I wanted to post this somewhere, and this seemed as good a place as any. It probably won’t interest anyone, so you can stop reading now.

I’m making this game to fund my long term goal of researching the economics of land rent. And sometimes – yesterday for example, I spend a whole morning making a single argument just a tiny bit more logical.  Other times - today, for example – I spend all my time reading about historical GDP and house prices, and spend no time at all making the game. I like to feel that I’m making progress in both areas, y’see.

Today I used a spreadsheet for the first time in twenty years. I needed to compare how a country’s GDP would change under different scenarios, and the only practical way was to spreadsheet it. So I did, and finally got some firm numbers for something I’ve suspected for years but been unable to prove:

If the government decided to buy all the land in the country, it would have to issue bonds with a good rate of interest, and hope to have the money by the time they mature. Which would be a lot of money – roughly three times the GDP.  And the bottom line is, they could, and they could pay off those bonds in just fifteen years. Fifteen years! That’s not long. And that does not involve a single cutback – indeed – most areas of government spending would increase.

(In case you think that buying all the land means I’ve suddenly become a communist, I should explain that in this scenario the government would immediately sell that land again, at a great loss. The reason for buying it is simply to ensure that current land owners do not lose a penny when land values crash. The theory I’m working on – that land rent is superior to taxation – is a very old and popular one, but has always had the Achilles heel that it reduces land values and is therefore opposed by land owners. So today I worked out the numbers for a thought experiment: what if the government just bought all the land at full; market value? It’s a pretty extreme idea, and I do not for an instant think this is the way it would happen, but it’s useful to consider extremes in order to test the limits of a theory. And the result? The numbers add up. And in only fifteen years! All because of the miracle of compound interest.

“What are you talking about? How can that be? Are you mad?” I hear you ask. Well that’s a topic for another time.

I’m one week into the combining games process: almost all characters and graphics are in one place. As usual, it’s turning into a huge job. For example, there are hundreds of places where images and characters and views are referred to by number, and all of these will have to change. Also, all the code will have to be combined, which could take ages. And if a single line or number is wrong anywhere then this would be a serious bug… so all the games would need to be tested AGAIN…

…and all so I can make a game that will be ten times bigger than any AGS game ever made. Over on the AGS forums, GarageGothic made a good point: I had all these problems with RunAGSGame because nobody had ever done stuff like that before. Well, nobody has ever made a game ten times bigger than any other AGS game either. How do I know that there won’t be other problems in the future? Maybe not now, but maybe in two years or five or ten?

He’s right of course.

So I’m starting to think the unthinkable:

what if these games did not join together?

At least not in the “walk between games” sense.

I mean, what if the games still LOOKED like they joined (sharing common roads, etc), and what if there was still some common map, but to get to a different game you actually had to, you know, finish the game and start a new one?

Would that really be such a big deal to players? What do you think? It could be argued that the story really only exists in the player’s head, so the difference between  walking between games and stopping a game and starting another is only trivial. Maybe it’s no big deal.

Or on the other hand, maybe a modular game would be one of gaming’s great breakthroughs and worth whatever it takes to achieve?

I am just so utterly tired of spending month after month just trying to get the game working at a basic level, instead of actually making games.

Any thoughts?

I’ve had a few bug reports from the new AGS3 versions of the games, mostly because of how AGS3 handles sprites differently. (It has an ugly habit of creating blank sprite loops then using them, resulting in characters becoming invisible – I thought I’d trapped them all, but clearly not!) With more people downloading  Les Miserables now that it’s free, I’m also getting more suggestions for general  improvements. E.g. there’s about a week’s worth of fixes and improvements to do. None of these is fatal, so I’m still leaving the games up, but I’ve added a warning to the downloads page.

Regarding the new combined game, it’s all systems go. It takes about two days to move all the resources (sprites, etc.) from each old game to the new combined game, then it will be another two days to convert each game’s code into a new combined code base that should work identically for them all. Then I’ll fix the bugs mentioned above, then test all the games again. In theory that can all be complete in ten more days. But long time watchers will know that it’s always a good idea to double that.

To end on a very positive note, my goal all along has been to get the game into state where I can focus purely on the story and not worry about code. By combining all the stories I’ll be closer than ever to that goal:

1. The best code will be in every game – for example, game 4 has more sophisticated code than game 1.

2. All the endless bugs relating to between game code will be over: never again will RunAGSGame cause me to tear my hair out!

