Not partaken of any daring ludology in ages, and haven’t completed one since my 3-hour abomination in LD12:The Tower (yes, that long ago)

I’ll be looking to do something in either PHP or Node.js for the serverside logic and Javascript/Canvas for the clientside logic, depending upon theme and ideas.

Anyway, true to my word, I’ve uploaded the fail version of it.

http://ed.onlinewar.org/viking-pre-test/www/

To scroll the display , use the numpad keys, but don’t go up or left at first (it can’t handle scrolling off the edge of the map)

Only tested in Firefox – requires HTML5 canvas element.

I’ll finish it!

I’ll still post whatever I’ve done at the deadline, but just don’t expect the “final” tag on that post.

My todo list for today (in order of priority):

- Improve map paging

- Biome parser

- Entity definitions

- Implement efficient scrolling

- AI

- Interface (at the moment there isn’t one, beyond keyboard)

- Create own tileset (I’m using the stock Dwarf Fortress one just now)

- Resizeable canvas/curses UI

- Resource management (food, grog, plunder)

- Crew management

- Weather / Day-Night cycle

- Dynamic vector field for wind direction

- Lose condition

- Town/Battle mode (separate game mode from the main “high seas” game. more traditionally roguelike – may not make the cut, which would be a shame if ship-to-ship combat was just like trad roguelikes but with blue tiles)

- Procedurally generated content (maps, biomes, entities, markov chains for town and people names)

I’ve used a 600×600 version of the heightmap on this post to populate the world data, there is also a JSON file that contains entity definitions and locations and another image map that maps cells to biomes, but those aren’t hooked into it yet, hence why there’s only the four types of tile at present. If I have enough time, the image maps and entity JSON will be fed in by a PHP script that will do the procedural generation.

Next stop, bed.

- Pirates Roguelike

- Tactical turn based variant of carrier command

As the former would be a client-side javascript game, and the latter as a proper web app, there’s a lot of divergence in the development process.

While I vacillate, have the sum total of my island-based efforts so far.

My tools will be wit, planning, and several foxes’ worth of cunning.

My base code will be my as yet-unreleased javascript curses-style renderer.

My game will, more than likely, be some form of roguelike.

Wizard needs sleep!

Well done to everyone who has finished, and good luck to those working up to the deadline.

Ed

http://ed.onlinewar.org/exploratorium/holyshit.png

No sign of a game, or interaction. But on the plus side I now have a Procedurally Generated Galaxy with 6.2 million stars in.

Shame they’re all the same. Bloody stars, can’t trust any of them.

The algorithm which I have used is in fact the same one that was used in Elite and Frontier, and it occasionally spits out a real word, or happily the odd bit of profanity.

I was a little worried when it gave me “quitbed”, though. I think it’s beginning to become sentient.

http://ed.onlinewar.org/exploratorium/foulmouth.php

I’m so sorry.

I shall partake this time round. Lessons learned from previous attempts were crystal clear; I should stick to what I’m good at. Well, I’m a web developer by trade, so instead of futzing about incompetently with triangles, I shall instead endeavour to create a web game, if I can come up with an appropriate idea for the theme

Tools:

PHP, Javascript, Eclipse PDT, Apache, GD Library, An Image Editor which may begin with the letter P

Pre-existing custom-built libraries:

Simplified MVC framework for PHP, with a single entry point for normal requests, Ajax requests, and scheduled cron tasks. I’ll share the code once the Model portion is not an unreadable mess.

Javascript library for handling Ajax requests.

A few custom JS widgets:

- Debugging console

- Wrapper for Canvas functionality

- Various interface bits and bobs

If realtime multiplayer interaction is appropriate, I may incorporate a Javascript/Flash socket bridge as long-polling over Ajax is a painful misery.

Good luck, everyone

I wont be able to start working on it until much later today, thanks to the joys of crunchmode overtime.

My guess is that I’ll do a prototypical sim city dealie in PHP/JS.

First of all, familiarity.

- I will only use techniques and concepts that I am already familiar with, rather than trying to combine learning and the time pressure, which caused me to explode spectacularly last time when Flex and I didn’t get on well at all. So I have decided this time to opt for PHP and Javascript as my game platform. I have no pre-written libraries though, although if pressed for time I might use a couple of PEAR extensions and JQuery.

Second of all, simplicity.

