as they are essentially non-semantic containers, able to be used for any purpose. The attributes are used to define functionality, for example the camera attribute sets the entity to function as a camera and kinematic-body makes it collide instead of go through objects. Attributes are also used to set position and sizes, often using JavaScript to dynamically define them.
Styling
Now we’ve got the HTML written, we need to style it. To do this we add A-Frame compatible attributes such as color and material. I recommend playing around, you can get some quite impressive effects fairly easily. Originally I wanted a light snowy maze but it ended up being dark and foggy, as I really liked the feeling it gave.
Note, you will probably need a server running for images to work. You can do this by running python -m ""SimpleHTTPServer"" in the folder where the code is, then go to localhost:8000 in browser.
Textures
Unless you are going for a cartoony style, you probably want to find some textures. I found some on textures.com, one image worked well for the walls and the other for the floor.
![]()
![]()
The
is used to define (as well as preload and cache) all assets, including images, audio and video. As you can see, images in the Asset Management System just use normal img tags. The ids are important here as we can use them later for using the textures.
To apply a texture to an object, you create a material. For a simple material where it just shows the image, you set the src to the id selector of the image.
Replace:
With:
This will automatically make the image repeat over the entire floor, in my case filling it with bricks. The walls are pretty much identical, with the slight exception that it is set in JavaScript as they are dynamically defined.
box.setAttribute('material', 'src: #texture-wall');
That’s it for the textures, for now at least. These will not look completely realistic, as the light will bump off the rectangular wall rather than texture itself. This can be improved by using maps, textures that are used to modify the shape and physical properties of the object.
Lighting
The next part of styling is lighting. By using fog and different types of lighting, we are able to add atmospheric details to the game to make it feel that bit more realistic and polished.
There are lots of types of light in A-Frame (most coming from Three.js). You can add a light either by using the entity or by attaching a light attribute to any other entity. If there are no lights defined then A-Frame adds some by default so that the scene is always lit.
To start with I wanted to light up the scene with a general light, type=""ambient"", so that the whole game felt slightly dark. I chose to set the light to a reddish colour #92455E. After playing around with intensity I chose 0.4, it added enough light to get the feeling I wanted without it being overly red. I also added a blue skybox (), as it looked a bit odd with a white sky.
I felt that the maze looked good with a red tinge but it was a bit flat, everything was the same colour and it was a bit dark. So I added a light within the #player entity, this could have been as an attribute but I set it as a child a-light instead. By using type=""point"" with a high intensity and low distance, it showed close walls as being lighter. It also added a sort-of object to the player, it isn’t a walking human or anything but by moving light where the player is it feels a bit more physical.
By this point it was starting to look decent, so I wanted to add the fog to really give some personality and depth to the maze. To do this I added the fog attribute to the with type=exponential so it looks thicker the further away it is and a mid intensity, so you feel a bit lost but can still see.
I was very happy with this result. It took a lot of playing around with colours and values, which is fun in itself. I highly recommend you take the code (or write your own) and play around with the numbers.
Movement
One of the reasons I decided to use aframe-extras is that it has a few different camera controls built in. As you saw earlier, I am using the universal-controls which gives WASD (keyboard) controls by default. I wanted to make it automatically move in the direction that you’re looking, but I wasn’t quite sure how without rewriting the controls. So I asked Don McCurdy for advice and he very nicely gave me a small snippet of code to get it working.
AFRAME.registerComponent('automove-controls', {
init: function () {
this.speed = 0.1;
this.isMoving = true;
this.velocityDelta = new THREE.Vector3();
},
isVelocityActive: function () {
return this.isMoving;
},
getVelocityDelta: function () {
this.velocityDelta.z = this.isMoving ? -speed : 0;
return this.velocityDelta.clone();
}
});
Replace:
universal-controls
With:
universal-controls=""movementControls: automove, gamepad, keyboard""
This works by creating a component automove-controls that adds auto-move to the player without overriding movement completely. It doesn’t even touch direction, it just checks if isMoving is true then moves the player by the set speed. Components can be creating for adding all kinds of functionality with relative ease. It makes it very powerful for people of all difficulty levels.