3. There should be no future surprises: everything about this game – except its size – is standard AGS code, as used and tested in hundreds of other games on millions of computers. And as far as I can tell, the size itself should not be a problem. If it is then I’ll just give up and make separate games, but let’s not even imagine that, because…

4. My goal has always been to have a vast game where there’s always more to discover: combining the games makes that easier than ever – e.g. if I want to add something to an old scene, or reuse an old scene, I can just do it, without any worries about complex retro code.

5. Long term, the goal was always to reuse more and more, resulting in  smaller and smaller additions of art and music. So in the early days, when there are less than fifty stories, it might seem odd to combine everything into one giant file. But toward the end, when there are around a hundred stories, any other strategy would seem crazy.

Anyway. Back to work. So much to do!

I just noticed all the replies to yesterday’s “one big game” idea, and now I’m undecided again. Maybe the situation will become clearer (to me) if I address all the issues raised:

Regular readers (bless you both) will know that the between-game code has been a huge headache over the past couple of years. For the past few days I’ve been getting feedback from the new games, and one issue has bothered me a lot: sometimes the between game code doesn’t always work for some people. It always works perfectly for me – I haven’t been able to duplicate any of the bug reports, and for most people the game works fine all the time.

The reported problems are erratic – e.g. the between game code will work perfectly ten times then break on the eleventh, or not work once then work perfectly forever. It might be something to do with how Windows accesses files and shortcuts? One thing is certain: it will take a long time to trace the problem.

I’m working almost blind here: I’m not a Windows expert, I can’t reproduce the bugs, and they involve RunAGSGame, a feature which has never been used in this way before. So it seems highly likely that problems like this will sometimes appear in the future. So I’m considering a radical solution: to cut the Gordian knot.

I’m looking at the possibility of combining all the games into one single giant file. Thus ending all between games problems forever.

This may not be possible. It would mean creating a game that, eventually, will be 5-10 gigabytes in size, saved games would be 10 megabytes each, and a lot of other stuff would need to be rethought. I’ll explore the possibilities tonight.

I just realized – the game centers on the destruction of the Bastille (the prison in Paris – its destruction is the key date that marks the start of the French revolution), and it was finished just in time for Bastille Day, 14th July. It’s a sign!

Uploading is complete, the new web site seems to work… see what you think!

Hopefully one or two blog visitors will want to sample A Tale of Two Cities in the next couple of days. When I get reports that all runs smoothly then I’ll start announcing the game everywhere else.

Remember that this uses a new engine that is not compatible with the old one, so if you want to try moving between games (as I hope you do) but you only want to buy A Tale of Two Cities (if you’ve already bought the other games and are short of cash) then you’ll need to download the new free Les Miserables game as well, so you’ll at least have two games to move between. the Les Miserables map then lets you choose locations in A Tale of Two Cities as well.

I think I’ve got rid of that annoying scroll bar on the forum and blog, but you may need to choose ‘reload frame’ to see it.

Please let me know if there are any problems. Or better still, if there aren’t.

P.S. I’ve also updated a lot of the non-game pages (those dealing with comics), because I’m still not happy with those. The main page and the best comics page are OK, but the rest need work. So I’ll maybe take a break and fix those up tomorrow while I wait to see if there’s any feedback from the main site. The land rent pages of course need a lot of work as well, but that’s a much bigger job so I’ll wait until Monte Cristo is finished before tackling those.

There are many benefits to living in a remote village, but quality broadband isn’t one of them.  We are now on day three of trying to upload the game files. After twenty hours of uploading I only had two files to go (at the breathtaking speed of  3 KB/s) then we had a thunderstorm and a power cut. The Java uploader has remembered SOME of what was half uploaded, but not the rest. So here we go again.

The folks at BMT say that they can transfer the files manually from my site on Monday, but I’m hoping this latest upload will be complete before that. I’ll keep you informed.

It only took ten hours on my not-very-broadband, but the games are uploaded. To save time I uploaded them to my own web space, and planned to then transfer them online to BMT (the payment provider).  However, true to Murphy’s law, I now discover that BMT cannot accept files from a URL, and I have to upload them directly to their site from my computer. Which is currently uploading 1.16 Gigabytes at the rate of 1.5 kilobytes per second. This may take some time.  :)

Update: BMT has excellent customer service – even though it’s Saturday they replied to my email within ten minutes. It’s possible that they may be able to do something. I’ll keep you posted.