s useful, but also maybe here’s some gotchas or here’s where it might not be useful, and maybe you want to use this other thing. Instead, what are some of the trade-offs of investing in a bespoke design system versus taking something off? The shelf, like a bootstrap yeah, that’s a big one and I’d say it’s tough, because tools like bootstrap and material design are already made.

They’re they’re, they’re well tested right, they’re in use by giant companies like this company called. Have you heard of Google before it’s like? It’s pretty big one. It sounds familiar yeah, so so a lot of these people right who are using tools like bootstrap and material design, they’re like oh, this has been tested by these. You know giant companies, so I could just sort of grab this and go and I don’t have to do all that work myself and that might be true and there are sort of instances of that um.

I think one of the big things that is important to sort of recognize and consider whenever you’re reaching for these tools is that it’s like you, don’t own it and it might be attractive from sort of you know, inefficiencies sake at first but as time goes on Right at the end of the day, your boss or your you know your product owners or your clients or whoever they are they’re going to say. Oh, we need to do this this way or we need to add this feature and all of a sudden, you’re you’re.

You have to learn and become sort of fluent in this other sort of system that you didn’t write, so so it can work and you can do things and extend things and customize things that works with the grain and these frameworks, but oftentimes. What I found is I work with clients that end up sort of working against the grain and they end up having to sort of undo a bunch of stuff and write a bunch of other custom stuff.

And then they end up in this sort of like weird messy middle ground, where it’s like. This is our sort of hacky stuff that we’ve done to sort of make things our own. But then also crucially, I’ll say that, from like a more of like a front-end architecture standpoint, I think that it’s sort of like safe, you know you got good bones to build upon, but like material design and boots actually offer a sort of anesthetic right.

They provide anesthetic and that could be helpful because again it’s like oh here’s, some good, looking buttons, here’s some good, looking form fields, here’s some good, looking components that I could use, but if Nike Adidas Puma, if you know Reebok whatever, they were all to use bootstrap For their redesigns, they would look frightening Lee similar right and that’s sort of not what they’re going for so there’s like there is this sort of branding aspect of it right this own ability that sort of gets lost whenever you’re sort of all using the same thing.

What are some of the challenges or unsolved problems of design of design? I mean I, I think, sort of specifically to design systems like a lot of there’s some things that are around sort of you know, tooling, and sort of figuring out how to keep design tools and tools, expanding quiff. You know what’s in code, that’s definitely one of the most. I feel like tangible sort of problems that but there’s a bunch of teams, doing a lot of work to try to solve that and startups and stuff that there are really exciting.

And so a lot of them look promising. And I don’t necessarily think that that’s you know far and away the biggest problem. That’s out there. I think so. Many of the problems with with design systems have to do with the sort of people have to do with communication and collaboration and sort of figuring out like how do we get this stuff adopted into our products right? How do we sort of communicate when things aren’t working as planned like? How do we sort of you know, establish solid processes for releasing new versions of the design system and letting everyone know like here’s one? You want to use the design system or here’s one.

It’s sort of safe to sort of you know, deviate from that system or build upon it or extend it, and how do you roll that back into the system? So a lot of that sort of coordinating a bunch of different people who are all suddenly relying on this, this design system product that stuff, I feel is – is still very tough to crack because it involves people and your you know the health of your your you Know design and development culture and like how well everyone sort of you know, collaborates together and like, of course, that’s that’s tricky right, so you could like you.

Could I could say things like here’s how I would create a governance plan for a design sister for a design system and here’s how I would you know, get these teams to work. You know and communicate more buts and you know easier said than done. Okay, so how much of a design systems success depends on the designers as opposed to the developers? What is their role in the success of it? I think, and – and this might be a little controversial design systems is sort of an unfortunate name because design systems are like.

Oh, this is about design, and it’s really not. The design system is, as I define a design system is, is how the official story of how an organization designs and builds tadaryl products and there’s a lot of ingredients to that story. And yes, like the design language, you know what what the brand colors are, and you know the the rounded corners or not of the buttons and stuff like that sure that that matters.

But that’s actually like a pretty tiny slice of what a design system entails, and so so when it comes to the success of a design system. So much hinges on that design system living in code and living as a thing that engineers and developers can sort of pull down into their application and sort of you know import a component and sort of see that design systems button or whatever show up on their Screen and then they’re able to sort of you know pipe in whatever sort of attributes and click handlers and whatever to sort of make it.

