Beeraholix

Came across this cool site that tracks beer news... Beeraholix

Hey there... blimpy beer

How awesome is this? A singed bottle of beer from the Hindenburg.. http://news.bbc.co.uk/2/hi/europe/8352032.stm

Google Adsense for BEER

Wow - just looking at the Adsense section on the right of this blog and it looks like these days it actually works! When I first set it up it didn't really know how to deal with BeerMogul, but I'm seeing ads for Dan Murphy's (a local liquor chain), a guide to pubs in north east England, and some RPG games - congrats Google!

What do you see there? Send me a comment - and have a click on them to raise some revenue ;-)

Nathan.

We're still here...

OK so I know it's been nearly 2 years since the last post, but after several false starts things are moving again. I'm taking a more simple approach. Apply the new look that we had designed ages ago and "fix" the god tool.

I've just finished applying the new look & feel to the main pages - the others will follow after re-launch, and I'm about to start on the god tool. This time I'm going to write it as a stored proc in SQL Server 2005 to avoid all of the performance issues I had when running it as ASP. Wish me luck!

P.S. Thanks to all those that kept posting comments on the blog - you're the ones that drove me back to it!

Regards,
Nathan.

God Tool?

Sure BeerMogul looks like a pretty simple game when you're a player - you log in, buy and sell properties & beer, then every day something happens when people either come to your pub, and brewers get to manufacture beer.

What is it that happens? We run the "God Tool". Probably the most complicated part of the system, the God Tool does a few things...

1. It simulates the manufacture of beer, one brewery at a time. It considers the quality of the brewery and the quality of the features installed, then looks at the investment that a brewer is proposing to make in order to determine exactly how much beer gets produced in a single day.
2. It processes all of the "contracts" which represent committed trades between brewers & pub owners.
3. It simulates a night at the pub! Punters are generated (including the amount of money in their pocket, a taste for a certain quality of beer, and a tolerance to price & quality variation).

As you can imagine, this was quite a time consuming exercise, and it used to bring the server to it's knees. It ended up being the downfall of the game - keeping it running successfully in a reasonable amount of time was VERY time consuming and demanding.

How are we going to fix it this time? We're looking to follow a "progressive" processing model. Rather than processing everything in one "big bang", we'll progressively process all of the God Tool activities throughout a 24 hour period. The good thing is that it will give people a chance to see the status of their properties change throughout the day - you'll be able to take a look and see how many punters are in your pub at any given time - or look at your brewery to see how your manufacturing schedule is going.

It's still a theory obviously, but pretty important. We have to get everything right if we want to make sure the game lasts this time!

Project Beerstar?

So we came across Project Rockstar today (http://www.projectrockstar.com/). From all accounts it's a successful browser based multi-player game - not too dissimilar to Beermogul! Looks like it was developed around the same time (2000/2001) and has the same basic elements - you sign up and receive money with which you have to establish a property (a band). You can then trade with that property to receive money.

I like the fact that players can choose to be record companies or venues (mutually exclusive) in order to trade off the good fortunes of others (by recording their material or charging punters admission). Also interesting is the fact that they have to restart the game occasionally when it becomes too lopsided (detailed here: http://en.wikipedia.org/wiki/Project_Rockstar) - this is a problem that we always encountered in BeerMogul - we never did find a solution.

The daily email is a little bland for my liking - wouldn't take much to spruce it up and use it to really attract players back into the game.

A site that looks after their advertising placement said that they receive 20M page impressions a month - looks like a great way to pay for the game. At 140,000 subscribers (only 10-15K active) that means the pages per registered user rate is around 143 per month (that's 20,000,000 divided by 140,000)! That sounds very high (I wish we still had the web stats from our original proof of concept), but very encouraging. Makes me think that if we bank on about 25 per month for Beermogul that we'll be very conservative and have an achievable target!

Have a look at the site and let us know what you think.

I can't play - I'm an alcoholic

I wrote the original BeerMogul code from the ground up. It was a bit of an experiment and a learning process - learning about ASP and learning about web development in general. I was learning as I went, and the result was a bit of a mess technically. As a result the system would experience regular issues - getting stuck in infinite loops, poor database performance due to badly formed queries, general IIS crashes etc.

Once the game became popular it was creating too much load on the shared server - to the point where the daily run of the God Tool would bring the thing to it's knees - drawing complaints from other customers of the hosting company. We were eventually thrown off the service (thanks Hostworks!) and had to quickly find another solution. We were lucky to find Webhost4life pretty quickly. Sure it was an unlikely sounding concept (pay once, website for life!), but it just worked, and SQL Server was included in the base plan for free! I quickly ported from Access to SQL Server (thanks to the fantastic Up-sizing Wizard) and things were back on track.

Given the unstable nature of the system I would sometimes have to write pleading emails to the support staff asking them to kill a database connection, restart IIS or even reboot the server. After doing this several times I had a brilliant idea - if I could get one of the support guys interested in playing the game, perhaps they would have more interest in keeping it running smoothly, and they might spot issues before I found out about them. The next time I had a problem (after the issue was resolved) I told the support guy (via email) about the concept and asked if he wanted to play - his response? "I'm a recovering alcoholic - I can't play".

A Lesson from Rock

So my band played a reasonable gig on the weekend, but we struggled in one main area. We have a couple of "core" songs that we have been playing forever - and we feel like we know them really, really well, but on Saturday night they completely bombed making us sound like bigger amateurs than we already are! The thing was that we also played some completely new songs - songs that we'd never played in public before and those ones went over really well - like we'd been playing them for years.

It's not like the "core" songs are really hard, they're not, it's not like we don't practice them, we do, it's not like we don't enjoy playing them, we do - It's more about the level of attention and concentration that we apply to them. When it gets close to midnight on a long Saturday it's really easy to mentally relax when we get to one of the "core" songs that we feel like we know backwards. The result is that we get lost in the song - was that the first chorus or the second? Is the bridge now, or next time? What note does that riff start on?

The other thing that becomes really evident is how reliant on each other we really are (although nobody cares to admit it). A small miss-placed drum fill, guitar lick or vocal ad-libbed in a live situation can really throw out the rest of the guys in the band - did that mean that he thinks we've already finished the bridge when we only just started? It doesn't take much to send the other guys off on a completely different course - even with the best of intentions.

Last week the same thing happened in the office (my day job of course - Beermogul doesn't have an office!). We execute complex software releases, unusual developments, subtle product enhancements, manage external contractors and negotiate large commercial agreements on a regular basis, but we came completely unstuck sending out a simple email announcement to our customers - a "core" activity that we've done a thousand times, one that is incredibly simple. We mentally relaxed on the simple task and completely bombed!

In an effort to increase our productivity after a recent pep-talk the manager involved played things a little different - "We need to do things fast if we're going to survive in this commercial world". By turning up the heat he hoped to drive his team a little faster, but things went off the rails with the wrong communication going out to the wrong group of recipients. A simple mistake but a painful one.

The lesson? Hmmm, pretty obvious I guess - concentrate on the "core"' activities, and don't let yourself mentally relax just because the job is a simple one because that doesn't mean it's not important.

Oh, and don't drink too many beers before you go on stage! (I knew I'd get a Beermogul link in somewhere.)