| |
Slightly cold news on page 10! Phlamey breaks the server. |
| |
This is a long thread. Click here to view the threaded list. |
| |
|
Jeffrey Lee |
Message #97575, posted by Phlamethrower at 18:41, 23/1/2007, in reply to message #97511 |
 Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
After much poking, thinking, poking, then more thinking, I've realised why the new plotter doesn't produce the same results as the old one. And to fix it will involve implementing a buffer to hold division results, which will be read from each pixel - so will double the number of memory reads needed by the core of the plotter. The buffer will only be small, however, so the speed impact shouldn't be so great.
However, there will also be an impact from the code that generates the buffer, which will essentially be a stripped-down version of the old plotter.
[edit]
And now it's working. Woo!
As expected, the extra divisions and memory accesses have slown down the plotter. It's now twice as slow as the horizontal plotter (But still 2.5 times faster than the old code it's replacing). There are still some optimisations left to do, but I think they'll be improving RiscPC performance more than anything else.
[edit #2]
Added the RiscPC optimisations. The Iyonix speed boost was fairly minimal (It nows runs at just under twice the speed of the horizontal plotter), but the RiscPC boost was much more significant. It runs 3 times faster than the old plotter - the game was running at the target 25fps  (But at 480x352, instead of my initial target resolution 512x384)
Of course there are "still more optimisations to do" (including writing the other half of the plotter, which will result in a further framerate increase), but for the moment I think the framerate is acceptable (50fps with R/B swapping and DMA on an Iyonix; 70fps with DMA; and 57fps with writing straight to PCI RAM, all at 480x352. Providing you're sat at the right place, I might add  ).
Once I finish tidying up the plotter, I'll be having a go at the city generator...
[Edited by Phlamethrower at 23:25, 23/1/2007] |
| |
 [ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #97628, posted by Phlamethrower at 23:57, 24/1/2007, in reply to message #97575 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
Aemulor appears to be fixing my bugs for me 
I found a 100% reproducable crash when entering one of the cars. The crash results in an immediate exit with a WIMP error box, which is of course hidden from view by the double buffering - so not only doesn't the game debug log show it, but I can't see it either. I have written software in the past to solve this problem, but it isn't 32bit compatible yet.
So I figured I'd try running the module using the Aemulor demo. Except that when running GrabError (or my ScreenGrab) module under Aemulor, the game doesn't crash! 
So now I have a reason to make some of my modules 32bit compatible
Also, the other half of the plotter has now been written (and is potentially what's causing the crash). All I need to do is replace the last couple of instances where the old plotter is used and everything should be nice and smooth |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #97630, posted by Phlamethrower at 09:02, 25/1/2007, in reply to message #97628 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
Hmm, why didn't I find that before?  |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98058, posted by Phlamethrower at 13:55, 3/2/2007, in reply to message #97628 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
I found a 100% reproducable crash when entering one of the cars. I forgot about this, and now I can't recreate it! (Well, apart from one instance where it completely locked up the machine) |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98082, posted by Phlamethrower at 13:30, 4/2/2007, in reply to message #98058 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
First attempt at a random city generator. Originally I was thinking of using the same algorithm as GAIO, but then I realised that it would almost always result in one long road stretching right across the city. So instead, I tried writing something to place randomly sized rectangles next to each other, without any overlap.
After a couple of tweaks, the algorithm is kind of working how I want it to. The only downside is that the vertical roads tend to be longer than the horizontal ones; I'll have to see how this affects things when I start adding buildings and proper roads. A more intelligent rectangle sizing function might be able to mitigate some of the problems. |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98085, posted by Phlamethrower at 13:48, 4/2/2007, in reply to message #98082 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
I thought I'd try tweaking it a bit...
Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger!
Internal error: abort on data transfer at &0002AFA0 [Edited by Phlamethrower at 13:49, 4/2/2007] |
| |
[ Log in to reply ] |
| |
|
Andrew |
Message #98087, posted by andrew at 14:04, 4/2/2007, in reply to message #98085 |
 Handbag Boi
Posts: 3439
|
That seems pretty good to be honest. I tried to generate a cavern system that was detailed but not *too* complex once and I couldn't do it without using SIN/TAN (polar?) functions. They gave a really nice cave layout but because I had no idea how to calculate SIN/COS/TAN on the fly (as I understand BASIC does) I couldn't try to implement it in assembler and instead used a lookup table.
