Unfortunately, <PlatformTarget>x86</PlatformTarget> is already in the project files. Checked last night.

Once again, I am really sorry it's not working. I've already started looking into SDL.NET, though.

All I do is leave waves of disappointment behind me 

Okay, clearly I love this because I've just been playing it the last hour and a half and haven't looked away from the computer.

Placed 15th. Have to stop myself!

Great work for such a short time. You really know what makes a shooter fun.

<Bezzy> OUCH!
<Bezzy> my mobygames rating just got down-bumped to 2.9
<Bezzy> which of you fuckers...
<Bezzy> HA HA HA
<jstckr>
<Bezzy> I'm not going to hurt you
<Bezzy> c'mon! fess up!
<Bezzy> we can have a jolly nice laugh about it

Sciere, thanks for the response.

As I said before, I don't particularly mind it being there... I just found the attention a bit confusing, because I had assumed that moby games was about finished games which were available to buy, rather than freeware/hobbyist stuff. But you've cleared that up. Don't bother taking it down, I guess, because I fully intend to develop it further, but slowly. Might as well be up there, if having a big catalog of all games ever is the point of it.

However: 3.6? Damn . You hurt my feeling!

I think I rediscovered the reason why nothing of mine ever gets released: the ratings cut deeper than I am willing to admit  .

Ohh! I had an idea like this, when I was thinking of fun to have with beziers! I'm an idiot for not following through because this looks really nice.

Awesome stuff!

When I "capture" the angels in my branches, it makes the branches split, right? I think I did it too much, because each of the branches died and sprouted too quickly again, so I couldn't keep capturing.maybe I missed the point, though?

I registered to clear up some stuff in the entry (and bump my score aggressively, of course  )

It is wrong, of course! But like, I don't feel like it's particularly justified being up there. It's an entry to a compo - not a finished game (I speak for myself only, obviously - there are plenty of games in the compo which feel basically finished and polished). Therefore, screw the scores (in my case, but not necessarily anyone else's).

And the score was something like 3.6 before I bumped it. Probably all it deserved. Out of ten.

I registered to clear up some stuff in the entry (and bump my score aggressively, of course  )

I dunno. I don't mind it being there, but it's presented like it's a final game or something. Who expects a prototype to get reviewed in the same breath as a final game? Feels a bit unfair, really.

Have people seen that someone has been adding compo entries to MobyGames, and rating them? This seems pretty out of order, although obviously I'm open to accusations of bitterness because our game gets a very low score. [but it's not a finished product!!]

In fact the "ratings" seems to be pretty wacky in most cases.

On the other hand, this pgwiki sounds like a great idea, and more of a repository of knowledge and examples, rather than the purported equivalent of IMDB that MobyGames is.

Pretty fucked off right now 

(can't post a link cos MobyGames is blocked here at work)

This has me in two minds. It's pretty cool that someone's taken the effort to do that. The description for mine is a bit off, but way, way more accurate than I ever would have expected.

But the game is really not meant to be finished. This is just a prototype for further development.

Don't particularly care about ratings. I'd rather hear an analysis of what's wrong with the game than some number, plucked out of the air on someone's subjective scale (but that's a general whinge. And I'm sure if I scored higher I'd instantly change my tune ).

But it IS slightly cooler!

[A Trip down Memory Brane]

Okay, so, I'm done with the Procedural Generation Contest. I'm not completely happy with how it ended up, but still, 22 days for a non-l33t coder like me... I shouldn't be too down on myself.

I wanted to keep K2 moving, albeit at a much slower pace than for the competition, so this is my thread to do that in.

I was looking back through some old files recently, and found a bunch of screenshots from the original K. I keep linking to

, but that is really blitzed by youtube's compression. I found an old, half-developed backup build of it, from before the time I introduced some really nasty bug which never got fixed, and I was sort of struck at how well the visual style holds up. I figured you might like to see some clearer screenshots from around 2001.

A Trip down Memory Brane

This was one of the original concept drawings I did. as you can see, I wanted more of a free camera, top down, remote-control style 3D game. Ultimately, it changed itself, though.

A ship concept from buddy JP, inspired by the flight-of-the-navigator ship. I wanted to go for some very abstract ships, although, as you'll see, there may have been some *ahem* Lucasian inspiration which distracted me from this goal...

Close up of the Tie Interceptor model. I actually made this in under half an hour, to explain to my friend how to model in milkshape. I'm not a modeller, obviously, but I was kind of proud of doing that in such a short time.
To get the line drawing effect, I simply loaded in models, figured out where the ends of polygon groups were, and drew gl lines around them. It was a cheap tricky to avoid making tools to individually craft the detail lines.

Experimenting with the landscape shading. This was using an environment map in a sorta bizzarre way. The same sort of idea is used for rim-lighting a lot in cartooney games, nowadays, like Mario Galaxy.



