New and Modified Scenarios

[Richard Bruns]

That scenario doesn´t reference the technology file; it has a copy of it, which is about ten revisions out of date by now. But the difference is probably due to the fact that I gave Jericho tech of 1 in everything. This meant that they could make warriors, so they could make everything. If you are playing any other civ, they won´t have that tech, so all they make is mobs.

[Martin the Dane]

Hi I have run into a few minor problems while modifying the Dawn scenario. I desperately want to have an event trigger at the turn after another event. For some reason if I use a delay of 1 the event triggers two turns after the previouse event, but if I use a delay of 0 (or ommit the delay tags) the event never triggers.

The reason I ran into this problem is that there is no <unitmoveevent>

What I use is

  <eventaction>
    <condition>Is Size 2</condition>
    <delay>1</delay>
  </eventaction>

[LDiCesare]

What do you mean by unitmoveevent? I will try to check the delay stuff, but if you need another event, it'd be well to code it. Can you elaborate a bit?

[Martin the Dane]

What I was originally looking for was an event that would trigger when a (specified) unit was moved. The reason for this is that I cannot be sure the player follows the instructions in the tutorial.

When we instruct the player to move the first unit two squares to the NW I want to have a message show up on the next turn to give further instructions and to tell the player that the unit has reached the destination. (provided the unit got there) If the player moved the unit in the wrong direction it could be good to inform him/her about that.

The event could have a number of conditions, like start and end location, unit-path-end-point, unitname, civ, etc.

The workaround I use for Dawn is to use a ≶singlevisitevent> The problem with that is that I have no way of knowing if the player moved the unit but did it the wrong way. I could add a conditional event to check if the unit was moved to far away from the desired point. But that will not work properly unless the delay can be set to one turn. (but I might do it anyway.)

[Martin the Dane]

Oops found another eventbug.

The <buildunitevent> can only be used once, as it is always true if a unit has been built. So it's not actually an event but a condition. I was trying to have a message pop-up to remind the player to stop producing wariors after building the second additional warior, but it fired prematurely.

[Martin the Dane]

An update to the above: It seems most events stays fired even if the conditions are not met any longer. Take the <destroyarmies> event. This is always true if the civ in the <civilization> tag has at some point had no armies. In the Dawn scenario this is always true for the barbarian tribes, since their armies are created by events during the game.

This would not be so critical if the barbarians could not occupy land, but they do, and I want to tell the player how to get back that land.

[Martin the Dane]

Ok Here is my attempt at a Dawn scenario, there are still a few isues that I would have liked to cover in the scenario, but either the events model does not allow it or I'm to dumb to figure out how to achieve what I want.

What I have changed is as follows:

Victory condition now requires a pop of 15 k, this should be achieveable around the same time as the major northern barbarians arrive.

The personell requirements for all units have been reduced , so Wariors are now 250, BarbarianArmy (the northern barbarian units) are 500, and Chariots are 1000. To keep the balance I have changed the costs and strengths as well.

Chariots are available at Tech 2, and research is speeded a little to make it all doable in ~80 turns.

Finally I have moved the hint about horses to a little later in the game, and changed most of the events, so now there is mostly only one message per turn. Furthermore I have tried to avoid long stretches without events.

Unfortunately I have not been able to add the events I wanted for the late game. I would have liked to add a few extra econ-instructions to the late game, but I have not been able to come up with any sensible things to tell the player so that'll be my next project.

Please let me know what you think.

[LDiCesare]

I just downloaded the new dawn scenario.
I checked the events and am a bit surprised by what you said, Martin. The buildunitevent data is normally reset every turn, and the destroyarmies event looks at all units present at a given time. This means destroyarmies is true when there are no armies, not when they have just been destroyed. Making a conditional event (spawning of barbarian units AND destroyarmies -> save a condition) should work for that one.
I'll check more in depth, but the buildunitevent should work the way you want.

[LDiCesare]

Martin, the scenario plays well and is easy enough to complete by following the steps.
I found a few things which may need correction:
One is a title for a panel: Horses is confusing. "Horses rumours" or something like that might be better. When I saw "horses", I thought I had found them even though it was not the case.
Another point is my unit stopped 3 squares from its start position after the second move when I gave it the order to move 2 squares NW. The message didn't fit the situation well. Also, the direction in which the squares expand is random, so the target square (2 squares NW) may not be in the black, depending on which area the population would expand.

[Martin the Dane]

Thanks for the comments. The one about horses, well I should have thought of that, but well I didn't. About the movement message, this is where I wanted to have a unit-moved-event. So I could show the right messages, depending on the move. But I'll try to think of a better text.

[LDiCesare]

I corrected the bug for buildunitevent. The correct code was not called because the superclass method wasn't called by the subclass... I sent a new code bunch to Mark, so this should be fixed soon.