Possibly I sent it to you once Phlamethrower but for some reason it didn't work on your machine although I did modfiy afterwards.
I got up to the point of particle plotting also using lookups with the aim of the player firing a particle weapon.
Then I had to find a job
[Edited by andrew at 14:05, 4/2/2007] |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98089, posted by Phlamethrower at 14:49, 4/2/2007, in reply to message #98087 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
I remember you writing that, yes. It's a shame nothing more came of it.
I've managed to tweak my algorithm more. I'd say the otuput is now pretty damn perfect. All I needed was two extra tweaks - previously it was only checking if the current rectangle is larger than the minimum size, but now it also tries to make sure it leaves enough room for the next rectangle to be larger than the minimum size. This got rid of all the nasty small areas (which would sometimes build up on the right of the screen as a big grey mass).
The second tweak was to make it so that a certain proportion of the rectangles stop at the same height as the rectangle to their left; this helps promote longer horizontal roads.
If you get the proportion right, then you can get a natural looking city like the one below
The output does look similar to what the GAIO room splitting algorithm would produce, but it is fundamentally different because there aren't any roads which stretch all the way across the map. |
| |
[ Log in to reply ] |
| |
|
VinceH |
Message #98096, posted by VincceH at 16:48, 4/2/2007, in reply to message #98085 |
 Lowering the tone since the dawn of time
Posts: 1600
|
I thought I'd try tweaking it a bit...
Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger! Bugger!
Internal error: abort on data transfer at &0002AFA0 You didn't copy and paste that error message, did you?
The number of times "Bugger!" appears is the give away - I'm pretty sure no RISC OS error includes that many. |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98097, posted by Phlamethrower at 16:58, 4/2/2007, in reply to message #98096 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
That was from the debug output. I wasn't entirely sure if some of the assumptions I was making in the algorithm were holding true, so I just made it print 'Bugger!' each time it detected a failure. And obviously that particular tweak I made caused it to fail |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98110, posted by Phlamethrower at 01:05, 5/2/2007, in reply to message #98097 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
I now have a road weighting algorithm working (which is used to calculate how wide each road would be). The basic algorithm is the one described here, i.e. make 10,000 trips between random locations and then make the most popular roads the widest. It only takes 15 seconds for a RiscPC to do the calculations, so time isn't an issue. But until I write the proper code for plotting the roads into the map I won't know if the algorithm produces sensible results or not Chances are it will require some tweaking, and I've already got a few ideas on how that could be done. |
| |
[ Log in to reply ] |
| |
|
Tony Haines |
Message #98146, posted by Loris at 19:15, 5/2/2007, in reply to message #98089 |
 Ha ha, me mine, mwahahahaha
Posts: 1025
|
- previously it was only checking if the current rectangle is larger than the minimum size, but now it also tries to make sure it leaves enough room for the next rectangle to be larger than the minimum size. Actually I meant to ask you about that in your random map generator article. IIRC you said that if it randomly chose a cut point within a room. If this attempt to split a room failed, it marked it as unsplittable. This might suggest that occasionally you might attempt to split a very large room (say, the first, large room) too close to the edge and fail, leaving nothing but empty space. |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98148, posted by Phlamethrower at 19:36, 5/2/2007, in reply to message #98146 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
What do you mean by 'empty space' - an actual empty area of the map that can't be reached, or just a big empty room? The algorithm will leave the latter.
Looking at the code, the algorithm works as follows:
1. If the room is wider than it is tall, it will attempt to split it horizontally. Else, split vertically.
2. Select the split point. This split point does have a chance of making a room that's too small. The chance of making a 'too small' room is inversely proportional to the size of the room (Simply because it picks a random point inside the room to place the split, and there's more chance of it picking a point next to an existing wall the smaller the room becomes).
3. If the split is going to make a room that's too small, and the room is above a certain size (I've chosen 16 tiles wide/tall), then pick another split location.
