If you're new here, you may want to subscribe to my RSS feed. Thanks for visiting!

Firstly thanks to everyone who helped get the word out about fav.or.it the level of support and overall interest in the project has been amazing. We are planning on revealing a few more secrets about the project over on the fav.or.it blog over the next couple of weeks but I am sure to keep reminding everyone at the time. As am sure a lot of you have noticed my article per week rate has plummeted over the last couple of weeks as progressively my blogging time has been eaten by a ravenous development monster.

I hope to return to a normal post rate after we have launched the beta and crunch time is over. The hope is that I come out the other end (preferably alive) but also with some great PHP examples to show off (I am playing with lots of microformat stuff at the moment) and a maybe with some lessons learnt from the experience.

I have in the past had to put in some long evenings for previous web projects but none have had the urgency of our current development and it is certainly starting to feel a bit like being back in the games industry with the excitement of getting something (hopefully) fresh and innovate out of the door.

Short story

While working for Empire Interactive I was given what to most people would have been their worst nightmare. Firstly it was a very niche game (Napoleonic strategic turn based warfare), secondly it had been in development for nearly 4 years (at the point of me taking over control) and looked like it would never end. On top of this the team had some serious deadwood including an artist who had a very unhealthy obsession with pigeons who he fed all day (rather than actually drawing art.)

I actually really enjoyed the challenge and after removing said deadwood got on with making the project move forward, this generally involved doing something I very rarely do (as a manager) which is to get involved with the code itself, but in this case the team had not a single clue how to write a AI for the game and needed a gentle push in the right direction. My code was not used but it first gave the lead coder motivation in that A) If he didn’t get it done he now knew I would do it for him B) He now had some working sample code to work from.

So six months down the line I have streamlined the team somewhat and we actually had a working product, beta had (sort of) been reached and I was down at the offices ready to pick up the gold master candidate to take to the CD duplicators to get it shipped. For those of you who do not know but when you take that disk to the manufacturers there is very little going back, the cost of duplicating 20 thousand CD’s then finding a bug is kind of expensive and not a good career move.

I had hoped it would have been an easy walk in, pickup disk, drive away. At 9pm and about the 10th build of the day I was not feeling a lot of confidence in the reliability of the final product. At 11pm I had probably drunk too much coke/caffeine, the duplicators have deadlines and the speed at which I was going to have to drive to get there was into the ‘get speeding ticket’ area. A decision was made and the final CD was in my hand. High speed drive through Oxfordshire had me at the door of the duplicators at 11:55pm, press buzzer; hand over CD, drive home slowly. Hope to god it is going to work.

Although the final push towards launching fav.or.it may be as frenetic the bottom line is that if it still has any bugs we know it is not going to cost us a fortune to rectify it.