[Martin the Dane]

Thanks Laurent,

I figured a way to get a more correct message for the move instruction, so now the message varies depending on the way the settlement expanded, and the move message, dosn't mention wether or not you reach your destination, just that the unit moved.

[Martin the Dane]

I have been looking at the <destroyarmies> event, again. As I want to use it inside a <eventunion>, but I get the following error: "Setter not found for destroyarmies in game.events.EventUnion"

Another thing I'm looking for is a way to delete a condition from the conditions list, this however might not be nesecarry if all events could be nested inside the four container-events, and they themselfe could be nested.

[Mark_Everson]

I have just sent out a Testbed V7.6! Here is the new stuff from the readme. Please be sure to look over the bold parts of my post. We need to take action soon on unifying the various scenario approaches or we'll have a train wreck!

As usual if someone else wants it other than Richard and Martin, just email me and I'll send you a copy. I will send off the new unified codebase to the programmers in a few minutes.

Good work Everyone!

Clash Testbed 7.6 (post-Demo 7.2.1 code) Feb 16, 2003

Please delete the entire old version of the clash testbed (if applicable) before installing this one.

New info:

v. 7.6 Is the post-Demo 7.2.1 code made available so that Richard, Martin, and others can mess with the scenario files. This version has bug fixes and new small features. Also Laurent is making good progress on the real AI, and some small improvements from that are starting to make it into the game.

IMPORTANT: This version of the testbed uses the old-style (before Richard's recent changes) scenario files. It does have Martin's new version of Dawn.

We need to figure out our new scenario standard and come up with a merged version soon, before drift between them causes wasted effort! I suggest that be a priority for the immediate future.

For now, to use Richard's scenario versions, just replace everything in the class subdirectory of this zip EXCEPT class\game with the equivalent from the Testbed 7.5.1.

New features:

1. Laurent has been charging ahead with the new AI. In this version enemies don't make boneheaded attacks. Sometimes unfortunately when they are stymied they just sit there though. Hopefully that can be remedied soon.

2. Owen has done a number of fixes involving colonization. The most noticeable are that colonies next to existing provinces are added to that province, and that you can now form colonies by disbanding military units.

3. I have gotten the econ orders working more-or-less properly for military unit obsolescence now. When Cohorts replace Legions the Legion order disappears and an equivalent Cohort one takes its place. For now you lose any Legion investment that remains, I'll fix that soon.

4. Laurent put in fixes for builtunitevent for Martin.


Thanks for all the hard work! I'm hoping we can get Demo 7.3 out in a few weeks. If Laurent can get the AI significantly better soon, I think between the tech advancements and AI we could just call it Demo 8.

[Mark_Everson]

Hi Martin, I finally had time to check out the new Dawn scenario. It seems pretty good overall. You have certainly included a bunch of things that we didn't have in there before that should help out newbies. You had said in your post that you are looking for other Econ items to include later in the tutorial. One idea I had is that if we can get an event to turn on ROI in the Econ GUI, going over how ROI works would be a useful thing. One thing we could demonstrate that with is to have a chariot production automatically turn off as Army supplies gets above 80% of taxes or so. There is already a right up of how ROI works in the economy section of the demo 6 manual.

Some other comments...

You did a good job on thinning out the bulletins, but it seems there are still an awful lot of turns where two come up at once. It may be that I did things in a slightly different order from you, and so things that were spaced evenly for you became bunched up for me. A long time ago I requested Gary to make an event class that was turn-based but fuzzy, such that if another event were fired that turn it would wait up to a turn or two before firing. Unfortunately he never got around to it.

There were an awful lot of grammar errors, and I realize that part may be hard for you because you're not a native speaker. There is certainly no point in you spending large amounts of time getting the grammar right when someone else can fix it fairly quickly. But lets see if we can get any volunteers to do it!

I thought the first barbarian attack was too soon. I had just started building roads, and then needed to switch off the road-building orders to make military units to fight off the invasion. It seems to me if that attack were delayed by five turns or so things would be much smoother.

I may be wrong about this, but I didn't notice an instruction that says about the tech button to see where technological knowledge is. If the comment isn't in their we should have that.

Population growth is much slower now then he used to be in the original version of Dawn. I think that is because the amount of Army supplies needed changed, and Army supplies pull food from the economy. You might want to try starting out with extra people, something like 8000, and see if the whole thing doesn't work better. That would also allow us to have a higher population target for winning than the fairly measly 15,000.

I got the "We Survived" bulletin more than once during the game. You might want to look into that.

Another bit of instruction you can give late in the game, is after all the barbarians are dead and you are trying to maximize population growth, you can suggest investing in Food Tech, which makes the farming sector more productive and therefore population grows faster.

I really appreciate all the work you put into this, and am looking forward to seeing the next version.