4. If the second split attempt fails, give up.
In their own way, all of these steps help the algorithm to avoid leaving large, empty rooms in the map. |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98153, posted by Phlamethrower at 23:23, 5/2/2007, in reply to message #98148 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
Playing with this road weighting algorithm is fun But it could take a while to get an algorithm that produces sensible results the majority of the time
The image below shows the result of the (slightly tweaked) algorithm. The brighter the colour, the more popular the road; the top 10% roads are in green, the next 10% in blue, and the rest in red. The red is a bit dark, though
Initially the popular roads were all in the middle of the city, and tended to zigzag a lot. Which is basically what I was expecting. So then I tweaked the algorithm to place a penalty on each route that involves turning a corner, and a further penalty for travelling through the middle of the map. As you'd expect, this has promoted straighter roads, and some more popular roads at the edge of the city.
My next step will be to try and make it prefer already popular roads - hopefully this will stop two popular roads from forming right next to each other (and may make the two earlier tweaks redundant).
The only other problem with the road weighting algorithm at the moment is that not all the roads are fully connected. Ideally I'd want all the green roads to be connected to each other, and all the blue roads to be connected to the green road. If I'm lucky, I'll be able to find a tweak to the algorithm - if not, I can add some code that tweaks the output later, by artificially raising or lowering the popularity of certain roads. |
| |
[ Log in to reply ] |
| |
|
richard cheng |
Message #98168, posted by richcheng at 11:31, 6/2/2007, in reply to message #98153 |
Posts: 655
|
That's quite a big ass city. Is your plan to generate it once per game, or once per mission? Or somewhere in between?
GTA has big cities, I'm aware, but they're basically static throughout the game, (In vice city at least. I presume it's the same for the earlier, top-down versions?). This gives you plenty of time to learn them. |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98169, posted by Phlamethrower at 12:03, 6/2/2007, in reply to message #98168 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
Yes, GTA 1/2 had static cities (Although a couple of small sections did change as you played the game). GTA 1 had 3 cities, all 256x256 (Same size as the above). GTA 2 had three sections of the same city, which AFAIK were 256x256 also.
At the moment I'm hoping to have around 4 main cities, with several smaller towns in 'wilderness' areas inbetween. I have tried making a GTA 1 city in the past, and it takes an incredibly long time - especially if you want it to be interesting and contain the same amount of detail throughout. So using a random city generator was an obvious solution (and an excuse to do some interesting programming )
Currently I'm thinking that the cities will different each time you start a new game. However, as you say, this might make it too hard for players to learn the layout. So instead I might make some/all of the main cities static, and only use truly random cities for the wilderness towns.
Depending on how advanced the city generator gets, I may decide to just use the generator to make the basis for a city, then do some manual tweaking afterwards to set up the gang territories, place unique structures, etc. |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98170, posted by Phlamethrower at 12:37, 6/2/2007, in reply to message #98169 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
I think I'm going to have to start thinking about the overall structure of the game soon. E.g. how important the main storyline will be; how many of the missions are going to be vital to the player's advancement through the game, and how many can be ignored; etc. Currently I'm leaning towards a fairly freeform survivalist game, where the end goal is there if you want to go after it - similar to Elite or a roguelike. The more freeform the game, the more scope there is for random cities or missions. Of course, the game may start to force you towards that goal, but you won't be getting storyline missions rammed down your throat at every opportunity.
I guess I'd better go back and reread some of my first posts on the subject |
| |
[ Log in to reply ] |
| |
|
Tony Haines |
Message #98204, posted by Loris at 20:08, 6/2/2007, in reply to message #98148 |
Ha ha, me mine, mwahahahaha
Posts: 1025
|
What do you mean by 'empty space' - an actual empty area of the map that can't be reached, or just a big empty room? The algorithm will leave the latter. Yes, I did mean the latter.
Looking at the code, the algorithm works as follows:
<snip>
Yes, I think thats how I understood it to work from your prior description.
But doesn't that mean that, for an initial 256*256 room:
There are around 256 possible split points (depending on how you handle your "ends" ) - at least one of which is always rejected.
If a too-small split point is chosen, it may be chosen again with the same probability.
Therefore there is approximately a (1/256)^2 probability of creating only one large room.
