So I attempted to make a game for LD25, unfortunately I am ashamed to admit I gave up in frustration after about 8 hours. There were a number of reasons:

The first issue, tried to make a sprite based game using Unity3D… now the clue is in the name, whilst Unity can be used to make 2D games it’s not its primary focus, and when your only target is web deployment there are probably better choices… Unity is great at making complex things simple (Physics, 3D Animation, Terrain Generation, Sound integration) but it’s also pretty damn good at making simple things complex (just drawing a freaking picture on to the screen).

So this time if I use Unity3D I’m going to do something in 3D and make use the out-of-the-box cool stuff Unity has (like physics!) however if I decide to try another sprite based game I’m going to use HTML5 canvas… you can’t get much more straight forward than context.drawImage(…).

The second issue, being based in the UK, I decided starting at ~2am probably not the best approach if I still wanted to be functioning by Sunday afternoon, so I got up at a reasonable time only to be horribly disheartened by all the people posting their already working prototypes after just 5-6 hours.

So this time I’m going media blackout, no checking the Ludum Dare posts – plenty of time to review those later – no twitter, no email. Also considering creating a sign “please don’t disturb unless you’re bringing sustenance”. This does mean I’m not going to post progress reports, but I’ll try to keep a log and post it all in one massive lump at the end along with the hopefully a finished game. I might relax this plan if I need some play-testers on Sunday.

The third issue, last time I did no preparation, I’d never tried to draw 2D sprites in Unity before, I’d not used any of the sprite manager plugins, I’d not practised creating my own sounds or music in years. Learning /remembering this stuff even enough to make place holder sounds takes a very long time!

So this time, I’ve drawn a few sprites with HTML5 already, I’ve worked out what libraries I’m going to use, I’ve got the wonderful SFXR to make some sound effects and tonight I’m going to be messing about with /learning the basics of LMMS in the hope I’ll be able to make some music with it some time over the weekend.

The fourth issue, a common one I think, I was overambitious. After 8 hours I just about had a sprite – which only I could recognise as a bear – you could move around facing in the right direction but not actually animating in any true sense of the word; some sprites that looked something like people (but again only to me it would seem) running about on semi-random paths (avoiding walls and such-like though) and the ability to make them disappear if you pressed space when facing them and were nearby… not really far enough along for 12+ hours in (5 hour delay before I started). It was at this point mostly frustrated by my complete lack of artistic skills that I went “Sod it, I’m going to go play a game rather than make one”.

So this time, whilst I’m also planning on using more appropriate tools, I’m still going to assume everything will take twice as long as I think it’ll take so I’ll try to think of a game to the theme I can make in half a day, that way the prototype might be finished on the first day and I can spend the second day polishing it. So it’s going to be something very simple, but much better to finish eh?

The fifth issue, I can’t drink caffeine (if I do I get heart palpitations) so I have to get a reasonably good night’s sleep, I can’t really crunch the way many people seem to live by.

So this time, well this one isn’t going to go away, but I’m planning on in-taking plenty of sugar and drinking lots of tea instead, and I can probably get away with a single cup of coffee or can of coke per day, so I’ll save that for the late evening in the hope it’ll help me continue to be productive into the night, I’ll just have to make do with maybe 32 hours of productive time (I’ve got to at least be able to get up for work on Monday too).

The final issue, I keep making excuses.

So this time I’m going to stop whining and keep working until the deadline, if it’s finished “yay!” if not, at least I didn’t give up this time.

Good luck everyone!

So I managed to finish a game this time! It’s available at http://goo.gl/4qI3O  it’s HTML5 and requires a keyboard. Works completely in Chrome, in other browsers some or all of the sounds fail to play. If it had a tablet or other device to test with I probably could add touch controls (I think it might work quite well with the game actually), but I don’t fancy trying to do it blind. So no touch controls, sorry!

On seeing the topic my first thought was, “oh what the f***”. My second thought was “Oh well this is a good excuse not use any art or music”. My third thought was “Actually might be a fun design challenge”, minimalism suggests to me, no UI, evident gameplay, no scores, no complex abilities.

I’m reasonably happy with it and I enjoyed working in JavaScript again (although any attempt proper OO broke down after a few hours), I think the most important change from last time I attempted it was the media blackout, mostly because other people are more skilled at this that me… especially in the art and music departments, best to ignore that whilst you’re working.

I think next time I try to make a HTML5 game I’m going to use WebGL even if it’s 2D because I would have liked to do some post-processing effects to pretty it up. Also when playing with higher top speeds for the circles it turns out even a reasonably low speed has a lot juddering.

This time I used howler.js for sound because it provides fall back for non-web audio api browsers… but given that the sounds only kind of work on those browers, I think I’ll just use the web audio API myself in future.