Building a map
Currently the maze is created randomly, which is great but means there will often be walls that overlap or the player gets trapped with nowhere to go. So to solve this, I decided to use a map editor (Tiled) so that we can create the mazes ourselves. This is a great start towards one of the stretch goals, levels.
I made the maze in Tiled by finding a random tileset online (we don’t need to actually show the images), I used one tile for the wall and another for the player. Then I exported as a JavaScript file and modified it in my text editor to get rid of everything I didn’t need. I made it so 0 is the path, 1 is the wall and 2 is the player. I then added the script to the HTML, as a separate file so it’s easy to update in the future.
var map =
{
""data"":[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 0, 1, 1, 0, 1, 1, 1, 0, 1, 1, 0, 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 1, 0, 1, 0, 0, 1, 1, 0, 0, 0, 1, 0, 1, 1, 0, 1, 1, 0, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 1, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 0, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],
""height"":10,
""width"":10
}
As you can see, this gives a simple 10x10 maze with some dead ends. The player starts in the bottom right corner (my choice, could be anywhere). I rewrote the random platforms code (from Don’s example) to instead loop over the map data and place walls where it is 1 and position the player where data is 2. I set the position so that the origin of the map would be 0,1.5,0. The y axis is in this case the height (ground being 0), but if a wall is positioned at 0 by its centre then some of it is underground. So the y needed to be the height divided by 2.
document.querySelector('a-scene').addEventListener('render-target-loaded', function () {
var WALL_SIZE = 5,
WALL_HEIGHT = 3;
var el = document.querySelector('#walls');
var wall;
for (var x = 0; x < map.height; x++) {
for (var y = 0; y < map.width; y++) {
var i = y*map.width + x;
var position = (x-map.width/2)*WALL_SIZE + ' ' + 1.5 + ' ' + (y-map.height/2)*WALL_SIZE;
if (map.data[i] === 1) {
// Create wall
wall = document.createElement('a-box');
el.appendChild(wall);
wall.setAttribute('color', '#fff');
wall.setAttribute('material', 'src: #texture-wall;');
wall.setAttribute('width', WALL_SIZE);
wall.setAttribute('height', WALL_HEIGHT);
wall.setAttribute('depth', WALL_SIZE);
wall.setAttribute('position', position);
wall.setAttribute('static-body', ');
}
if (map.data[i] === 2) {
// Set player position
document.querySelector('#player').setAttribute('position', position);
}
}
}
console.info('Walls added.');
});
With this added, it makes it nice and easy to change around the map as well as to add new features. Perhaps you want monsters or objects. Just set the number in the map data and add an if statement to the loop. In the future you could add layers, so multiple things can be in the same position. Or perhaps even make the maze go up the y axis too, with ramps or staircases. There’s a lot you can do with relative ease. As you can see, A-Frame really does reduce the learning curve of 3D and VR on the web.
It’s Not All Fun And Games
A lot of examples of virtual reality are games, including this one. So it is understandable to think that VR is for gaming, but actually that’s just a tiny subset. There are all sorts of applications for VR, including story telling, data visualisation and even meditation.
There have been a number of cases where it has been shown virtual reality can help as a tool for therapies:
Oxford study finds virtual reality can help treat severe paranoia
Virtual Reality Therapy for Phobias at the Duke Faculty Practice
Bravemind: Virtual Reality Exposure Therapy at the University of Southern California
These are just a few examples of where virtual reality is being used around the world to help people feel better and get through some very tough times. There have also been examples of it being used for simulating war zones or medical situations, both as a teaching and journalism tool.
Wrapping Up
Ten years ago, on this very site, Cameron Moll wrote an article explaining the mobile web. He explained how mobile phones with data plans were becoming increasingly common, that WAP 2.0 included the XHTML Mobile Profile meaning it would be familiar with web folk. “The mobile web is rapidly becoming an XHTML environment, and thus you and I can apply our existing “desktop web” skills to understand how to develop content for it.”
