The Multimedia / Games Blog of Adam Kidd

Meh

2008-06-29T18:11:00.001+01:00

I'm going to make a new blog as part of my portfolio website I'm going to be working on shortly.

This one irritates me..

What's a hiatus?

2008-05-13T20:42:00.001+01:00

I've not posted here for over a month now, since I only got the internet back a couple of days ago after moving flat, it didn't cross my mind really until now.

A fair bit has happened since I lost the internet - I've started working on the unreal level design, got a job (starts tomorrow, yay money!), and realised how incredibly little I know about maya at the moment. The latter is the biggest problem right now, since I need to use maya to create static meshes for the unreal level!

As a "quick" remedy, I'm going to watch the 13 hours of video in the 3dbuzz advanced modeling dvd's for polygon modeling up to and including the modeling of a socom pistol. Having done this, I'll need to learn how to do texturing in maya for exporting to unreal, which should be a much shorter task. I'll also have a sweet socom pistol model which I can put in a modeling portfolio, which will be nice!

That's got to be done by the end of thursday ideally so I'm setting a tentative deadline for that being done by then. Following that I need to start making the level in unreal so going to be getting some tracing paper out and making sure of exactly what I want to do with the level. Since I had an idea today of having a long glass corridor I clearly need to think more about it since that would be just about the lamest thing you could have in an unreal level.

I type too fast.

3rd year projects o_O

2008-04-04T23:20:00.003+01:00

Just pretty much skim-read the information on the third year projects, and I think the project I'm working on with Pawky Designs will be suitable for my live client project, but as we're aiming to get that released in its first stages by september, I don't know if it will qualify!

There is undoubtedly a LOT of work to do on this site, probably enough to get a first class masters qualification in fact, and the potential for ongoing add-ins and updates is very significant, and I could likely spend the entire year just adding more things to the site to make it even better.

I have a lot to think about in regards to what I would like to do after I finish because I'm still teeter-tottering between web design and 3d animation, and a combination of the two doesn't appeal greatly to me as I've seen it done so poorly before.. but then perhaps that's why I could be good at it.. because once I've finished the final year I will undoubtedly be advanced in both areas (since I will be putting a load of time into improving my 3d work) rather than just the web design as I am currently.

*shrugs*

Tiny things

2008-04-03T22:01:00.010+01:00

Just had a few thoughts while working on some of the things for the project outside uni, first of all I forgot to mention in my previous post that it is not a solo project, it is a joint venture with a company called Pawky Designs based in the Hive.

It's a student company, and is owned by the guy that made My Magical City (see www.mymagicalcity.co.uk) which won an award (I forget the name..) with a prize of £10,000 while it was in the concept stage which is what funded it to get where it is now.

I will shortly be starting a business in the hive myself if possible as well: I've got an application form ready to fill out but I'm not sure how I'm going to go about it since at current I would need to be in the hive on 1-2 weekdays a week for 12 weeks starting in june, which is clearly not ideal for job seeking and making money over summer! I may have to stoop to the level of call centre work again as the flexible hours and generally high level of pay could make it worth my while to stay up here in Notts.

The new project is slightly similar to my magical city in that its going to be using (true) isometric pixel art, which I've done a little of in the past:

Note that the above is not truly isometric, as it uses an angle of 26.565 degrees instead of 30 as it should be. This is the standard you will see on the web however made by people such as eBoy and Gunstar Red as it gives very clean lines.

I've never been one for copying however, and while it may be slightly easier to work in false-isometric, I'm all for us making our own style!

I've started to make a few pixel people (and tv's apparently) which you can see below:

Fubar body:

Much improved:

Working outside of uni

2008-04-02T11:55:00.002+01:00

Been starting work on a secret project outside uni (NDA's stop me from saying anything really!) and it is going to be absolutely awesome once its done, problem is, the coding required is nothing less than immense and is a far greater challenge than the one I had on that damned game I was making!

