Illyriad: The Journey from Concept to HTML5

Andrew J Baker talks to Illyriad CEO James Niesewand about the HTML5 MMORTS Illyriad, from initial concept to release to adoption of HTML5 and continued development. As well as advice for new and old developers of persistent HTML5 games.

1. How close to Illyriad's initial concept is the current version of the game?

I'd say we're probably just under half way to the original vision!

Illyriad was born out of a late-night and well-oiled conversation between gamer friends -  none of whom had any experience whatsoever in the games industry - where we discussed the current state of gaming and bemoaned the lack of anything truly new in the strategic empire-building genre. 

We wrote down everything any of us liked most about classic games in our favourite genres, from Civilisation and Dune to MOO; from HoMM, to Ultima, to later games such as Asheron's Call and Eve Online.

Around about dawn, we came to the conclusion that there wasn't a single game out there that encapsulated everything we wanted from a game - and especially nothing in the free-to-play persistent browser-based game ( pbbg) world, and the only real option was to build our own.

A few weeks later we'd fleshed out the skeleton of the "ultimate" game we wanted to attempt to build.  After working on some proof of concepts in our spare time, much hand-wringing later, a couple of us gave up our RL jobs and started work on Illyriad full time.  We knew, however, that if we set out to build this "ultimate" game and then release it we'd actually never get it finished, so we set ourselves a goal of producing a "minimum viable product" that we could put live.

About 6 months after we had really started on Illyriad, we were ready to launch.

Since launch we've regularly deployed new content - whether that's implementing fractally generated world maps complete with biomes, or introducing NPCs and Factions, or tournaments and siege warfare - and all these pieces of new content are on the original sheets of paper, which even to this day informs the master development plan.   Items further down the list include everything from unit pathfinding to item crafting, religion and guilds, airborne and naval warfare etc. 

2. Did you have any prior experience of developing, playing, or administrating a browser-based game?

No, no experience on developing or administrating a browser-based game (bbg) whatsoever, neither Ben nor I! 

Sure, we'd played a lot of of the strategy bbgs but none had really satisfied our cravings for a deep and meaningful gaming experience; of being in a place in a living, breathing world where your actions really do matter, where danger and risk always lurks, but where communities and friends can flourish, where lone wolves as well as massive alliances can thrive, where trade, magic, combat, diplomacy and questing can be found in PvP as well as PvE - basically, that whole sense of "epic" was missing from most every game we had tried playing.

Having no experience in writing a game made Illyriad a big gamble for both of us.  However, both Ben and I are avid gamers as well having between us a ton of experience in the commercial world.  

We work together very well, as we both have a passion for game design and thinking laterally to solve a problem and have skillsets that complement each other.  I tend to focus on the business-logic of the gameworld, and Ben looks after the programmatic side, especially at the front-end interface.  Ben once described the way we work together as "I look after the horizontal and vertical and James looks after the depth", which encapsulates it quite neatly.  We weren't really concerned about whether Illyriad would succeed or not, I think we always felt it would simply because there wasn't (and still isn't) anything like it out there.

4. What did you use in the early days of Illyriad's inception regarding acquiring artwork, etc.?

Neither Ben nor I are graphic designers in any way, shape, sense or form, so the original graphics for the game were produced by my brother, who I badgered and cajoled into producing everything from the logo to the original site design, world map tiles and building tiles.

However, I couldn't continue to sponge off his goodwill indefinitely, and as soon as Illyriad began getting known (and some money was coming in) I had the good fortune to meet a freelance game industry producer, Kevin Hassall, who came on board to run our external collaboration and production.  Kevin introduced us to a bevy of amazing content writers and graphic designers who have produced the majority of the faction content, lore and map artwork you see today, and who are building the 3D models we're going to use in the WebGL implementation.  Since then, Nathan Bebb has joined us as Art Director to provide an inhouse overview of the creative artistic process, as well as producing much of his own work on the research technology tree graphics in his own indomitable "propaganda art" style.