You know, breathe life into it, make it real, but you they sort of get that stuff for free right. If all you have is like a sketch library or some like Zeppelin file or some like like little, it’s a style guide thing where it’s like: here’s, our colors and here’s or whatever, like there’s so much that gets lost yeah if all the developers are doing is Like copying and pasting some hex codes in there, you know sort of crappy like development environments, and it’s just you end up with a bunch of spaghetti, even if they’re all using like the same color blue.

It’s not like systematize right. So what you want to get to is, you want to say like if we change our brand color blue – and this actually just happened on a project of ours – got a brand color blue and actually it wasn’t passing the accessibility level that we wanted, and so they Actually had to sort of you know: tweet the the color blue in order to make that sort of pass. You know because sort of cut the accessibility, mustard and with a design system like you literally, have you know a variables file or is these design tokens? You sort of tweak that value there and then that ripples out to the entire sort of design system right and then that gets packaged up in a new release of the design system in code.

And then you know next time the developers pull that down. Those sort of get and see those updates, so so, coming back to it’s like yeah, like the design language part of it, like the look in the feel of it that matters, I’m not going to say it doesn’t matter, but it’s almost just like you’ll, like do Your thing make it look good like I, you know. I trust you be systematic about it right. Thinking about motifs that are going to sort of like you know, translate well the different components, but, like so much hinges on like getting that stuff into a place where it’s consumable by the actual sort of you know, environments that users will be interacting with your products And that’s what we spend, probably the overwhelming majority of our time and effort on is actually like building out those libraries with components right, an HTML, CSS JavaScript.

You know bundling that stuff up and like sort of working with development teams to make sure that they have what they need in order to use the system successfully. So, to what extent should a design system anticipate the chaos of user-generated content like errors and long names? What is the actual like breaking point of a design system yeah? Well, I think that the breaking point of the design system has everything to do with how well you consider all of that stuff right.

So if, if it’s user-generated content that you need to account for and you’re in your UIs, then you have to you know, consider things like character limits and things like that. But you know there’s many other flavors of that as well. You know internationalization right, right-to-left languages or just you know, German will wrap onto multiple lines, and things like that – and this is where I think again – sort of designing and building components in isolation is a bad idea because you could sort of you surf fall into the Trap of saying like well, here’s this like perfect scenario where you know everything’s filled in and the card has this nice sort of you know image I found from unsplash and it’s like really nice.

Looking and you know, as it happens, the users name is Sarah Smith and Sarah doesn’t even have an H on it, so it just fits so nicely onto one line and of course, the reality of of our user interfaces is anything but that, and this sort of Also comes back to like the trap, was sort of relying on these sort of static design tools to sort of handle that they’re up until very very recently, there weren’t even conventions in place to sort of handle like dynamic data, so that’s sort of how we handle That this is where atomic design as a methodology – I think, really shines.

So what atomic design does is basically helps people consider the whole the pages, the actual product screens in various states and configurations, as well as the sort of parts of that hole right. So the underlying components that build up those screens and at the page level of atomic design, what we’re able to do is articulate here’s. What our homepage looks like with this. You know the fall campaign with the leaves – and you know this tagline and this call to action button that takes people to this and and whatever, but then you’re also able to say, okay and then here’s what this that same page looks like in German or here’s.

What that that same page looks like with you know the Christmas campaign and oh that’s, sort of image that we’re using that has a bunch of Christmas ornaments that actually is sort of you know, impacting the the readability of the text. That’s sitting over that image or something like that right, so you could start seeing where the UI starts falling down and then what you’re able to do is is sort of take that and learn from that and sort of go back to that hero component.

At a more atomic level and sort of say, okay, we’re going to maybe add a variation of the hero component that adds like a little gradient overlay so that the the legibility of the text always sort of you know pops over the the image a bit more. So how we sort of do things like in our own workflow, with that we sort of will create sort of you know, try to represent the whole bell curve. So it’s like what does a card? Look like what does sort of like a kitchen sink card? Look like with like the maximum character count that you might be able to sort of upload as a user or something or what happens if the user uploads the profile picture, what if they don’t right, and so all those various states and sort of you know, mutations Of the other component, so to get that sort of commonly used case down.