So at the moment, hoping for a september release which may or may not be likely, but will be easier to determine closer to the time! Other than that I'm seriously working on improving my php, flash, maya and mysql skills ready for doing this the best way I can.

Other than that, I've also got to start on the UT level shortly (this week!) as I have big plans for it!

This is awesome

2008-03-15T22:41:00.000Z

Need I say more? Watch this!

Link

Recovery

2008-03-12T20:54:00.002Z

After a very long 36 hour stint before the hand-in, I've now had time to recover (which involved lots of sleep and a few pints!) I'm going to keep on working on the game, though with a few changes to make it more suitable as a full platform game engine, and also create my own flash-based level editor for the engine. It will in a few months be a very nice engine that I can use to create platform games (and a few other types of games) relatively easily!

I will not be posting the game as it stands now on this blog, nor will I for a while, as I intend to procure sponsorship for the game once it is completed, and, as it is an original idea, I don't want it to be stolen by random internet users!

I will shortly be beginning work for Pawky designs in the Hive on their website mymagicalcity.co.uk [plug!] and other work they get soon as their in-house programmer, which should be very good not only as a paying job (on a per contract basis) but also for improving my knowledge of various programming languages and also improving my portfolio and design skills.

So lots going on at the moment, will post more soon!

Late night programming

2008-03-07T01:13:00.002Z

Given that I've got so much work to do on this game, I'm not going to sleep over the coming few nights until I've met my targets for that day.

For tonight that means I need the second half of the game working. Unfortunately due to a very busy day flat hunting, its now 3:30am and though I've made loads of changes including fixing the bug that made the character warp occasionally (woohoo!) I still haven't done that necessary part.

So now over the next hour, or two if I'm unlucky (I think it should be fairly easy, sods law says it won't be) I'll be adding this functionality and then uploading my test here! Then sleep!

The deadline approaches..

2008-03-06T02:46:00.002Z

It's late at night and I can't sleep, partly because I keep thinking about all the cool things I could do with my game, but probably mostly because I slept till 2pm today!

Unfortunately, with the deadline fast approaching, my game seems far off completion, and to be fair, despite spending already probably around 200 hours or more on the game, there is still a huge amount of features I want to add, some of which I simply won't be able to because my eyes are bigger than my stomach so-to-speak.

So focusing on the important parts is key over the next few days, and without fail I will have the second half of the game working tomorrow (eyes peeled for a demo on here people!) and will hopefully have started adding the enemies/things to make the game challenging.

It seems like I haven't done much so far from what there is to show, but there has been a huge amount of thought, some learning of advanced trigonometry and matrices, and some seriously complex coding (for me at least, this is probably the most difficult coding I've ever done!) to get the game where it is right now. What is left seems simple in comparison to the hurdles I've overcome to get the character moving smoothly!

So now the targets for the next few days which I will meet unless I run into some unlikely difficulties with some parts are:

Thursday: Add second half of the game in, have character moving smoothly in all cases possible.

Friday: Add the bad guys and the end screen with scores etc. Add checkpoints if possible.

Saturday: Create the full level and boss level and do some testing to make sure it works. Add start screen/preloader and checkpoints if not done on friday.

Sunday: Improve character graphics and add necessary animations

Monday: Miscellaneous tweaks and phidget implementation. Add a tutorial and high score board if time allows.

Tuesday: Any remaining tweaks/improvements.

So in all, a rather busy few days coming, I think the coming days should increase the total time taken to over 300 hours, which will have been well worth the time spent!

It's been a long week..

2008-03-03T12:47:00.002Z

..but finally, I've managed to get my character moving smoothly over slopes. I tried various methods of rotating the character and trying to hit-test him against the slopes, but the hit test simply wouldn't work. There was absolutely no reason why it wasn't working, and to be honest I think I probably ran into a bug in actionscript.

However, using a single pixel bitmap to work out how far the character needs to rotate worked perfectly and here is the proof in all its smooth slope-gliding glory!

Link

Of course!