There are further details on our slightly anarchic development process in a develop-online magazine interview.

3. What originally influenced you to adopt some of the new features that HTML5 has to offer?

It was a no-brainer for us.  Things like producing a usable, multi-layered interactive world map that works cross-browser with drag-and-drop, mousewheel zooming etc… in HTML4….?!  You were much better off using a google map mashup for that because every time a new browser version came out that broke it, it was someone else's problem.  The ambitious nature of what we really wanted to create in Illyriad meant that we couldn't escape HTML4 fast enough!

We were also hitting some limitations on the client-side as well.  You can squeeze every drop of performance out of an HTML4 browser and still want more.  Some of the Network I/O features, as well as having proper local storage in HTML5 make life very much faster for all our players.

5. What have been the biggest challenges so far that you've had to overcome and how have you overcome them?

HTML5 has been like a breath of fresh air to us.  Things that we wanted to do (but were shelved because they were practically impossible to do properly in HTML4) are now a reality, and are going live every week.

The biggest challenge with HTML5 has been the general lack of libraries, forcing us to re-learn and dust-off many "old school" techniques.  However, we love 'coding-in-the-raw' and so this has been fun more than anything else. 

6. What advice would you give to other budding HTML5 developers that are considering or beginning to embark on a game of their own?

Apart from "don't"? ;)

Well, firstly I'd strongly recommend spending as much time on the planning phase as possible.  Especially in an MMO envrionment, it's hard to "nerf" something you've already released without upsetting the playerbase and in the early days that can be difficult to recover from.  You have to balance that, however, with the reality that you need to get something out that people are going to enjoy playing.  It *will* take longer to get off the ground than you thought.

I'd recommend taking a modular approach to development, if your project is ambitious, and especially if it's an MMO or a sandbox game.  Decide on the minimum sets of functionality and playability and implement them, and only them - everyone loves an expansion that comes once you've had time to assess what's actually going on in your game world and what areas really need emphasis or alteration.  Metrics and statistics are key to helping you understand not just whether you're implementing your vision, but also whether the players themselves are participating in your vision.

You should solicit feedback and engage with everyone you can.  Sometimes these can be of a technical nature, such as the excellent #bbg irc group on freenode that Ben frequents, or simply communicating with your players via a forum where they can tell you what works, what doesn't and what should be changed etc.  Your game only exists to be played, and you have to listen to what the players say - no matter whether you agree or not.  Write a blog about your experiences with HTML5 - and update twitter (we came late to this and regret it, nowadays).

7. Which elements of HTML5 would you encourage PBBG developers to integrate into their existing game first?

I'd go with Canvas and the gfx elements.  The immediate boost of seeing something that you've never seen before in a browser is going to help your confidence hugely.   It's also going to differentiate your game from the HTML4 pbbg games, instantly, as HTML5 is still very young and fresh.

Avoid tinkering behind the scenes for ages.  Whilst there are many amazing Network I/O features in HTML5, you could spend a lot of time on aspects that are entirely invisible to the end-user, and which they won't appreciate.

Sure, you can't neglect these areas in the long term either, but if you have an existing game then you are presumably handling the traffic adequately already.  If you don't have an existing game, your traffic on launch is unlikely to test the old ways of client-server communication and so your players won't even notice.

8. Was HTML5 always a planned consideration for Illyriad?

We'd decided from the very beginning that Illyriad was going to be an HTML-only game without using any third-party plugin software.  So, well, in a nutshell, yes!

9. What do you see as the next 'big thing' for both Illyriad and perhaps PBBGs specifically?

We are very excited about the possibilities inherent in both the audio element and WebGL.  Our WebGL demo is only just beginning to scratch the surface of the kind of things you can do, and having fully rendered 3D in the browser, using the client's hardware acceleration is really going to shake things up.