We can look at that and laugh a little, we have come a very long way in the last decade. Even people in developing countries with very little money have mobile phones with access to a web that is far more capable than the “desktop web” Cameron was referring to.
So while I am not saying virtual reality is going to change the world or replace our phones, who knows! We can use our skills as web folk to dabble, we don’t need to learn any new languages. If on the 2026 edition of 24 ways, somebody references this article and looks at how far we have come… well, let’s hope we have used our skills well and made the world just that little bit better. And if VR is a fad? Well it’s fun… have a go anyway.",2016,Shane Hudson,shanehudson,2016-12-11T00:00:00+00:00,https://24ways.org/2016/first-steps-in-vr/,code
294,New Tricks for an Old Dog,"Much of my year has been spent helping new team members find their way around the expansive and complex codebase that is the TweetDeck front-end, trying to build a happy and productive group of people around a substantial codebase with many layers of legacy.
I’ve loved doing this. Everything from writing new documentation, drawing diagrams, and holding technical architecture sessions teaches you something you didn’t know or exposes an area of uncertainty that you can go work on.
In this article, I hope to share some experiences and techniques that will prove useful in your own situation and that you can impress your friends in some new and exciting ways!
How do you do, fellow kids?
To start with I’d like to introduce you to our JavaScript framework, Flight. Right now it’s used by twitter.com and TweetDeck although, as a company, Twitter is largely moving to React.
Over time, as we used Flight for more complex interfaces, we found it wasn’t scaling with us.
Composing components into trees was fiddly and often only applied for a specific parent-child pairing. It seems like an obvious feature with hindsight, but it didn’t come built-in to Flight, and it made reusing components a real challenge.
There was no standard way to manage the state of a component; they all did it slightly differently, and the technique often varied by who was writing the code. This cost us in maintainability as you just couldn’t predict how a component would be built until you opened it.
Making matters worse, Flight relied on events to move data around the application. Unfortunately, events aren’t good for giving structure to complex logic. They jump around in a way that’s hard to understand and debug, and force you to search your code for a specific string — the event name‚ to figure out what’s going on.
To find fixes for these problems, we looked around at other frameworks. We like React for it’s simple, predictable state management and reactive re-render flow, and Elm for bringing strict functional programming to everyone.
But when you have lots of existing code, rewriting or switching framework is a painful and expensive option. You have to understand how it will interact with your existing code, how you’ll test it alongside existing code, and how it will affect the size and performance of the application. This all takes time and effort!
Instead of planning a rewrite, we looked for the ideas hidden within other frameworks that we could reapply in our own situation or bring to the tools we already were using.
Boiled down, what we liked seemed quite simple:
Component nesting & composition
Easy, predictable state management
Normal functions for data manipulation
Making these ideas applicable to Flight took some time, but we’re in a much better place now. Through persistent trial-and-error, we have well documented, testable and standard techniques for creating complex component hierarchies, updating and reacting to state changes, and passing data around the app.
While the specifics of our situation and Flight aren’t really important, this experience taught me something:
Distill good tech into great ideas. You can apply great ideas anywhere.
You don’t have to use cool kids’ latest framework, hottest build tool or fashionable language to benefit from them. If you can identify a nugget of gold at the heart of it all, why not use it to improve what you have already?
Times, they are a changin’
Apart from stealing ideas from the new and shiny, how can we keep make the most of improved tooling and techniques? Times change and so should the way we write code.
Going back in time a bit, TweetDeck used some slightly outmoded tools for building and bundling. Without a transpiler like Babel we were missing out new language features, and without a more advanced build tools like Webpack, every module’s source was encased in AMD boilerplate.
In fact, we found ourselves with a mix of both AMD syntaxes:
define([""lodash""], function (_) {
// . . .
});
define(function (require) {
var _ = require(""lodash"");
// . . .
});
This just wouldn’t do. And besides, what we really wanted was CommonJS, or even ES2015 module syntax:
import _ from ""lodash"";
These days we’re using Babel, Webpack, ES2015 modules and many new language features that make development just… better. But how did we get there?
To explain, I want to introduce you to codemods and jscodeshift.