2008-02-27T11:30:00.003Z

Got my code semi-running after yesterday, unfortunately the character immediately falls through the floor, even though it is detecting he hits the floor! It seems in my rush to create most of the states for the character, I forgot one of the most important - where he's standing still! So at the moment after he's created he never hit tests at all, so thats on my list today!

Since this code is all written and just needs bug testing now, the character should be moving properly by the end of today, a couple of days later than I'd hoped but this is one of the last major milestones before the game begins to look like I intended!

Once the character moves and the second half of the game is in, I can concentrate on performance indicators, start, finish and loading screens, and finally, making the actual level for the project and improving the graphics. After that, its simply a task of adding more features to the game, which will be a particularly interesting part!

Don't code too much at once

2008-02-26T21:37:00.003Z

Have just added a LOT of code to my game today, but can't test it yet! Need to write another four functions before I'll be able to test it, and then there will be a lot of debugging I expect. Nonetheless I should have a working version by mid-afternoon tomorrow ready to post on here, and I can then add the second half of the game in finally!

I really hope the code I've spent so much time on today doesn't have many bugs..

Next on the list..

2008-02-24T21:12:00.003Z

Tomorrow my target is to get the character moving realistically once the secondary hit test function works correctly. After this I'll be adding the second half of the game in.. which will hopefully also be done tomorrow.

For those of you who happen to be reading this that aren't on my course, watch this space over the next few days as the game is really going to start coming together. As I haven't yet explained to anyone outside the course what my game concept is, hopefully the next few posts will be very interesting to some of you!

Next on the list after this game is completed (well technically it's just a single level rather than a full game) is to create my own website which is long overdue. Once it is complete, this blog will be moved there and I will be creating a secondary blog which will include only highlights from this and hopefully more interesting posts.

Following this is finally getting the 3dBuzz work done so come the new term of University I'll be making some awesome characters. This part may be delayed if the new project is due shortly after Easter, but we shall see!

More, more, more!

2008-02-24T21:02:00.003Z

I've been working on the character hit tests today, now that the new hit test in the previous post works perfectly, I've been adding the secondary hit test to the character. Unfortunately, like all things seem to be with coding this game, adding this functionality negates some of my other code which will henceforth be useless.

This will however enable far more realistic jumping and collision reaction, and furthermore will allow me to move the character up slopes realistically, so the advantages are clear. Just the fact that when he jumps and hits a ceiling he'll start to fall down is nice in itself, and the knowledge that the code I'm writing will be re-usable to the extent that I can EASILY make another platform game afterwards is inspiring to say the least, and in itself, the only reason I am continuing down this difficult road.

This game: Many, many hours of coding to get it working as I want
Next platform game I make: A few days of graphical changes, a bit of game-specific AI coding/minor tweaks and constructing the levels!

hitTest changes

2008-02-23T09:27:00.002Z

Have recently mostly recoded my game to improve the hit-testing - I had planned to do this for a short while as the speed increase provided is around 100%!!

The reason for this is, before I was creating an entire world-size bitmap to perform hit-tests against with the character, meaning it would create a 21x16 tile bitmap in memory - so thats 336 tiles it had to copy before even doing the hit test. It would then have to redo this after the hit-test to refresh the screen in the correct location.

In the new version, it creates a mini-bitmap of the tiles surrounding the characters centre point and does the hit-tests on this instead, so instead of 336 tiles being loaded into a large bitmap, only 9 tiles are loaded into a much smaller bitmap.

This of course will have an impact on not only how fast the screen will redraw each frame, but also the hit tests will use far less calculations since it only needs to test the non-transparent pixels in the 3x3 tile mini bitmap, so speed increases all round!

Not only that, but I also changed how the array is created to a way that makes it around 90% quicker to access after creation (something to do with Flash' optimization code) than how I was using it, so overall now the game should be running MUCH faster!

Here is the game now, with the mini bitmap it is performing the hit-tests against in the top left corner (move the character around and you'll see!)

Actionscript 3: Learning more

2008-02-20T12:44:00.004Z

I'm taking a break from coding the game today in order to learn Actionscript 3 a lot better.