But yes, here's our next big thing:  http://www.youtube.com/watch?v=0bZmt3yQzmE [Video: Illyriad's WebGL testing]

In the future, browser-game is going to challenge, for the first time, the graphical and audio supremacy of the Triple-A box-set game.  We can't see any reason why an incredibly deep 3D game like Eve Online couldn't be developed in Chrome on HTML5 in the future.

10. Can you imagine having used alternative technologies to HTML5 (Java applets, Flash, etc.) to implement features of Illyriad had elements of HTML5 not been available to you?

Nope, we simply refused to contemplate plugins from the very start.

The simple fact is that, as a developer, being able to develop once on one platform, and then have the game work across every browser, OS and device, looking and feeling the same regardless is worth its weight in gold.  You don't have to worry about whether the device supports Flash, or even if it does, whether the version of Flash is compatible with what you're trying to do.  What with devices (and corporate firewalls!) becoming increasingly security paranoid, there's also a great advantage to not requiring any kind of download.

11. Do you have any notable tips and tricks you've picked up whilst working with HTML5 that you're prepared to share with us?

Yes, absolutely - we're very open to helping out with what we've discovered, and the more esoteric, the better. 

We publish a blog (“Beneath The Misted Lands”) where the dev team put in their best tips, tricks and workarounds (with code samples) for things from closing memory leaks to managing request throttling.   

We'd also recommend a look at the Box2d (Of Angry birds and Cut the Rope fame) tutorials on Seth Ladd's blog as they offer some very interesting possibilities. 

12. Which areas of the HTML5 specification do you think that Illyriad and other PBBGs would likely benefit from the most?

Most have already been mentioned above, but we'd also like to emphasise the browser local storage access in HTML5.  Gone are the days of caching simply cookies and images; you can now (invisibly in the background) upload fairly hefty objects such as detailed 3D WebGL textures so that when they're called on in the game they are ready locally.  With the right combination of gameflow design and background caching there's a great benefit there, and the bandwidth limitation of a pbbg (vs a DVD-installed game system) can be largely negated.

13. What more can be done to encourage developers both new and old to begin working on PBBGs of their very own?

Well (highly specific to the UK, I guess) but I don't think the government's cancellation last year of the proposed tax relief system for video game developers has helped, so give Her Majesty's Govt a good kicking from me! ;)

On a more serious note, though...  Well, the possibilities in HTML5 are amazing, but they do currently require some fairly heavy old-school programming knowledge, especially at the gfx end.  When someone comes up with a really good WYSIWYG editor for some of the HTML5-specific elements, then that'll help immensely.

I also don't think people should be afraid of taking the plunge.  When we first started Illyriad we were most concerned on our SWOT analysis about a big, established strategy/empire game coming along with their IP and their already-created 3D models and, basically, blowing us out of the water with their established playerbase advantage.  I'm deliberately not naming any names here, but you can see these old game franchises (the ones that have simply taken their old titles and stuck the word "Online!" at the end of it) desperately failing to do anything that anyone actually wants to play.  16 players at once or occasionally collaborative 2-player quests does not make a game "Online!" in any meaningful sense.  The original BT CompuServe MUD in 1987 was more "Online!" than that.

Finally, I'd say that you have to embrace the openness of the web: People do genuinely want to help.  Illyriad initially used a small army of unpaid volunteers to write quests, bughunt, make suggestions, crunch numbers for game balance purposes etc etc; and there's a genuine willingness in the communities, be that pbbg, or HTML, or simply your playerbase to help make the game better.

Andrew J. Baker, speaker at the first HTML5 games conference onGameStart, developer of the work-in-progress HTML5 game Fleeting Fantasy and promoter of the browser-based games channel (#bbg) on irc.freenode.net

About Illyriad: Illyriad is a browser-based HTML5 Massively Multiplayer strategy game, in which players take control of cities, engaging in city-building, quests, diplomacy, warfare and trade. Developed by London-based Illyriad Games Ltd., Illyriad went live in 2010 and was officially launched on May 16th 2011.