Xerra's Blog

Drunken coder ramblings

General update —

Hi guys,

Sorry for going over a month without an update. I’ve been so busy with the new job that, while I’ve still been working on the game, I’ve not made time to blog about what I’m actually doing.

At present I’ve spent the last couple of weeks going back into the title sequence I started earlier in the year as we decided it was more important to get all that out of the way before getting the transfer game done. Our reason for this is that the transfer game is a self-containted, mostly scripted, room for the project so, as a result, will be easily ported into the main project once it’s completely finished, and tested.

The title sequence was pretty-much the first part of the whole project that I started working on before I dropped it half-done to try and put the transfer game in place so that Aaron wasn’t held up on his droid code. However Aaron rejigged his development list to work on other areas first and leave that fun part as the last part of the game to go in. He’s almost ready to make a start there but we haven’t tried merging projects yet so felt this was something we should get under our belt now both for future reference as to the difficulties, and so we can both modify how we start projects and handle cross-over elements, to make working as a team go even better with the next project.

So I’ve rewritten almost from scratch the old, poorly written title screen/attract mode sequence to pay tribute to what Andy Braybrook did, but give it a modern update. And it’s looking pretty good now, barring some graphics changes.

Some screenshots and a walk-through of what I’ve been doing will be in the next update – hopefully at the weekend.


Who knew walls could be so tricky —

The Paradroid transfer game is actually turning out to be a bit of a bugger to get in place. After a couple of rewrites I’m actually going back to basics and putting the code together on paper to try and get a proper build routine for random maps. At present everything is switched off apart from the generation of walls on one side of the playfield circuits so they can be tested. And retested. And then retested again.

I’ve been breaking it down into stages with this, as it’s so important to take into account other elements around an area before a decision is made to plonk a wall down there. Actually getting it on screen isn’t the problem but it needs to follow the rules so that circuits can be activated through it, and that multiple walls don’t screw each other bit.

Start on the left side of the play area and start reading the first line horizontally deciding randomly wether a wall should be put down or not.

If a wall is to go down then I need to follow this decision up by deciding if this wall will have two circuits leading into it, and one leading out. Or should I have one line leading into it and two lines coming out.

I will then need to check the surrounding blocks to ensure I’m not going to break another walls 3 blocks by creating it. I say will then as I’m not at that part yet.

Decide to build my wall and place the blocks.

Depending on which way round the circuits go – one wire into two, or two into one, I then need to clear either the middle line in front of the wall up until the end of the play section then plonk a closing zapper to tidy it up, or clear the upper two lines behind the wall as it should only have the one wire going in.

I will then need to decide if I have another wall to maybe close off the two sections and then put it into a single wire again, perhaps, just like the original Pardoid did. That will naturally mean rebuilding the middle line again from the wall and this time clearing the lines above and below this wire from the front of the wall.

This is all for just one of the 11 lines going horizontal, so I repeat for the other 10 and will need to make sure that when i put down another wall they aren’t too close to one already down so the circuit would be broken.

After all that it’s going to be running through the board again, looking at each location and deciding wether to put any blockers, zappers or switchers on, and ensuring they are in legal postions.

That all make sense so far? Probably not. It just amazes me how Andrew Braybrook did it back in 1985 – and in 2k of memory using 6510 assembler. I’ve been at this for over a couple of months now – allowing for a real life couple of weeks break I needed to take for work reasons. Yes, I have a day job, too.

The reworking of the walls will fall into place soon, which looks like it’s going to be the hardest part, so I’m hoping the rest will come together much quicker after that. It’s a good job Aaron doesn’t have those droids in the game yet as he’ll be needing the transfer game slotted into his project by then.

I’m actually almost looking forward to going back and completing all the title screen stuff once this is over. Next time I’m volunteering to do the maps or something 🙂

Here’s a snippet of the terrible code that’s currently in there for generating the walls. I have a build script that can be called with multiple arguments for doing stuff like clearing the play areas and filling with wires, throwing random chars in the area to check I’m not overwriting the background display and testing animating pylons etc. That’s why it’s such a lot of lines at present.

As with other code I write, I often tend to hack it together to get it all working and then will go through it again tidying up the messy parts and optimising to an extent. I always comment the hell out of something when I first put it together in case I don’t finish that part in that session. I will then often tidy up the comments later just in case someone else does read the code, so they don’t fall asleep with too much information.

The code shown above will likely look nothing like this once it’s actually finished.

As it stands, however, I can just press a key when running to generate a different map which is how I’m testing if my walls work. As you can see from the screen shot below – they don’t. Ignore the vertical/horizontal text – that’s a hangover from when I put in the control pad code.