Most importantly I want to be able to split my code up into a few seperate actionscript files so it's easier to see whats what and edit individual parts seperately (hopefully) without messing up anything else.

Other than this, I think its about time I learnt more about Actionscript 3. After all it's hugely different and the many things I've learnt during this project so far are only lightly touching upon the surface of its true power.

I figure that one day spent today will save me many more hours over the coming 2 weeks before the deadline, meaning the game could be even better!

Even if it doesn't save me that much time, it is still a worthwhile investment as when I am creating my portfolio page over the Easter holidays of uni, I will be able to use the improved programming knowledge to aid my (re-)learning of PHP/mySQL which I haven't used for a long time, but fully intend to use to create a back-end for me to update the site easily.

Wasted time but light on the horizon

2008-02-19T12:08:00.006Z

I spent most of yesterday (well, about 14 hours) trying to get my character moving nicely but as it reached the last 30 mins, I realised that I can't tell how the character is hitting things at the moment; I can't differentiate at the moment whether the character is hitting something with his head or his feet.

This causes undesirable behaviour such as if he jumps into another object he can jump higher because he's hit something which resets the jump counter.

I've thought about various ways to solve this and this is what I've come up with (in order of thought):

1. Use various points around the character and if the whole character hits something, perform a hittest on those objects and see which ones are colliding at the same point as the character is.

Advantages: Gives some indication of which part of character is colliding.

Disadvantages: if the character hits a spike or anything in the middle of himself, only the collision on the character will register and this may still cause undesirable effects.

2. Use four half-character hit tests to determine what quadrant he's hit. This is basically the same as above but with bigger hit test areas.

Advantages: Won't miss any spikes and a general area where he's been hit will always be generated.

Disadvantages: Still a lot of inaccuracy in the detection. This may be absolutely fine however.

3. Find the four tiles surrounding the character and put ONLY these into a bitmap. This will hugely improve the speed of the game running as it only has to load 4 tiles each frame for hit testing instead of 300 as it does currently.

Using these four tiles id numbers, identify them as slopes or solid ground etc, and use calculations to figure out where the character should be based on that.

Advantages: Very accurate and covers how the character should be angled upon the slope immediately.

Disadvantages: Will take a lot of coding and may not be possible in the scope of this game.

4. Again use the four tiles around the character, but this time identifying them as slopes or uneven/flat ground. Instead of calculating where the character will have to be when he lands, find out where his hands and feet need to be seperately and change his animation accordingly.

Advantages: Most realistic animation possible. [Excuse the incredibly bad paint image, didn't have much time!]

Disadvantages: Actually animating the character may be difficult. This may likely by alleviated by the fact the game is solely black and white.

I'm thinking at the moment that the most likely choices will be either number four or numbers 2 and 3 combined.

I've also been thinking about using verlet integration for the game, but right now I've got no idea how to understand what it is since American mathematics is written using so many symbols it means nothing to me!

Troubles with movement..

2008-02-18T23:48:00.003Z

Been having a lot of trouble getting the character moving today, but I've just had a moment of potential hope as I've just found an excellent site on hitTests which may very well help me hugely with this stage of the programming.

Here's the tutorial:

Link

The end result is very close to what I want for the character. Obviously since my code is different I'll have to modify it a bit but its a great start! Not sure if it's AS3 but hopefully it is!

He jumps, he scores!

2008-02-18T21:59:00.002Z

Mr Block has moved to a different world, and is now affected by gravity.

It's not all bad for him though, he's learnt how to jump!

Here he is for anyone who wants to see him jumping:

Code speed explanation!

2008-02-18T21:57:00.003Z

The code isn't actually impossibly fast, in fact, it could be almost twice as fast (so I will be changing it shortly!)

It turns out, flash optimizes its arrays if they contain only integers, which my code happened to, but since I was using split to create the arrays, I can actually nearly double the speed the arrays are accessed, as well as speeding the build time by around 20% by changing from using split to either push or fixed length arrays!