I gave cel shading a go, but because of some bug in the normal calculation, it didn't really look that great. Remember, this was in the days before shaders dominated everything. There's a lot of old-school hacks using the nuances of opengl which were clever at the time, but are now a bit redundant.

The original landscape engine was by David Hillier, and was meant for more realistic terrains. I dabbled a bit, and decided that my texture art skills would not be up to scratch, hence the totally abstract look.

Starting to get ships into the game.


Contours and shading, here, are generated by two passes (using oldschool texture combination). One is an environment map, and the other is a transparent texture, except for the stripe.

O look, that's not a tie fighter rite?

This was a slightly newer version of the engine, donated by David Hillier again.
Contours/shading turned into a clever trick. They used one texture, which had a stripe on the Y-Axis, and the shading blended along the X axis. A flat-on texture co-ord generator was used for the V component of the texture (to generate the contour), and the U component used the environment map method, generating the shading. Or something like that. I forget.

Starting to get game play in. The key idea to K is that the free-roaming cross hair fixes problems of parallax error. Most games really lock your camera behind the ship, because the shots don't line up properly with targets, but when you project the cross hair out to the same distance as the target, you reduce parallax error immeasurably. This allows for a freer camera. Unfortunately, all kinds of other problems meant that I had to have my camera a bit closer.

I gotta check this out tomorrow, dude. So tired, but looking forward to it.

[edit] Cool stuff!

As self appointed 3D reticule expert, I'd say you need to tween the crosshair depth between its current position and a target's position (so that it snaps less).

Or if you want to be really fancy, you generate a plane based on the three closest enemies to the crosshair, and project your line of fire onto that plane to give you your crosshair position - that gives you a "free" blend when navigating between enemies.

I wasn't sure how to get the colors shooting to work. Guess I should just read the docs, right?

Thanks for the feedback, everyone, and especially haowan for clearing up what's needed (cos I wasn't even sure :/ ).

The lockons: Yeah, they're not perfect. The "front edge" of the lockon arc doesn't line up with the crosshair (a bug I described earlier but didn't manage to fix in time). Also, the icon might be confused for a health bar. All fixable though.

As far as jerky controls go, yeah, there's some issues some people have been having. Might be a combination of mouse sensitivity and frame rate... I'll have to try it on crapper/better/more machines.

I was playing with the idea of "dodging" by jerking the controls in a direction. It kind of works, but it needs a lot more tweaking before I can write it off, and it depends on a lot of other things than just control implementation: You need more telegraphing of the bullet path to be able to dodge it. I guess I'm experimenting with a "tie-fighter meets bullet hell" mechanic... maybe it's impossible, but it'll be interesting to try it.

But yeah. I didn't have enough time to polish everything. Sorry about that.

Third page? Screw it.

BIRTHDAY BUMP

I think I really like C#, and would probably lose a lot of productivity going to C++ (also, that's my day job).

but yeah, I'm moving out of XNA after this... or maybe keeping the stubs, but making my own wrappers for graphics, sound, and input, so that I can switch between OpenGL and XNA's directx easily.

Plus, with OpenGL, I can probably get back that intentional alpha draw-order error effect which I didn't manage to nail in XNA (though it's probably due to a weird render state setting somewhere, which I couldn't find in time).

But yep. This game will slow burn for the time being.

There are no rules about what you do with your game past the deadline. It's completely up to you whether you continue developing it or stop.

I guess you need to keep the "finished submission" version of the game there for posterity, though.

Phew. Done.

I think that I didn't really make the most of the procedural generation part of this competition. All my geometry is parametrically generated, but the idea of shooting enemies, which would upgrade them next time... i didn't really have time to do it justice. It's there to some degree, but so much more could have been done with it if I started a little earlier.

Oh well. Hope you enjoy it for what it is: a prototype for a better possible game.

Seems that my problem really is with XNA itself, as most games that aren't XNA but still need .NET 2.0 work, like Dyson.

It doesn't work for me, and I'm not sure; either I failed at installing XNA or XNA simply doesn't work under x64.

X   N   A     

404

Fixed, thanks!

New, final version up. Refer to the first post!

Or if you're lazy: http://www.bezzy.net/Downloads/K2PrototypeFinished.rar

Thanks for the mirror, Haowan!

So far, the main bug seems to be that XNA is a pain in the arse to install. I'm really sorry about that, everyone. Should have learned my lesson after D@sher... but XNA is the thing I was most familiar with, and fast to use, so I didn't have any other choice.

Any other bugs/feedback would be highly desirable so that I can polish off anything outstanding tonight.

Bugs so far:

*Second time you play, at game over, the "next wave time" is huge, and you have to exit manually.
*Score on HUD doesn't catch up to regular score fast enough.
*No sound/graphics for thrust (last minute feature).
*Game is a mere shadow of what it was supposed to be.