[subkamran]

This is awesome but man, reading the canvas portion was a bit distressing. I wonder why they didn't use a game engine to do this? All the work they did has been implemented already in several JS game engines, such as the one I help maintain (it's free and OSS), https://excaliburjs.com. We support all the features they needed including mobile & touch support. They could have also used Phaser (http://phaser.io) too I bet... that has WebGL support for even faster rendering on supported devices.

[madlee]

Hi, I wrote the majority of that part of the project (canvas stuff) & that section of the article – the simple answer is that I have a lot of experience working with the canvas API directly, but little to no experience using any of the popular JS game engines out there (I played around with Phaser years ago, but not very much). I don't think it would've saved me any time to be honest.

[subkamran]

That's totally fair, I get the sentiment. Great job nonetheless!

[Mahn]

...it's not like you have write assembly to get it done, the native canvas API is fairly straightforward. A game engine is a bit of an overkill if all you want to do is place pixels on a canvas.

[subkamran]

But they wanted a lot more than that. Engines like Phaser work hard to take care of browser quirks for things like PointerEvents vs. TouchEvents vs. MouseEvents or supporting mobile devices. Sure, it seems simple at first until you run into those kinds of problems and reinvent the wheel... learning an engine isn't terribly complicated but I understand the sentiment for a one-time project. It just seems like they did so much other planning but didn't want to plan the UI implementation to the same degree?

[brianwawok]

What is the learning curve on a game engine for not game engine developers?

They used tools they knew and knew how to scale. Almost always the tool you know is better than the perfect tool. They know redis and websockets, and they made it work. Beats using some engine know one on staff has ever touched?

[subkamran]

I guess it depends on the engine, or maybe that was a rhetorical question. It doesn't take a month, or even a week, though--maybe to master them but not to learn the API. We've worked hard on our little project to design our engine for non-game developers or veterans, it's only a few lines of code to move a square, or draw an image, for example. But I understand the sentiment, especially for a one-time project.