So code will soon be even faster!

Mr Block is suddenly stopped on his rampage

2008-02-17T18:39:00.005Z

I've added a basic testing character in now and he moves around - not properly yet, but he moves! I'm about to fix the most obvious problem with it at the moment - when he hits any objects side, he gets stuck in the object.

This is caused by my code creating the moved world after the character is moved, so its just a bit of rearranging needed.

Soon he will rule the world! Muahaha!

My code runs impossibly fast!?!?

2008-02-16T12:36:00.004Z

I recently posted my code so far on the actionscript.org forums in case anyone else attempting to make a tile-based game like this may find it useful.

It seems however after a post today by friendly forumite "sgartner" that my code shouldn't possibly be able to run as fast as it is! Here's the post:

http://www.actionscript.org/forums/showthread.php3?t=160386

So basically, because I happened to make my code with the split() function it seems to have made the arrays faster to access than if I had created them in any other way!

I hope that I've discovered some unbelievable flaw in actionscript that will make peoples code run faster, and that it's not just a (unlikely) coding error in the changed examples by sgartner!

This should turn out to be extremely interesting, and I want to find out if it will have the same effect with single, linear arrays or if it is limited to multi-dimensional arrays as I have used.

Physics? Meh..

2008-02-15T17:33:00.003Z

OK, so I've sadly missed a couple of days where I could have worked as I took a temporary job for 2 days (and valentines day - watching jumper was a must!) but during that time I've been thinking of what will be the easiest way to get the character moving.

The most efficient way I can think of doing it is by implementing a rudimentary physics engine (well just gravity really for now) to do it.

I figure this will also enable me to easily add realistic jumping into it as if there is a force constantly pulling down on the character, all I need to do to jump is oppose that force and put a decay on the opposing force over time. Assuming the decay is calculated properly (I'm thinking exponentially right now, but I'm not too sure really!) then it should automatically give the character a jump that slows at the top and speeds up on the way down, like real life!

So my next task now is to get the character in the game and, hopefully, moving!

Watch this space..

Flash keyPolling

2008-02-13T13:57:00.002Z

Just found this keyPoll class which tests if a key is down or up in flash. My code currently does this but it could be better!

This class seems interesting and I might see if I can modify parts to improve my code..

Here's a link if anyones interested in taking a look:

FlashGameCode.net

They also have a few other classes and even code for a particle system in flash! Definitely worth a look.

Here's a few other sites which other's may find useful:

8BitRocket: This guy has some really good tutorials on actionscript 3 game development including pixel level collision detection (lots of you will need this!) and other things.

Fisix Engine: A physics engine in flash! Looks very interesting!

Papervision 3d: A 3d rendering engine in flash - import your 3d models and animate them! Actionscript 3 has revolutionised flash development..

WOW 3D Physics Engine: A 3D physics engine in flash! Trumps all the others but isn't complete yet - still is going to be amazing when done - think of the games and things you'll soon be able to create in flash! I can't wait!

Hope someone finds these as interesting as I did!

5 minutes of coding later..

2008-02-13T13:27:00.004Z

5 minutes of coding done after the previous post (plus a shower and breakfast!) and I've fixed the most important bugs in the new scrolling screen.

Now it scrolls smoothly and doesn't have any black gaps on the sides, and also doesn't jump when it gets to the final rows/final columns tile, so overall is FAR better from just a couple of minutes changing where a few things were.

There is one small bug left to fix - when you get to the last row/column, it tries to load a nonexistant tile from the next row/column (which is causing it to pause until you get away from that row/column) - just need a quick bit of code to fix that, but I know exactly what to do so this part (should) be simple :)

Edit: Just fixed it and updated the link - there is now no pauses at all when navigating the map - works perfectly at this stage!

Here's the better scrolling version:

Next steps:

1. Update comments/reorganise code so its cleaner and easier to read.

2. Add the character and start getting him moving!

I'm looking forward to this step :D