Work in progress, people…


Website update —

Thought I’d put myself out there a bit more with some social media buttons and add a lot of Twitter integration as it seems a lot more people are starting to get interested in our #ParadroidRemake blogs and tweets. While I was at it I also added in dedicated pages to some old games I’ve been involved with including a link to iDef, now it’s not currently mentioned on our website due to the Paradroid build up page.

Our developer licence for Apple is probably due to expire soon so may as well see if anyone else wants to download it from the AppStore for free before they take it offline – as they do for expired developer contracts.

On the right is also a link to another new page which is where I have been working on my kind of memoirs, for want of a better description, of my experiences being a teenager as the home computer revolution came in. I’m not sure where it will go if I ever finish it, but I put up a draft (read rough) copy of the first couple of chapters in case anyone was interested in reading it.

Progress on the transfer game for Paradroid is still ongoing. It’s pretty tricky trying to get random layouts that are always workable in a game.

Brief update —

Sorry, I’ve been a bit quiet of late. Real life stuff slowing down my coding a bit but I’m still beavering away. Aaron’s been a lot busier so today I’ve been testing out all the droid info screens that he’s been plugging into the console. Now that these are in and working – just a few cosmetic changes to be made – the actual console part is almost finished. Aaron has the icon for statistics plugged in as well, which is something that the original game never actually had but got added by another coder years down the line, along with other improvements, when a lot more was known about the C64 hardware. Not to mention compression improvements allowing a lot more data to be held in memory. Quite how this will be setup is yet to be decided but it will probably get done once there’s proper gameplay running near the end, so that there’s stuff to actually track. He’ll no doubt blog about how he’s done the console screens once he’s written up all the fun he had doing the lifts.

The links on the right hand side of this page if you want to catch up with what he’s been up to. Certainly more than me this month /guilty-look


Gamepads —

Aaron and I had a very important business meeting this weekend at our head office (cough) The Coach and Horses. Among the things we discussed were throwing ideas about on a potential next project after Paradroid, and debating wether we need to have gamepad support in Paradroid. As I’ve mostly just used keys/mouse on the games I play on the Mac, and just stick to playing most big games on the PS4, I hadn’t given it much thought until now.

The thinking is that it’s a C64 game we’re remaking that was designed to be played on an old cube-telly and a Competition Pro or Quickshot joystick, so it’s common sense really. Bonus points to the old farts who can remember both those joysticks. I had a Competition pro that lasted me through my Vic 20 and C64 years. Never used anything close to being as good, although the PS4 controllers I’m almost comfortable with now.

So I spent Sunday afternoon digging out my old wireless XBox 360 game pad as it turns out GMS2 does support either those or PS3 controllers. No PS4, otherwise I’d have just attempted it with that.

Another snag was that Microsoft (obviously) don’t provide drivers for using their XBox controllers on an Apple Mac. I mean, why would they?

Fortunately there’s a driver out there for people who’d like to use one anyway, and it was relatively painless to get the device linked up to my Mac and then start looking at the code side of things.

In the old days of programming a game to work with a joystick, it was pretty much just peeking a location to see if the number within had changed, which would indicate the switch was joined and the game should react accordingly. Nowdays there’s start buttons, back buttons, triggers, shoulder buttons, directional sticks and four flippin’ fire buttons. Only basic joystick use is really required for this game so I set up the transfer game in its current state with enough code to be readable and for Aaron to make use of the source to do the same with the main game.

Once the games finished I’ll probably take out all the joystick code and stick it into some kind of reusable script so gamepad support can be dropped in to anything else we write that follows Paradroid.

It was nice getting away from the workings of the board building routine that I’m still tinkering with , while still working on the game itself as that way I don’t feel that I’m wasting time when something needs a bit more thought. I just do another part of the game instead that won’t need much thinking, such as tinkering with sounds.

No new screenshots this time as there isn’t much new to see as yet. Nip over to Aarons blog via the link on the right and you can see an example of the in-game graphics being switched to heavy metal, original etc.


Transfer game progress update —

More work on the transfer game this afternoon.

I’ve been trying to keep the game handler down to fewer objects so it’s easier to work with all the elements without having to factor in with object calls as it could get kind of messy with all the sub-elements that are going to come into effect in this mini-game. I’ll rely more on scripting for the grunt code stuff such as drawing the game board as it’s easier for Aaron to read my work and understand what’s going on. There will probably be an element of tidying up and improving areas after the grunt work is done because that’s part of the optimisation process, although it’s probably not going to produce any noticeable improvement in the game speed as it’s not exactly hitting the hardware as it is. But one day maybe someone else may read the code so it’s nice to be seen as doing your job properly.

I got the mechanisms in place to have a toggle-able game state in so I could both implement the movement of the player selector for both sides, and also test out some random mapping. I don’t have it in place where the script actually does draw a proper playable game board yet, but it’s got test code in to throw out random tiles so I can check everything is being drawn in place and that my movement works correctly within the tiles on both sides. It’s a little strange working with a tile as the actual player controlled element rather than a sprite, but sometimes it’s cool to mix things up a bit as everything is a learning thing with a new development system.