A codemod is a large-scale refactor of a whole codebase, often mechanical or repetitive. Think of renaming a module or changing an API like URL(""..."") to new URL(""..."").
jscodeshift is a toolkit for running automated codemods, where you express a code transformation using code. The automated codemod operates on each file’s syntax tree – a data-structure representation of the code — finding and modifying in place as it goes.
Here’s an example that renames all instances of the variable foo to bar:
module.exports = function (fileInfo, api) {
return api
.jscodeshift(fileInfo.source)
.findVariableDeclarators('foo')
.renameTo('bar')
.toSource();
};
It’s a seriously powerful tool, and we’ve used it to write a series of codemods that:
rename modules,
unify our use of AMD to a single syntax,
transition from one testing framework to another, and
switch from AMD to CommonJS.
These changes can be pretty huge and far-reaching. Here’s an example commit from when we switched to CommonJS:
commit 8f75de8fd4c702115c7bf58febba1afa96ae52fc
Date: Tue Jul 12 2016
Run AMD -> CommonJS codemod
418 files changed, 47550 insertions(+), 48468 deletions(-)
Yep, that’s just under 50k lines changed, tested, merged and deployed without any trouble. AMD be gone!
From this step-by-step approach, using codemods to incrementally tweak and improve, we extracted a little codemod recipe for making significant, multi-stage changes:
Find all the existing patterns
Choose the two most similar
Unify with a codemod
Repeat.
For example:
For module loading, we had 2 competing AMD patterns plus some use of CommonJS
The two AMD syntaxes were the most similar
We used a codemod to move to unify the AMD patterns
Later we returned to AMD to convert it to CommonJS
It’s worked for us, and if you’d like to know more about codemods then check out Evolving Complex Systems Incrementally by Facebook engineer, Christoph Pojer.
Welcome aboard!
As TweetDeck has gotten older and larger, the amount of things a new engineer has to learn about has exploded. The myriad of microservices that manage our data and their layers of authentication, security and business logic around them make for an overwhelming amount of information to hand to a newbie.
Inspired by Amy’s amazing Guide to the Care and Feeding of Junior Devs, we realised it was important to take time to design our onboarding that each of our new hires go through to make the most of their first few weeks.
Joining a new company, team, or both, is stressful and uncomfortable. Everything you can do to help a new hire will be valuable to them. So please, take time to design your onboarding!
And as you build up an onboarding process, you’ll create things that are useful for more than just new hires; it’ll force you to write documentation, for example, in a way that’s understandable for people who are unfamiliar with your team, product and codebase. This can lead to more outside contributions: potential contributors feel more comfortable getting set up on your product without asking for help.
This is something that’s taken for granted in open source, but somehow I think we forget about it in big companies.
After all, better documentation is just a good thing. You will forget things from time to time, and you’d be surprised how often the “beginner” docs help!
For TweetDeck, we put together system and architecture diagrams, and one-pager explanations of important concepts:
What are our dependencies?
Where are the potential points of failure?
Where does authentication live? Storage? Caching?
Who owns “X”?
Of course, learning continues long after onboarding. The landscape is constantly shifting; old services are deprecated, new APIs appear and what once true can suddenly be very wrong. Keeping up with this is a serious challenge, and more than any one person can track.
To address this, we’ve thought hard about our knowledge sharing practices across the whole team. For example, we completely changed the way we do code review.
In my opinion, code review is the single most effective practice you can introduce to share knowledge around, and build the quality and consistency of your team’s work. But, if you’re not doing it, here’s my suggestion for getting started:
Every pull request gets a +1 from someone else.
That’s all — it’s very light-weight and easy. Just ask someone to have a quick look over your code before it goes into master.
At Twitter, every commit gets a code review. We do a lot of reviewing, so small efficiency and effectiveness improvements make a big difference. Over time we learned some things:
Don’t review for more than hour 1
Keep reviews smaller than ~400 lines 2
Code review your own code first 2
After an hour, and above roughly 400 lines, your ability to detect issues in a code review starts to decrease. So review little and often. The gaps around lunch, standup and before you head home are ideal. And remember, if someone’s put code up for a review, that review is blocking them doing other work. It’s your job to unblock them.