Of course, as like a starting point but like you really do have to represent like here’s, the extreme and here’s the empty and sort of everything in between as well and the only real way to test. If that actually works is by sort of plugging in real products and Aereo’s into your user interfaces and by sort of having that best sort of atomic design system wired up where, like the pages, informs and influences the underlying components, you’re able to sort of make changes To those components with which, then, you know, inform and influence that the actual page design, so it’s sort of like a virtuous cycle between like the design system and the pages and screens that that system builds.

Finally, what resources would you recommend for people eager to learn? More about design, Cisco there’s a lot I feel like. I have a hard time, keeping up with them anymore. There’s a there’s a number of really great resources, one that I help maintain is a resource called style guides i/o, which is a collection of, I think, we’re over like 200 50 examples of public design systems and style guides that are out there in the wild as Well, as sort of talks and books and resources and tools around a design system, so that’s just like an open source resource repository that people contribute to and sort of, submit poor requests to.

There is design dot systems which is maintained by Gina Ann who’s done so much work for the design systems community. She has a clarity conference, which is a conference dedicated to design systems. We have a podcast, which is a little bit in hiatus, but where we interview people that work at different organizations who have spun up their design systems and what they’ve learned and sort of you know struggled with as they’ve as they’ve done it.

Stu Robson has a really fantastic design systems newsletter. That’s part of the design, dot systems, sort of universe there and then there’s also a slack group all about design systems as well. So I’d save it like that sort of has me covered for sure and again there’s like a lot of activity there and new stuffs happening every day and people are learning from things you know from each other and plugging them in at their organisations and sharing what They’ve learned and like that’s really for me, the most exciting part of all of this is just sort of you know.

Here’s some concepts here are some things that we’ve found useful share. Those people take them, learn from them validate or invalidate them and sort of share. What they’ve learned and then everyone benefits from it, so your book is also available for free to read online right where it is yeah yeah, so you could read it at atomic design. Brad Frost, calm great breath. This has been great. Thank you so much for pyrite.

Thanks so much for having me, you can check out the links to everything we talked about in the description below thanks for reading, we’ll see you next time.


 

Categories
Online Marketing

Service Workers – The State of the Web

My guest is jeff posnick, he’s on Google’s developer relations team and today we’re talking about service workers and how they’re elevating the capabilities of progressive web apps. Let’S get started all right, so Jeff, thanks for being here in the context of web technologies. What does it mean for a worker and what does it actually do so? The whole idea of a worker has been around for a while.

Traditionally there were web workers and it’s basically serves as almost like a background thread for the web, so a worker can execute JavaScript code, that’s kind of independent from the context of your actual web page and it’s a great way to kind of offload processing or I Do tasks that might take a certain amount of time without slowing down the main thread for your web page and yeah, that that’s kind of should been the traditional model for workers on the web.

So now what does it mean for a Service Worker? What does that? Actually do the service workers builds kind of on that concept and adds some superpowers really things that you were not able to do before so a service worker is similar to worker and that it’s, you know, running independent from your actual web page and it doesn’t have Access to things like the Dom you know or the global scope of your web page, but unlike workers, it could respond to specific events and some of those events relate to network traffic.

So one of the really cool things and most common use cases for a Service Worker is to respond to outgoing Network requests that your webpage might be making, and you can kind of sit in between your webpage and the network and almost serve as a proxy that You control and you could write code to take advantage of things like the cache, storage, API and say hey. You know, I know how to respond to this particular request without having to go to the network.

I could just use this cache response and thereby saving you know the uncertainty and unreliability that comes with coming against the network. It also enables capabilities like push notifications, etc. Yeah so there’s a whole bunch of kind of event based listeners that you can set up in the Service Worker, including responding to portion of vacations. That may come from a notification server and you know fetching requests and people other kind of interesting things are kinda slated for the future as well.

So what’s the status of its implementation and support? Yes, the service workers are well supported right now in modern browsers. So pretty much anything Chrome or chromium based, Firefox, Safari and edge at the moment, it’s great. They all have at least a basic level of support for service workers and some of the enabling technologies, like the cache storage API, so they’re they’re ready to use right now.