So, having tidied up various elements of the other parts of the game controller for the transfer system, I’m unable to do anything else until I actually have game boards to work on so I can put in the code to actually start activating individual circuits.

It will be straight in at the deep end for my next blog entry.

As a side note, Aaron and myself have found a project management system we both finally liked, rather than hacking together some kind of shared spreadsheet. We created our project on EasyNote. This is a deliberately simple system of setting tasks and assigning them between people with the usual options to comment, set sub tasks, deadline dates and status of elements. Dead simple to use and invite members into each one so we’ve got all our development notes running from this now.

It appears to be free and has a gig of storage for what looks like both attached files and actual projects, so we should have plenty more space to use it for future projects after Paradroid. As I said, it’s free so this is no advertising plug. I just thought it’s worth a mention as we’re finding it so handy.

Have another screenshot. This one shows the board handler spewing out random tiles just to check the layout is ok. You’ll get one with an actual random generated map one day, I promise !!

Transfer game progress —

Creating the transfer game was always going to be a major task in Paradroid due to the complexities of having a random element to every game screen generated but having to make every one actually playable. I knew at the start that the actual routine to generate the screen was going to need to be written one element at a time to ensure each one worked on their own, before adding in the block stuff – which is where 3 lines are used to create an opening bar and the circuits from this can either progress into other bars, or close off into a single circuit again. This coupled with the other elements that can be part of this structure, such as colour swappers, reverse polarity switches or just plain dead ends, will make this require a lot of testing and map generating within the code to get it absolutely right.

So, progress on the creation, means building the other elements up first, such as the droids on screen with their respective class numbers, and sides selection implemented, so we can get the right amount of turns given accordingly.

Hence the screenshots shown here.

And another shot here shows us having the 999 droid under our influence.

Quite why we would want to drop down into a lowly 123 maintenance droid from this unholy beast of destruction is another story that I’m likely never to go into.

So our droid of choice is now selectable within the transfer game by us. You guys won’t be so lucky as I’ll disable the option to do this once it’s all working so you’ll need to find another way to cheat like me. If we let you …

We went to great pains to ensure that the layout is as close to Paradroid as we can get it, while obviously adjusting graphics size to fit on our predetermined screen size of choice – mainly for compatibility reasons as we would like the end-game to run on practically any old PC out there. DirectX restrictions allowing, of course.

So the markers can already be seen for each players movable circuit selector in black on each side. Once they move down from their parking spot they can’t go in there again and naturally can only fire into circuits that are available. If there’s a blocker in the next column to them then the original game would not waste a shot in such a pointless attempt to try and activate it so we won’t either, for example.

The circuits area which will be modified by the random map element are currently highlighted with yellow and purple animated elements to show conductivity across the line as they will show in the game once the circuit is lit up by the player or computer. You can’t see them animating in a screenshot, however, so you’re just going to have to take my word on this for now.

Creating the actual map is being done by using tilemaps rather than sprites, so a lot of my time has been spent working out how to actually put these on screen and working correctly. Fortunately I’ve had the assistance of Aaron in doing this, as he’s already been playing with that feature of GMS2 from the start in creating Paradroids many deck/map arrangements.

Where I’m currently at is putting in the code to control movement by the player up and down the line of circuits which shouldn’t take very long so I can get to the fun part of tying in the drawing of a random map every hit of the space bar.

The transfer game links in a timer to select a side to play and for completion of all your shots on the circuit board itself, but I’ve disabled all  this for now so that the testing can be done first as there will be a lot of map studying taking place to ensure there are no flawed circuits in any of the play areas.

It all sounds so easy when it’s written into a blog but you can be sure that the code element for this part is probably going to take me longer than all the work I’ve put into the title sequence so far.

Onward then…


Paradroid music —

While Aaron and I were looking into wether we could actually do our own clone of Paradroid, I also went looking for some alternative music as I didn’t fancy the idea of using the droid chatter system of the original game.

Luckily I found that a chap called P I N (Michael Pin) had done a techno mix using a background sound and some of the sound effects from the actual game. You can listen to it now on his soundcloud page. The track is called “Do you remember Paradroid”

Michael has agreed to let us use the music and could also compose something new for us if we need another track later on. Naturally, we are really happy about that as, due to the game Paradroid not belonging to us, we never planned on ever trying to sell it, so we were never going to be able to pay for anything we used that was created by anyone other than ourselves.

Aaron had been tinkering around recreating some of the Paradroid sounds in code downtime but we also have a large sound file that Michael supplied which he used for his music project that we may also be able to make use of. Hopefully this will make the sound effects part of the project a lot quicker to get done.

Here’s also the original YouTube video Michael did when he created the music.

