2. Previous Working Relationships
Once we had surmounted the initial hurdles in the deal and convinced the involved parties that Baldur's Gate was indeed worth revisiting, we were able to bring all the involved parties together and sort out the core terms.
The deal was a bear simply because of the number of parties involved: Atari, Hasbro, Wizards of the Coast, BioWare, EA and us (Beamdog). We managed to get the terms sorted quickly through great former relationships. It actually took longer for me to get through the wall at Atari and talk to the right people than it took to put the core deal together. I remember meeting the fellow who had shut me down for a few months at GDC once we had the deal underway. He offered, "You must hate me," when we met in person, to which I responded, "Not anymore."
I'm very happy to have had some great friends such as Richard Iwaniuk, Greg Zeschuk, and Ray Muzyka, who were exceptionally helpful in getting terms agreed to by the involved parties and signed off in a reasonable timeline. Richard was able to line up the key people on that end and quarterback the requirements. Derek French was also a great asset at BioWare, as he assisted us in our spelunking for code and art assets throughout the course of the development. (Key lesson here is without someone on the ground running a contract in each group, it just doesn't happen.)
I'm very happy to have worked with many great people in my former life at BioWare. During my 15-odd years there, I started as one of the original founders, worked on design, programming and art on Shattered Stee", led Neverwinter Nights from a one sentence description of "everything in the D&D Box" through seven years of development and two expansion packs. I served a tour as director of technology and returned to my first love, directing video game projects, before leaving in 2009. One of the great things about the game industry is the amazing people involved in it and I look forward to "paying it forward" whenever and however I can.
3. The iPad and the Touch Interface
We proved you could produce an epic, massive RPG in the tablet space, and the market was starving for such a project. The success has been awesome and the feedback from the fine fellows at respected media outlets like Touch Arcade and Pocket Tactics has been a real positive.
Beamdog co-founder and veteran developer, Cameron Tofer, went crazy on the tablet interface early on. We had an iPad version running and semi-functional almost a year before we shipped. Every new build brought radical changes as we explored the logical means to perform actions in Baldur's Gate using a touch design paradigm.
In the end, Cameron's approach broke down to favoring common actions and giving them the least complex control mechanism. In BG, the most common action is to scroll the screen, so we used the photo app UI standard of a touch drag for scrolling. The second most common is moving/interacting, which we mapped to a single touch. We initially favored group selection, but after a number of clumsy iterations we realized the vast majority of actions in BG are performed with either a single character selected or the entire group, and as such, group selection could be moved to a modal button. When you play the tablet version of BG: EE, it is the constant iteration you feel.
We've added to this level of polish recently with a further enhancement we call smart radius. Smart radius allows the user to have the precision of a mouse with the less accurate nature of a touch interface, making it easier to enter buildings, pick items from the ground and so on.
We'll continue to iterate as we go, making the game better as we move along. From an overall UI design, we wanted to keep the original feel, but re-base around the new resolution of 1024x768. My priority in UI was to make the user portraits as large as possible. We listed all the required UI elements and slowly re-constructed all the art and panels around the new portrait sizes, we also tried to embrace some UI guidelines from Apple for the iPad, trying to keep buttons over 44 pixels in height (we only failed in a few spots like class selection). The overall process was challenging, as we tried to balance re-writing the entire UI code while keeping the existing complex (and overly integrated -- damn you, Scott Greig!) systems working.