My final thought is, I had to make myself do a steering behaviours project a couple of months ago and I kind of thought to myself at the time, “this is a waste of time”… however this game’s AI is pretty much exclusively steering behaviours. So that’s good encouragement to continue to do ‘learning’ projects as well as actual game projects.

So, yay finished a Ludum Dare, it was fun. If you try my game I hope you enjoy it.

Yay. So this’ll be my third Ludum Dare, and with luck my second successful Ludum Dare.

I shall be using either WebGL or Unity3D, honestly depending on how I feel when I start! (as well as: bxfr / paint.net / (sublime text 2 / node.js) || (Visual Studio Community))

I’ll be aiming for a 2D game, with as inspired by the key note something a bit different, which for me is more of a focus on art / visuals than I normally take. Soo that’ll be some focus on art / visuals rather than just geometric shapes.

If I use WebGL I’ll be using my engine/framework which has been up on github for quite a while, if not given much love recently https://github.com/delphic/Fury. It’s not much more than an organised bunch of helper functions and state rebinding minimisation at the moment, still this should be fine for just rendering unlit sprites!

Being in the UK and fond of sleep I’ll not be starting until several hours in when I get up at a relatively normal time, so I’ll likely work under a blog / twitter blackout so I don’t get demoralised by much further ahead all those who start on time are (and / or just how much better people’s ideas tend to be).

Good luck and have fun everyone!

So the phrase post-mortem is quite accurate, in that I did not finish nor submit my game. The experience probably best summed up by my tweet on Saturday evening.

Nope nope nope, too much time on tooling / workflow not enough on making the game. Not enjoying… I’m out. #LDJAM

— Harry Jones (@_delph) April 18, 2015

Here’s a gif of what I did manage (distressingly probably 5/6? hours work)

The important thing when you fail, is to learn something from that failure, I think the lessons for games jams to be taken from this experience for myself are as follows:

1. Use tools you are familiar with
2. If you’re trying something new (e.g. pixel art animation) give it a try before the jam too – doing a warm up over the week before is a good idea! – and make sure you’ve got your workflow / import process sorted before hand – that’s a getting to proof of concept and repeating a couple of times at least!
3. Seriously… use tools you are familiar with
4. If you set yourself rules about media blackout to keep focus and motivation, keep to these rules!
5. Did I mention? “Use tools you are familiar with”
6. Before taking part in a jam make sure you’re comfortable with the time pressures involved, e.g. if Ludum Dare is a bit too much might #1GAM be more your style?
7. Oh and… you should use tools you are familiar with
8. Provide yourself a workspace that encourages you to work on your game and does not distract you
9. Really, just use the tools you are familiar with
10. Short deadline game jams – especially when you’ve only managed a few more before and struggled with making something you were proud of – are not the best place to learn new skills, they’re somewhere to expand / experiment and hone existing skills

The last one might be more about my personality type. I find have unrealistic expectations about what I can achieve when first learning something; it’s not bad to know what you want to be able achieve, but until you try something it’s silly to try to guess how long it’s likely to take to get to that quality.

I enjoyed moments of  the jam, and it was nice to code in JavaScript for a while (not to mention the difference in load time between Unity WebGL export and doing it yourself pointed out the gains to be had from ‘doing it properly’), and got some insight / reminded myself as to why code can descend into entwined state / flag based hell if you don’t think you have time to give the architecture proper thought (false economy coding!), which as echoed in point 10 suggests that perhaps for me slightly longer deadlines / time periods are more suited to giving good motivation!

It's been a while since I tried the compo but going to have a go this time!

Will be using one of my JavaScript libraries:

• Hestia https://github.com/delphic/Hestia or
• Fury https://github.com/delphic/Fury

plus aseprite (https://www.aseprite.org/) for images.

If I get around to attempting sound and music I suspect I'll do something programatically directly with the Web Audio API.

Will be developing on a Ras Pi 4 as much as is possible (i.e. when not using aseprite), so also will probably be using my own text editor (https://delphic.me.uk/blog/hekatetexteditor), because why not!

I'll be toying around with the Web Audio API this evening hoping to know it a little better before the jam starts.

Good Luck and Have Fun!

I've gone from this

to this

Okay, so yeah obviously today was mostly code, but there's a mini life sim going on in the background! And an actual full game loop (although no main menu screen).

I did create that new pixel font from scratch because the micro font that was ready in my JS library was a bit too small / hard to read, and I had to write a util to parse fonts from images because for some reason before now I thought it was a good idea to transcribe the pixel info into a JSON array by hand... yeah I probably should have done that before the jam. Well the joys of using your own libraries, jam's certainly show what you should have done already.

The in progress code is on github and I'm committing as I go https://github.com/delphic/LD46-Keep-It-Alive (yes the game code is all in one huge JS file :)).

Tomorrow will be about art, content and polish. Whilst I'm no where near the 10% mechanics, 90% polish advice you often hear, 50% mechanics, 50% polish is much better than I normally do.

Good night everyone!