So web sites may experience Network reliability issues at any. Given time, would you recommend service workers for every website? Should they all be using one? Well, I mean it’s tempting to just throw a service worker up and see what happens. I would suggest to take a little bit more of a considerate approach before adding a Service Worker to your web app. Ideally, a service worker will kind of play the same role that your web server would play and maybe share the same logic for doing routing and templating that your web server would normally respond with.

And if you have a setup where, like your web server, for instance from a lot of single page apps, the web servers just can respond with some static HTML that could be used satisfy any sort of request. That’S pretty easy to map into a Service Worker behavior. We call that the app shell model or a service work role say: hey. You know, you’re navigating to XYZ URL. I could just respond with this HTML and it’ll always work.

So that’s a really good model for using a serviceworker. If you have a single page app we’re also seeing some success with partners or using models where their servers implemented in JavaScript, they have some routing logic and they have some templating logic. That’S on JavaScript, and that translates over really well to the serviceworker as well, where the serviceworker you just basically fill the role that the server would normally play.

I would say if you have a scenario where your back-end web server is doing a whole bunch of complex templating and remote API calls and language that is not JavaScript. It really might be hard to get your serviceworker to behave exactly the same way. So in those scenarios I mean you can add a serviceworker and we have some kind of provisions in place to not pay the price of having that serviceworker, intercepting all requests and then not doing anything and just going on against the network.

There are waves of saying, hey, you know we have a serviceworker, but we’re not going to be able to respond with HTML for navigation requests. In those scenarios it is still possible use the serviceworker for things like ok, show, custom offline page when you detect that a user’s network connection is down or implement a kind of interesting caching strategy, like still while revalidate for certain types of resources.

So it is still possible to add a serviceworker in those cases, but you won’t necessarily get the same performance and reliability benefits that you get when your serviceworker really respond to all navigations with HTML by essentially having a network proxy juggling requests and responses. Is there a latency cost of having a serviceworker yeah, so I mean you’re you’re running JavaScript code, that’s sitting in between your web app and then a work and that’s not for me.

Some of it depends upon whether the serviceworker is already running. One of the kind of neat features about a serviceworker is that just it’s particularly to preserve battery on mobile devices? It’S killed pretty aggressively. It doesn’t just keep running forever in the background. So sometimes you do have to startup the serviceworker again and there is a cost involved in that startup. There’S a really good talk from the chrome dev summit that just happened a couple of months ago that kind of goes into some metrics and real-world performance.

Timings of you know exactly how long it takes to startup a serviceworker, seeing tens to hundreds of milliseconds depending upon the actual device and things like the storage beautiful device. So you are going to be paying that cost. Potentially, when you’re using a serviceworker – and you know again – that’s really why it’s important to make sure that you have a strategy in place for responding to requests, hopefully by avoiding that work and just going against storage API.

Ideally, and if you’re doing that, then you should see the service worker give you an that positive in terms of performance, you know paying tens, maybe even hundreds of milliseconds is nothing compared to the multiple seconds. Simply didn’t see that you might expect from making a network request each time you navigate to a new URL right. What’S the saying the fastest request is the one that you never need to make indeed yeah.

So what are some anti patterns that you’ve seen the way that people have implemented service workers? There’S a lot of power involved in using a Service Worker? It is just JavaScript that you could write that will pretty much do whatever you want, so you can do all sorts of crazy things, some of which are kind of cool as proof of concepts, but not necessarily things you want to deploy to production in terms of The things that we’ve seen kind of as pain, points or things that are pretty easy to, unfortunately get wrong when implementing a Service Worker.

I think one of the things that it’s most common is caching requests and responses, as you go without having any sort of upper limit on the amount of data that you’re storing. So now you can imagine a website that maybe has a bunch of different articles. Each of those articles has images it’s pretty easy to write a serviceworker that just intercepts all those requests and takes the responses, save some in the cache, but those cached responses will never get cleaned up by default.

There’S not really any provision in the cache storage API for saying you know stop when you reach 50 or 100 entries, or something like that, so you could very easily just keep using up space on your users devices and potentially use up space for things that are Never going to be used again, you know if you have an article from a week ago and you’re caching, all the images and that article that’s kind of cool.

I guess if you’re going to be visit article immediately, but if it’s a page that users never going to go to again, then you’re, really just caching things for no reason. I would say that really one of the important things before you implement your serviceworker kind of have a strategy for each type of request and say: here’s my navigation requests that are being made for HTML; here’s how I’m going to respond to them here.

