Saturday, April 05, 2008

Game Development Status 4/5/08

monthly total 2 h 24 m; yearly total: 24 h 30 m

Things are moving right along. I created a new test scenario in the database. I found 6 places in the code where scenario numbers were hard-coded into configuration and reporting logic. I coded around them and got it to work. This test scenario is the one I have been using for the last couple of years. It's based on the Opening Moves, but has only one division on each side: French IV/2 and Russian 8th Div.

In general, I'm continuing to work on decoupling the AI from the main thread. I've made good progress. Movement is working fine. I'm now working problems related to conduct of seems AI infantry units are firing at target units further than one hex away (a no-no according to the rules). I think the problem is that the AI units' locations are not being updated after movement. When the AI plots fire, it is basing this on its movement plots' locations, not on realithy.

Aside from this I'm working a small number of bugs that have come up during the AI decoupling. This helps to take my mind off the large number of bugs already on the list. I'll go though a stabilization pass shortly.
Battlestar Galactica Rapid Recap

Nice. I hope they do another one after the dust settles. Beats taking the time away from other things to watch the episodes.

Friday, April 04, 2008

Wish I Could Get a Redo on the Game

I'm studying for the Microsoft 70-536 Certification test. This covers .Net 2.0 Framework Application Development Foundation. (Visual) Basically, this means "how to get stuff done in .Net". All kinds of stuff. I'm working through the chapter on Serialization now and just read a sidebar on serialization internals that talks about how the ObjectManager handles forward and backward references to other objects. This is a recurring problem for me in PC Eylau. As I add new data to be saved, I frequently have to write code to fix object references after deserialization (loading a game file).

I really wish at times that I could take a few hundred hours and re-implement the game in C# in .Net. This would fix all my serialization problems and all the memory leak problems, it would give me better exception handling, and I'd be able to use generics for type-safe collections. Plus, I think I could trim off a lot of unnecessary code...C# (heck, VB.Net, too) and .Net 2.0 are a bit more terse than native VC++ and MFC for quite a few things I'm trying to do. Of course, on the downside I'd have to reinvent some wheels.

Maybe I'll do this in Eylau 2.0. For now, I must spend the available time to get the darn thing (AI) to work!