I suppose you might argue that that is a pretty small probability. But there is the same chance of around a half empty (ie - wall-free) level, and considerably more for an around a quarter empty level, and so on.
When you've had a few splitting events it probably doesn't matter, it is just an interesting large room, full of aliens. But starting a game and suddenly having every monster on the level converge on your position might be a little disconcerting. |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98209, posted by Phlamethrower at 21:35, 6/2/2007, in reply to message #98204 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
Yeah, the fact it can still create large empty rooms is a bit of a problem, even if the chances of it happening are very small. If/when I do some more work on GAIO I'll probably fix it.
But a more serious issue is that I currently can't decide what I want DeathDawn to be like
One thing is clear, though - the game engine was designed around the idea of having distinct levels/maps, so any multi-city game design will have to give the player an incentive to stay in each city for a reasonable amount of time. Otherwise it will just be a case of driving from one side of the city to the other and waiting for the next loading screen to finish.
But the good news is that if I can't decide on a multi-city storyline/game design I can just design a single-city one, which will (presumably) significantly reduce the amount of time it takes for me to finish the game and for you lot to start playing it. Plus I can save the multi-city design for a sequel |
| |
[ Log in to reply ] |
| |
|
Tony Haines |
Message #98226, posted by Loris at 12:19, 7/2/2007, in reply to message #98209 |
Ha ha, me mine, mwahahahaha
Posts: 1025
|
But a more serious issue is that I currently can't decide what I want DeathDawn to be like On most platforms you'd have to avoid being too like GTA; but that probably isn't an issue on RISC OS. Personally I'd go for something with a bit of character of its own, rather than trying to clone it.
One thing is clear, though - the game engine was designed around the idea of having distinct levels/maps, so any multi-city game design will have to give the player an incentive to stay in each city for a reasonable amount of time. Otherwise it will just be a case of driving from one side of the city to the other and waiting for the next loading screen to finish. What you could do is what I wanted to do for a long time, and have several cities.
By this I mean find a few good seeds for your 'random number' generator which create interesting cities (potentially with different styles). The advantage of this is that you don't have to work all the bugs out of your generator, you just need to screen seeds to find good ones.
You'd need some reason to travel between them occasionally/frequently, so they wern't just 'levels'.
Depending on your game, you'll need some sort of open space between them - conceivably you could have different methods of transport which you wern't usually in control of (train, boat), then you won't have to worry too much about putting interesting things in there.
And you'll probably need big gaps between cities whatever you do, so I'd tweak your generator to make countryside for those. This should be easy - a few relatively straight roads, the odd village and an alternate spriteset. |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98227, posted by Phlamethrower at 12:43, 7/2/2007, in reply to message #98226 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
You seem to be thinking of all the same things I've been thinking of. Which is nice, but doesn't help me out of my current situation much
Depending on your game, you'll need some sort of open space between them - conceivably you could have different methods of transport which you wern't usually in control of (train, boat), then you won't have to worry too much about putting interesting things in there. I've also been thinking that the different transport types could be a way of preventing the player from moving between cities too quickly - e.g. when you first arrive in a city, you need to spend time trying repair an old railway engine, or steal a plane from the local criminals, etc. Or maybe a rival gang is controlling the gate out of the city, and you need to dispose of them in order to continue further. |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98335, posted by Phlamethrower at 23:34, 9/2/2007, in reply to message #98227 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
I still haven't thought up any big, sensible ideas for a storyline, so I think I'm just going to start with something simple. You will start in a city, the city will be fairly ordinary, and you will be doing a variety of missions for the different gangs/corporations in the city. We can just assume that the player's character is criminally motivated and feels that doing these missions is the right thing to do.
I still need to find some motivation for the player to move to a different city, though. In GTA 1/2 it was just a case of raising enough money (typically by completing most/all the missions), and then being allowed into the next city. This is a fairly weak system to use, and won't fit very well with my plans for interconnected maps.
There are a couple of more sensible ways I can think of unlocking the later cities - either you are granted access to them (e.g. by being given a new form of transport), or something forces you out of the current city (e.g. you become a notorious criminal and suddenly everyone is after you - or some disaster strikes the city that makes it almost impossible to live in).