The image requests. I’M making you know, maybe it doesn’t make sense to cash them at all, or maybe certain it only cache certain images and not others. So thinking about that – and that really just means getting really comfortable with the kind of network info panel in the browser’s dev tools and just seeing the full list of requests are being made. You know. Sometimes your web app is making requests.

If you don’t even realize it’s happening and it’s coming from the third-party code and your service worker ends up seeing that too, so you want to make sure that you know what your service work is doing. You know what your web app is doing and just one other. I would know that a lot of times and kind of pain, point and things that could go wrong with me using a service work, but just has to do with controlling updates to resources.

So you know you are stepping in between. You know your web app and a web server you’re responding, potentially the cached resources, if you’re not sure that those cached resources are being updated. Every time you make changes to your actual website and you read – apply to your web server, it’s possible that your users will end up seeing stale content kind of indefinitely, and this is a trade-off like seeing stale content, but avoiding the network gives you performance benefits.

So that’s that’s good for a lot of scenarios, but you do need to have a provision in place for updating and making sure that you know. Maybe the user sees still content then the next time they visit the site. They get fresh content. So you know you could do that right. Unfortunately, you could get that part wrong and the users can end up the frustrating experience. So you maintain a tool called work box j/s.

What is that? What does it do sure so? Work box is open source and a set of libraries for dealing with service workers and kind of all aspects of building service workers. So we have some tools that integrated with build processes, including you know we have web pack plugin. We have a command line tool. We have a node module and that aspect of the tools, basically, is something you can drop in your current build process and kind of get a list of all of the assets that are being produced.

Every time you rebuild your site along with kind of some fingerprinting information like say you know, this is a particular version of your index. Dot HTML work backs will keep track of that for you and then it will efficiently cache all of those files that are being created by your build process for you and that just helps ensure that you don’t run into scenarios like I just described where you’ve rebuilt.

Your site – and you know you never get updates to your previously cached resources and we also have some tools as part of work box, that kind of harm or execute at runtime. That’S part of the serviceworker, so some libraries for doing common things like routing requests. We have there’s just kind of some canonical response strategies for dealing with caching, so things like still while we validate or going cache.

First, we have implementations of those strategies inside of work box, and then we have some kind of like value adds on top of what you get with the basic serviceworker spec in the cache stored specs. So we actually have an implementation of a cache expiration policy that you could apply to the caches that would otherwise just grow indefinitely, but using work box you could say, hey. You know it actually like to stop.

When I reach ten items and purge the least recently used items and just cache when that happens, and a few other kind of ran two modules, we see it as a bit of a kind of grab bag for all the things that somebody might want to do With a serviceworker and we kind of ship them as individual modules, you can choose the ones that you think would be useful for your particular use case. I don’t want to use something, that’s fine, you don’t have to incur the cost of you know downloading it or anything like that.

Do you foresee some of those caching and expiration policies making their way back into the cache storage API yeah. I mean it’s kind of interesting whenever you have something: that’s almost like a polyfill for some behavior on the web. You know whether that ends up being implemented back into the standards, and you know the the actual runtime could just fade away and just use the underlying standards.

And you know I’d like to see that. I think that where cost has been really great for kind of enabling folks to ship service workers in production and seeing the types of things that they actually need, when you’re shipping somebody in production and a lot of times when you could do that and get points. As a vision thing like yeah, you know it is actually important to have run time, cache expiration.

That could then be used. You know when going to different standards, groups and saying hey, we really do need to extend. You know, what’s supported natively in the platform, to take care of this really common use case. You know what that actually happens or not remains to be seen, but you know I think work box is positioned to help folks with kind of that initial, proving that these things are necessary stage kind of take it from there.

So, in terms of adoption, according to the HTTP archive, less than 1 % of websites tested actually include a serviceworker which is kind of a misleading number. For two reasons. The first is that it’s actually growing at a very fast rate and the websites that do include it are actually pretty popular websites. Can you give us some examples of those yeah? So I think you know the raw number of URLs unique URLs might be on the lower side, but I think in terms of traffic, you know sites as big as Google search have deployed a serviceworker for some types of clients.

