Cerberus Daily News: The Making Of

So, the new site's pretty awesome, isn't it? It still has a long way to go, but I thought I'd throw you some ideas about where we started out from. When I first brought up recreating the site with an all-new look to the Elite Few who would help out in the endeavour, I started out wanting a design that was visually interesting, and touch-friendly. If you're on CDN with an iPad, you'll notice that everything fits neatly. This was entirely intentional. I'd argue that we're not quite perfect, but remember that this is site version 1.0. Dippy's working hard all the time, and I'm sure I could cajole Aphin into doing more touch-up work.

Here's the first, raw sketch, a look at the CDN of the future:



I doodled it up one day at work, scanned it, and sent it in. Following that, we built up a basic "this is how things should go" flowchart. This time I had a much better app to work with (OmniGroup's OmniGraffle), so things look a little better:



And, finally, with a basic idea in my head, I finally sat down and did some Serious work in Pixelmator to sketch up the site proper. I still wouldn't mind if the site inched closer towards this look, but don't think that I'm not thrilled with what we have so far. If you're curious, I took my inspiration from Polygon and The Verge, which I think are excellent examples of clean, interesting design for news sites.

Here's the original (and possibly future) look of CDN:



I'm sure Dippy and Aphin have some interesting tidbits they'd like to share. Or, maybe, they'd just like to curse me out a bit for having insane demands. I'd been sitting on this stuff for a while, and I'm sure that some people would enjoy the whole "making sausage" aspect of the site. So here it is!

The last picture looks amazing! You guys have done an amazing job, kudos to you all!

So here's my side of the story, one day Harrad mentions his idea for designing the site and looking into the different variants of phpbb, so as that goes on I start to think.

The number one gripe that the old board had was the one character/one account thing. People miss posting and other stuff were a real issue. The other issue was how everything even the RP posts looked like a message board with a signature and side stuff.

So on a cool evening, (I think,) I asked if Harrad would be cool if I tried something out that would allow a single character have multiple accounts. This fostered his interest and so a few weeks later I came with a small, working prototype.

These are some shots of this first prototype:




Notice how even in this early stage we already were testing it, this is hyper important, if you don't test early and often you might miss issues until they are too late.




And here we see the main selling point. Even at this point people who used it could see that the ability to have one account / many characters was 'good' and that doing so could potentially save many headaches.

I should have run a DIFF between the current incarnation and this first prototype to see how much code is the same but I suspect that much of it has changed over time, I scrapped and restarted the project at least once only saving some the most basic of tech; The Bracket Code engine.

Bracket code is the name for the bbCode inspired system we use for formatting the text. There are however some clear differences between the two; bbCode has a lot of legacy an quirks I wanted to minimize these. Bracket Code system is reliable enough to function within the environment and because I made it from scratch I understand it.



The most important lessons I learned from this project is that I should make more project so I can estimate my time better. Secondly communication sometimes failed, when I said one thing others understood another for example I estimated that the 8th of October would be the moment the major subsystems would be working while Harrad and others understood it as being ready to deploy which was not the case. Thirdly Leadership and accountability, I believe we would still be struggling with it if we didn't tighten our pants and started being accountable to each other.

I was falling behind the coding, concentrating on insignificant details while I should be coding, and others were in the same position. At the start of September we just agreed that we should be accountable to each other, that each of us should be able to honestly what he did for the project that day, understanding that we all have day jobs and lives so that not every day could be productive.

In the end I would argue the project was successful, could it be better, sure and it will get better. Bugs gets squashed, design gets refined and new features will happen. In the end we've made a new forum system from nothing something that can't be said to happen every day.

As for the future, I hope that once some more features come online and some more debugging is done I will be able to release it as an open source project. I honestly think that the 'niche' of roleplaying forums can have a real benefit from this system.

You guys should found a company and sell this thing. Seriously. I bet my ass there is a market for a light-weight community management system.

It's flattering, but there's dozens of different CMS products out there, and many of them have years of people building and tweaking them. The real strength of Valiant is that it's very effective for a niche situation, but the type of people who would benefit most from it aren't likely going to spend money. That said, unless plans change, my understanding is that Dippy will be releasing the source code, and hopefully with a bit of advertising and word-of-mouth, we'll see more sites adopting the software.

Because really, it's absolutely badass.