Do you remember Paradroid?


The transfer game —

Even though the title sequence still needs some work to get it up to standard, we decided that it was something to come back to later as there will soon be enemy droids going into the game. Obviously, without having the transfer game in place, it’s going to be just the shoot-em-up aspect of Paradroid we’ll be testing and fixing, so it’s time to get that done now so all aspects of enemy droids can be set up in one go.

We delegated a rough task structure at the start of the project and decided that I would be the one who would be coding the transfer mini-game. With us this doesn’t mean that I just get on with it and Aaron will look at it when it’s done. With all aspects of the game we spend a lot of discussion time making sure we have as much assistance as needed to get each section done, so the last weekend we both spent a considerable amount of time at Aaron’s working on the graphics for the game and working out how it’s going to be coded. As in either Sprites or a tile map system.

Common sense made us choose tile maps as we have the option of resizing to get everything fitting nicely on screen, and it will be much easier to actually keep track of them in game as it’s a map system.

So here’s one of several maps we’ve set up for both testing the graphics, getting the tile-sizes right, and being a useful reference for how to put the random element into code when generating a new screen.

This one functions as how the screen will be generated at the start. My code will overwrite the blank areas with the actual layout of the circuits, blockers and switchers. Not sure if those are actual names Andrew used but I need variable names for them so they will do for me.

Andrew stated in his original “Zzap 64” diary that it was pretty easy to get this transfer game working properly and did it over a weekend, as I recall. I have as much time as needed, realistically, and a much better development system. Not to mention I don’t have to worry about fitting it into 2k of memory either. And I doubt I’ll still get it done anywhere near as quick as he did but here’s hoping.

This is pretty much how it looked in my editor before I started working on creating it. I’ve not got Aaron’s handy C64 colour macros in the project yet and stuff like all the other graphics needed, like robot definitions in, but it gives an idea of how I’m working with GMS on this horrible Parallels environment. Can’t believe I was happy working with that when I first tried it…

The tiles we needed to make up the entire game turned out to be relatively low once we had put some through in to what could be re-used. We didn’t screenshot the original game and use GMS to tile a PNG image because we wanted to keep it tight so we could macro the tile map blocks and keep it simple to keep track of redrawing the map after every circuit is activated. Creating our own tile set also gave us the advantage of getting the dimensions right as we wanted to make it fit exact to our current fixed screen size but retain the original number of usable elements in the game. We had to play the original game many, many times and tinker with the graphics a lot to get this just right. Hopefully we did cover every tile actually needed but I’m not going to be totally sure on that until I actually start moving things around on-screen.


Bug fixing and text formatting —

Started ripping out the temporary text lines I’ve been using for testing the scrolling information. Had all sorts of fun and games creating a routine to load in a text file and then display the text using a DS_list. Because I’d been editing it on the virtual PC  it had inherited all sorts of stupid formatting codes that weren’t visible and caused the formatting to go all over the place. Eventually got that all sorted out and finally had reasonably formatted instructions doing what I wanted.

Got her indoors to help out choosing some colour combinations to swap around for the displays as the idea is to break up the routine by displaying the original Paradroid loading screen that Andy had on the C64 between runs. Due to the way the surface scrolling works, and how i’ve clumsily used two objects, I’ve got to streamline it all again back to just one system doing all the background stuff as things are going all over the place when I try to add little features here and there. Put that down to a job for next session along with some advice from Aaron about using two views to save constantly creating and removing objects while running the title sequence.

I’ve mentioned previously about using a file loading system to bring in the text so it’s easy to modify it on the fly and eventually turn it into internal data once it was finalised. I was hoping to just stick with the load system as it works pretty well, but it turns out that external files added to the project aren’t built into the main executable at compile time, so it would be easy for anyone to modify the text after installing the game to do all kinds of naughty stuff like changing credits, or potentially breaking the game with overloaded text lines. So I’m going to have to shunt the formatted text into data within the code, although it’s not something I plan to do right until the end.

The only external data for the game is going to be probably ini files for saving and loading configuration stuff, and maybe a save game system, if we decide to build that in.

Got another build of Aaron’s part of the game today. As well as already having all the maps set up, he’s now also got the lift UI and code in, so it’s possible to test all the different decks of the ship. There’s a couple of position faults, and wrong lift shafts highlighted issues still to get fixed, but it’s all working pretty well.

I talked him into doing a test run of double speed movement for your droid if it’s moving around on an empty deck. It can be quite tedious wandering around trying to find those last couple of droids when you’ve almost cleared the ship. Wether it stays like this, becomes a feature you can activate, or a hidden thing has to be decided as I like it but he doesn’t.

Time to email and see if i can get someones permission to use a tune of theirs for the game music now. Hopefully they will say yes and also help out with some of the sound effects as it seems they’ve managed to emulate loads of them within the actual music itself.