ID:1089
 
Well, after about 9 hours of work (plus 3 hours for breaks), my 4k game is mostly done. I've got fall damage, crushing damage, chat, troll control, some graphics, bug-hunting, minor gameplay tweaks, and 67(!) bytes to go.

There is still some compression I can do; I've already saved a huge amount of space by stuffing data into the map file (useful loophole, that). Once I compress all my remaining procs down into one-liners (and half-liners and sixteenth-liners) and #define the hell out of it, I should have a little more space to play with.

The gameplay needs some work right now; the beginning of the game in particular is quite slow. Changing the starting positions on the map would probably help... but that's an issue for another day.

And now it's time to say goodnight. Goodnight!
We should recruit BYOND developers in Japan/China/India, so that development on critical projects such as this can go on 24 hours a day!

(I was up past 3:00AM myself last night, and I already had many hours of sleep deprivation, so I'm on what you might call a "natural high"...wheee...)
Critical? Heh, hardly. =) I have a whole month, after all, so it's not like I'm pressed for time; just space!

Recruiting someone from Japan wouldn't help much, anyway, as they're nearly in the same timezone as me. The US or even Britain would be closer. Heck, you guys could be recruiting me. ;-)
Good luck on finishing your 4K project!
67 bytes over? If you haven't already #define'd the crap out of it, then that figure can shrink to something closer to 500 under. I usually have #define working at a steady pace early in my projects, and tweak later to get extra oomph. Not to mention, if you're only 67 bytes over and your code hasn't been compressed to one line (let alone all procs to one line each), that compression alone may do the trick. In fact if your code is even looser than a semi-compressed state, you're liable to save a lot more.
Thanks, TerranWolf!

Lummox: I'm actually 67 bytes under. Well, 36 now. I have the following defines at the moment:

#define u #define
u l locate
u li(a,b) l(a in b)
u BK (l(/obj/b)in T)
u DE density
u C {X--;continue}
u cl(q) (l(/obj/c)in(q))
u is icon_state
u o overlays
u t istype

There are a few there that could be single-letter ones, but I want to get the game mostly completed first. Anyway, I have about 30 lines of compressed code, and two mostly-uncompressed procs taking up about 20 lines together. Most of that isn't strictly grouped under typepaths either. So yeah, I'm not too worried, even though I've still got a bunch of things to code. =)
Good gads, that's absolutely nothing for #defines. I always start out these projects with E for else, R for return, W for while, and F(a,b) for for(a in b). Usually any given project will benefit from those. It sounds like you could probably shave off at least half a K with more macros.

Incidentally, does that locate(a in b) syntax actually work? It should be (locate(a)in b) I believe. locate(a in b) should be the equivalent of either locate(0) or locate(1), which will always fail.
Yeah, probably.

I did notice that, but it seemed to be working so I left it alone... but on the other hand I just tried using locate(a in b) and it failed, so something wacky is going on with my code. =P Should fix that...

I don't actually use return at the moment. I use while 4 times, else 5 times, for(a in b) a and whole bunch of times, though, so most of those are probably good candidates (I'll do the maths later to make sure).