On TweetDeck, we actually try to keep reviews under 250 lines. It doesn’t sound like much, but this constraint applies pressure to make smaller, incremental changes. This makes breakages easier to detect and roll back, and leads to a very natural feature development process that encourages learning and iteration.
But the most important thing I’ve learned personally is that reviewing my own code is the best way to spot issues. I try to approach my own reviews the way I approach my team’s: with fresh, critical eyes, after a break, using a dedicated code review tool.
It’s amazing what you can spot when you put a new in a new interface around code you’ve been staring at for hours!
And yes, this list features science. The data backs up these conclusions, and if you’d like to learn more about scientific approaches to software engineering then I recommend you buy Making Software: What Really Works, and Why We Believe It. It’s ace.
For more dedicated information sharing, we’ve introduced regular seminars for everyone who works on a specific area or technology. It works like this: a team-member shares or teaches something to everyone else, and next time it’s someone else’s turn. Giving everyone a chance to speak, and encouraging a wide range of topics, is starting to produce great results.
If you’d like to run a seminar, one thing you could try to get started: run a point at the thing you least understand in our architecture session — thanks to James for this idea. And guess what… your onboarding architecture diagrams will help (and benefit from) this!
More, please!
There’s a few ideas here to get you started, but there are even more in a talk I gave this year called Frontend Archaeology, including a look at optimising for confidence with front-end operations.
And finally, thanks to Amy for proof reading this and to Passy for feedback on the original talk.
Dunsmore et al. 2000. Object-Oriented Inspection in the Face of Delocalisation. Beverly, MA: SmartBear Software. ↩
Cohen, Jason. 2006. Best Kept Secrets of Peer Code Review. Proceedings of the 22nd ICSE 2000: 467-476. ↩ ↩",2016,Tom Ashworth,tomashworth,2016-12-18T00:00:00+00:00,https://24ways.org/2016/new-tricks-for-an-old-dog/,code
290,Creating a Weekly Research Cadence,"Working on a product team, it’s easy to get hyper-focused on building features and lose sight of your users and their daily challenges. User research can be time-consuming to set up, so it often becomes ad-hoc and irregular, only performed in response to a particular question or concern. But without frequent touch points and opportunities for discovery, your product will stagnate and become less and less relevant. Setting up an efficient cadence of weekly research conversations will re-focus your team on user problems and provide a steady stream of insights for product development.
As my team transitioned into a Lean process earlier this year, we needed a way to get more feedback from users in a short amount of time. Our users are internet marketers—always busy and often difficult to reach. Scheduling research took days of emailing back and forth to find mutually agreeable times, and juggling one-off conversations made it difficult to connect with more than one or two people per week. The slow pace of research was allowing additional risk to creep into our product development.
I wanted to find a way for our team to test ideas and validate assumptions sooner and more often—but without increasing the administrative burden of scheduling. The solution: creating a regular cadence of research and testing that required a minimum of effort to coordinate.
Setting up a weekly user research cadence accelerated our learning and built momentum behind strategic experiments. By dedicating time every week to talk to a few users, we made ongoing research a painless part of every weekly sprint. But increasing the frequency of our research had other benefits as well. With only five working days between sessions, a weekly cadence forced us to keep our work small and iterative. Committing to testing something every week meant showing work earlier and more often than we might have preferred—pushing us out of your comfort zone into a process of more rapid experimentation.
Best of all, frequent conversations with users helped us become more customer-focused. After just a few weeks in a consistent research cadence, I noticed user feedback weaving itself through our planning and strategy sessions. Comments like “Remember what Jenna said last week, about not being able to customize her lists?” would pop up as frequent reference points to guide our decisions. As discussions become less about subjective opinions and more about responding to user needs, we saw immediate improvement in the quality of our solutions.
Establishing an efficient recruitment process
The key to creating a regular cadence of ongoing user research is an efficient recruitment and scheduling process—along with a commitment to prioritize the time needed for research conversations. This is an invaluable tool for product teams (whether or not they follow a Lean process), but could easily be adapted for content strategy teams, agency teams, a UX team of one, or any other project that would benefit from short, frequent conversations with users.
