My home desktop montior's been looking a little peaky lately. Hope it doesn't die on me soon; that monitor's been around for - what is it, six, seven years? While it might be time to look for an upgrade, I've grown attached to the old thing and would hate to have to replace it.
Development is not going as fast during the week as it did on the weekend. I've been a bit out of sorts and not thinking as straight as I should, so I tend to mentally drift off when programming. I'm also stuck on Blocky Man music; nothing I've tried seems to work. At least though things are moving now; I'm feeling much better about that. Guess I'll get more done on the weekend.
Part of why this present stage is so slow is the design of the graphics library is one of the more complicated bits, and I'm figuring out how it should go as I go along. Hence I tend to rewrite things a fair bit once I get uncomfortable with the way things are going. It's also a bit dry and dull in parts, but at least I know this is a one-time cost. Once I've got this engine finished, it should be much easier to write the Mark II version with a previous code base to work off. I'm also extremely slow when it comes to deciding on function names; I can never feel comfortable with anything I decide! Not sure how normal that is.
I'm also a bit worried about my present tendency to use singletons for most things, such as the logger, graphics and input systems. Well, the logger is pretty darn hard not to make a singleton as it's needed by everything, but the other ones it's debatable. I suppose I'm just a bit wary of global variables, although it this case it's a lot neater as everything is compartmentalised. For the Mark I engine I'm sure it's fine to do whatever the heck I want with singeltons to see if it'll be a problem.
The danger is the tendency to daydream about what the next project should be rather than work on finish this one. That's already started, of course, and I've got some good ideas recycled from the ones of the past. There's plenty of ideas that were scrapped because I wasn't quite ready for them from an art perspective; but now I'm a bit more comfortable with drawing so my range of possible games has increased. Also working with Blocky Man has left me hankering to try to some neo-retro arcade type games; the sorts of ones I'd have loved to have been able to make back in the eighties. But all that's got to wait until Ice Slider is finished, which is all the more reason to keep on at it!
One other issue I'm sure one of you can help me with is the size of OpenGL textures. At present I'm using OpenGL as the display mechanism for my 2D sprites, which I'm presently planning to use as sprite sheets, storing multiple images to one texture. However, I'm not sure what maximum size I should make these sheets. I'm (fairly) sure they have to be a power of 2 in dimension, but do most video cards these days support 256x256 or 512x512? And if I wanted bigger pictures (say for a logo or background), that'd mean I have to split them up into blocks, right?
One of things gamedev related things I'm doing (while I'm procrastinating over deciding function names) is reading through some of the latest books I bought. At present I'm skimming through "Designing Virtual Worlds" by Richard Bartle, which is an interesting book on world design for MUDs and various other multiplayer online worlds. I actually have no interest whatsoever in making a MMO game myself, but I am interested in how they are made from an academic game design perspective. I haven't read through much so far, and I'm skipping around a lot, but from the small bits I've read all I can think of - why the holy heck do so many people on these forums actually want to build one of these?! I knew the technical challenges were great, but the design challenges seem to be even greater (for example, just getting a working economy is a Herculean challenge), and even if you do a half-decent job you'll still have a significant proportion of your players hating your guts (just down to human nature).