The first option is OK, but can make things seem a little contrived. The second option is a bit more dramatic, but if you're able to revisit previous cities then it could potentially be a game-ruining experience - e.g. at one point in GTA 3, the mafia will turn against you. And the shotguns they use do a lot of damage, making it very difficult to travel through their territory without getting killed - much more difficult than with the other gangs that hate you. Also, if the disaster is big enough, it might not leave the city in an interesting state (e.g. if the population dies, it won't be very interesting at all to go back and visit)
One good thing I have realised though is that there's no particular reason why I need to prevent the player from going out and visiting other cities before the missions in them become available. In fact, I could make all the missions available from the start - but that might just lead to people getting lost and feeling like the game has no real direction (Mostly because it won't). And the more free I allow the player's movements to be, the more I get drawn towards my original idea of making DeathDawn a game of survival - that the game world is virtually limitless, you're able to go pretty much anywhere you want, and something forces you to keep moving if you want to survive (war, zombies, killer bees, council tax, etc.) Or there's some far away goal that you're searching for.
The trouble with this approach is that it would involve a lot of dynamic content, which could quickly get boring if not done right. Compared to a story-driven game where if you don't like it much you will at least know that there is an end to the game and it may be worth your time playing through to the end before giving up.
This almost makes me want to write to game modes - one story-driven one, and one dynamic, survival based one |
| |
[ Log in to reply ] |
| |
|
Andrew |
Message #98336, posted by andrew at 00:03, 10/2/2007, in reply to message #98335 |
Handbag Boi
Posts: 3439
|
Sounds like you like the idea of open-endedness. Personally, I'm drawn towards this unless there's a particular story I would want to tell. The "far away goal " sounds good though - something Elite arguably should have had but with plenty of discoveries along the way. |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98344, posted by Phlamethrower at 10:05, 10/2/2007, in reply to message #98336 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
Yeah, I think I might try the open-ended approach. The only thing that was stopping me before was that I couldn't find a sensible reason for the player to stop for a while in each city - but now I don't think that's much of an issue. The player might naturally want to explore the city (escpecially if there's the potential for some better weapons or vehicles hidden in it), and it should be possible to modify the map generator to run in the background to get rid of the pesky loading screens.
Of course, this will elevate it from the status of being a simple GTA clone to being a much more complicated free-roaming survival game that will take ages to balance |
| |
[ Log in to reply ] |
| |
|
Phil Mellor |
Message #98918, posted by monkeyson2 at 21:28, 21/2/2007, in reply to message #98344 |
 Please don't let them make me be a monkey butler
Posts: 12380
|
GTA for free, there:
http://www.rockstargames.com/classics/ |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #98920, posted by Phlamethrower at 21:33, 21/2/2007, in reply to message #98918 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
Yeah, but no source code |
| |
[ Log in to reply ] |
| |
|
Andrew Poole |
Message #98922, posted by andypoole at 22:00, 21/2/2007, in reply to message #98918 |
Posts: 5558
|
GTA for free, there:
http://www.rockstargames.com/classics/ And has been there for a few years now |
| |
[ Log in to reply ] |
| |
|
Phil Mellor |
Message #98930, posted by monkeyson2 at 00:28, 22/2/2007, in reply to message #98922 |
Please don't let them make me be a monkey butler
Posts: 12380
|
GTA for free, there:
http://www.rockstargames.com/classics/ And has been there for a few years now
 |
| |
[ Log in to reply ] |
| |
|
Jeffrey Lee |
Message #100341, posted by Phlamethrower at 13:41, 20/3/2007, in reply to message #98930 |
Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot Hot stuff
Posts: 15100
|
...and for my next trick, I suppose I should get this finished and released.
But I still haven't reached a definite decision on setting/style
Does anyone have any preferences? |
| |
[ Log in to reply ] |
| |
|
richard cheng |
Message #100346, posted by richcheng at 15:05, 20/3/2007, in reply to message #100341 |
Posts: 655
|
Olde Londone Towne!
It's a right cockerney knees up, Mother Brown! |
| |
[ Log in to reply ] |
| |
| Pages (22): |< <
15
> >|
|