You know partners that we’ve talked about using work box, in particular in the past and Gleevec Starbucks has a nice progressive web app, that’s implemented Pinterest as well, and there’s also some sites that you might have heard of like Facebook and Twitter that are using service workers. Not using work box but using them to kind of unlock things like you know, they’re progressive web app experience – or you know in some cases just showing notifications, which is important part of you know being on the web and having parity with native apps.

So you know, I think that the actual number of you know you visits to web pages is probably much higher than the 1 % number would indicate, and you know I mean there are challenges with adding a service worker into especially legacy sites. You know it does. Take that coordination that we talked about before tree, making sure that your service worker actually is behaving in a similar way that your web server would behave and yeah that doesn’t always fit into existing sites.

So a lot of times we’ve seen when working with partners in particular, is like you know: you’re planning a rewrite, re architecture of your site anyway, that’s a great time to add a service worker in and just kind of take care of that story as well. Are there any options for CMS users who may be using things like WordPress or Drupal? So there definitely are, and I think that you know first of all, I’d work for everybody back to another talk from the most recent chrome dev summit.

That really goes into some detail about the WordPress ecosystem in general, so they have a really cool solution, some folks from the dev rel team that Google have been working on it and I think it kind of works around one that that problem. I was saying where the architecture for your kind of back-end web server needs to match up with the serviceworker implementation I kind of just sending a baseline.

So it’s not an attempt to take any arbitrary, WordPress site that might be out there, which might be executing random PHP code depending upon you know what kind of themes and extensions and all the other stuff is going on. You really are not going to be able to successfully translate that into just a general-purpose serviceworker, but the approach that was subscribed and this talk. It seems to be building on top of a kind of a common baseline of using the amp plugin as a starting point.

So any site that has gone through the effort of kind of meeting all the requirements for using the amp plugin. So it means I don’t know the full set, but I think, like not running external scripts, not doing anything too crazy with other plugins. That’S inserting random HTML on the page building. On top of that, you can then have a serviceworker. That’S like okay. I actually do know how to handle this subset of activities that you know WordPress is doing when it’s using the unplug in and it can automatically generate that serviceworker for you.

So again, it’s part of a migration story. I think it’s not going to just drop into any existing legacy WordPress site, but it does give a nice path forward for folks who are planning on rewriting anyway are planning on making some changes anyway, and plugging into the CMS ecosystem is great way to increase adoption By tens of percents on what yeah absolutely so, what kinds of resources would you recommend for someone who’s just getting started with service workers? We have a lot of material available, some of which is more recent than others.

I would say that the things that I worked on most recently are the resiliency section of web dev. So if you were to go there kind of have something I would walk you through the various steps of thinking about adding a service worker to your website or just really thinking about making your website more resilient in general. So it’ll talk about you know identifying your network traffic it’ll talk about using the browser’s HTTP cache effectively, which is kind of your first line of defense, and then it all kind of go into how you could add work box to an existing site and the various Steps involved there, so if you want kind of a guided path, I would say that’s one option we’ll biased.

For that. I would say that if you want to just learn more about service workers in general and material written by my colleague, Jake Archibald, it’s probably the best that for folks who really want to deep dive on things, he was somebody who worked on the actual serviceworker specification And you know he knows more than anybody else about these things, so he was a really great article talking about the serviceworker lifecycle, just all the different events we have fired, and you know how you have to handle those events differently and implications that they have for You know the state of your caches and updates, and things like that so diving into that would be kind of my recommended starting point, and he has another article that talks about kind of a cookbook almost for recipes for caching, so implementations of the stove are valid.

A pattern cache first pattern: if you wanted to implement it yourself, instead of using work box, he kind of walks through the process. There is that the offline cookbook, yes, the offline cookbook, and if you want something, that’s really offline, there’s some actual physical books that that are pretty cool, related to service workers and progressive web apps in general. There’S a new book written by Jason, Grigsby, eight in particular, that I would recommend and just kind of talks a little bit about, I’m necessarily some of the technical aspects of service workers, but more about why you should think about adding a service worker to your site And why you might want to build progressive web app in general and that’s a really cool book, that kind of takes it from a slightly different angle, but gives some good perspective great Jeff.

Thank you again for being here. Absolutely you can find links to everything we talked about in the description below thanks a lot and we’ll see you next time.


Website management packages are important for any business these days. Check out the video from Allshouse Designs to see what can be done for your company and yes, for how much.