The process I use requires a few hours of setup time at the beginning, but pays off in better learning and better releases over the long run. Almost any team could use this as a starting point and adapt it to their own needs.
Pick a dedicated time each week for research
In order to make research a priority, we started by choosing a time each week when everyone on the product team was available. Between stand-ups, grooming sessions, and roadmap reviews, it wasn’t easy to do! Nevertheless, it’s important to include as many people as possible in conversations with your users. Getting a second-hand summary of research results doesn’t have the same impact as hearing someone describe their frustrations and concerns first-hand. The more people in the room to hear those concerns, the more likely they are to become priorities for your team.
I blocked off 2 hours for research conversations every Thursday afternoon. We make this time sacred, and never schedule other meetings or work across those hours.
Divide your time into several research slots
After my weekly cadence was set, I divided the time into four 20-minute time slots. Twenty minutes is long enough for us to ask several open-ended questions or get feedback on a prototype, without being a burden on our users’ busy schedules. Depending on your work, you may need schedule longer sessions—but beware the urge to create blocks that last an hour or more. A weekly research cadence is designed to facilitate rapid, ongoing feedback and testing; it should force you to talk to users often and to keep your work small and iterative. Projects that require longer, more in-depth testing will probably need a dedicated research project of their own.
I used the scheduling software Calendly to create interview appointments on a calendar that I can share with users, and customized the confirmation and reminder emails with information about how to access our video conferencing software. (Most of our research is done remotely, but this could be set up with details for in-person meetings as well.) Automating these emails and reminders took a little bit of time to set up, but was worth it for how much faster it made the process overall.
Invite users to sign up for a time that’s convenient for them
With a calendar set up and follow-up emails automated, it becomes incredibly easy to schedule research conversations. Each week, I send a short email out to a small group of users inviting them to participate, explaining that this is a chance to provide feedback that will improve our product or occasionally promoting the opportunity to get a sneak peek at new features we’re working on. The email includes a link to the Calendly appointments, allowing users who are interested to opt in to a time that fits their schedule.
Setting up appointments the first go around involved a bit of educated guessing. How many invitations would it take to fill all four of my weekly slots? How far in advance did I need to recruit users? But after a few weeks of trial and error, I found that sending 12-16 invitations usually allows me to fill all four interview slots. Our users often have meetings pop up at short notice, so we get the best results when I send the recruiting email on Tuesday, two days before my research block.
It may take a bit of experimentation to fine tune your process, but it’s worth the effort to get it right. (The worst thing that’s happened since I began recruiting this way was receiving emails from users complaining that there were no open slots available!) I can now fill most of an afternoon with back-to-back user research sessions just by sending just one or two emails each week, increasing our research pace while leaving plenty time to focus on discovery and design.
Getting the most out of your research sessions
As you get comfortable with the rhythm of talking to users each week, you’ll find more and more ways to get value out of your conversations. At first, you may prefer to just show work in progress—such as mockups or a simple prototype—and ask open-ended questions to measure user reaction. When you begin new projects, you may want to use this time to research behavior on existing features—either watching participants as they use part of your product or asking them to give an account of a recent experience in your app. You may even want to run more abstracted Lean experiments, if that’s the best way to validate the assumptions your team is working from.
Whatever you do, plan some time a day or two later to come back together and review what you’ve learned each week. Synthesizing research outcomes as a group will help keep your team in alignment and allow each person to highlight what they took away from each conversation.
Over time, you may find that the pace of weekly user research becomes more exhausting than energizing, especially if the responsibility for scheduling and planning falls on just one person. Don’t allow yourself to get burned out; a healthy research cadence should also include time to rest and reflect if the pace becomes too rapid to sustain. Take breaks as needed, then pick up the pace again as soon as you’re ready.",2016,Wren Lanier,wrenlanier,2016-12-02T00:00:00+00:00,https://24ways.org/2016/creating-a-weekly-research-cadence/,ux