- I will only make a game that can be feasibly completed within 24 hours, and will then use any time left over to hone and improve it. Last time I tried to make something hugely ambitious and failed miserably with very little time left, and was forced down the route of starting from scratch and cranking out an ass-nasty flash “game” with 2 hours left. This is not full of win.

Third of all, prioritisation.

- Sleep when tired, eat when hungry, drink when thirsty. If something interminable is defying my attempts to resolve it, I will go away and do something else while my brain processes it, rather than wasting 7 hours trying to fix it (yes, AS3 XML handling and Flex SDK debugger I’m looking at you.)

Finally, have fun.

- Last time I didn’t enjoy most of my time on LD because I spent so long stuck on the one problem (loading assets from xml files), which is not fun. By going down the route of PHP and JS (Both really fun languages to mess around with, and flexible enough to have workarounds for most problems), this should hopefully stop me getting helplessly stuck at the bottom of problems.

Good luck to everyone!

I decided to go for option 1, the ultra-simple 3d modeller.

The other options didn’t set my “ooh shiny!” reflex ablaze, although they are probably significantly more doable in the time frame.

In the same vein as Ludum Dare I have given it a Latin name – Carpe Lutum (Seize the Clay).

I have put all the woes I had with Flex in LD12 behind me – i had so much trouble because I was trying to treat it like normal Flash, and it will resist you every step of the way if you try that. Use it like an app platform and it purrs like a kitten.

Here is a screenshot of my interface as of hour 5:

Ideas:

- Ultra-ultra simple 3D Modeller for very low-poly models

- Asset Management and loader (Graphics, sound, models)

- Maths libraries (I suck at maths though)

- Multiplayer server? (already written the main guts of this in PHP, just need to get it finished and working)

- Tile/map editor for my RetroRemakes entry.

First I’d like to say well done to everyone who took part and completed a game within the time limit, and thank you to everyone for giving me honest reviews. I had great fun taking part and learned an awful lot about game development during and after the contest.

(post mortem continues after the break)

I have to say that I agree with most of the ratings and reviews given for my entry TowerAssault0. Although it’s not bad for something that took ~2.5 hours in the creation, it was entered into a 48-hour competition and on that basis my entry stinks, and I’m glad everyone judged it on that basis as this gives me something constructive to build upon for next time.

What went right:

- Actually getting something almost playable out the door.

- The graphics turned out quite nicely by my own standards.

- From my aborted earlier attempts, I now have the beginnings of a fairly robust 3d game engine in flash, and I now know my way around Flex/AS3.0 somewhat and should be able to start developing with it properly.

- I’m quite proud of my submission considering how little time I had to spend on it.

- The learning experience. I learned heaps about deadlines and problem solving through this, and will be eager to use the fruits of these lessons in the next competition.

What went wrong:

I don’t want this section to turn into a hand-wringing self-deprecating whine, but a lot of things seemed to go wrong for me in this contest (most caused by myself)

- Lack of preparation. I basically started the competition with no knowledge of flex or AS3.0, and had to learn as I go along, and had not previously developed a game to completion since Amos Pro on the Amiga some 13 years ago. In hindsight I should have done some warm-up projects using the platform I was targetting.

- Overambition. Using a new platform is bad enough, but trying to create something completely unfamiliar (a wireframe 3d first person tower exploration game a la the freescape games on the 8/16 bit computers of yore) while combatting the above problems was completely beyond me.

- The awful awful online Flex documentation.

- I left the decision to abandon Flex/AS3.0 for my competition entry far too late, and then tried to port stuff back to Flash8/AS2.0 and continue on the weaker but more familiar platform, with some success.

- Failing to learn from my earlier mistake, it was very late in the competition before I realised that I would have literally NOTHING to show for my effort. This led to me to a choice between the total failure of not submitting anything at all, and the poor fallback option of rushing a quick-and-nasty flash game out the door. Had I decided to abort earlier than I did and go down the route of quick-and-nasty flash game I might have been able to add more gameplay to my entry and make it better.

- Out of time - I wasn’t able to finish my entry, the plan had been to give the player several attempts to hit the castle before game over, but as I ran out of time I wasn’t able to code all the things required to reset the game back to initial settings keeping the castle at the same distance, so I forced it down the route of a browser refresh. A cheap and nasty way around it. This got the most criticism of all, and deservedly so.

- Spent too long in IRC and not enough time writing code. Enough said

I’m going to rustle up a post-LD version and will post it here in the next couple of days when I get a chance to work on it.