Demo 7 Download and Comments

[Martin the Dane]

First: I totally agree that there should be some way of replacing the lost men in a unit, and also that there should be some experience bonus to every unit based on the battels the unit has been fighting.

Well guess this should be in the military model thread, so I'll move on, and possibly poste something there.

Originally posted by Richard Bruns
It should definitely be impossible to build roads in enemy territory. After all, if you have an army there to protect the builders, the game calls it your territory. I think that roads through uninhabited land should also be prohibited. A good coherent model for road builidng would be:

You use the current interface to plan a road. Funds are allocated automatically, as per military tactics. If the road is in a province you control, the money comes out of the province economy and it is automatically built. Otherwise, you need to move a unit with engineers over the road builidng spot, and the money for the roads comes out of that unit´s supplies.

Obviously we don´t have unit supplies introduced yet, so until then a road can be built from civ funds if a unit is on top of the planned site.

I like the idea of automatic alocation of resources to roadbuilding. But I don't think all the funds for the roads should be drawn from the province, but that's a minor detail.

I do not like the idea of having to micromanage roadbuilding in uninhabited land, having a road building unit that could build roads as it moved along would be ok, as an enemy could kill the unit in order to stop the roadbuilding. This unit should then have a very low speed and an upkeep based on the roads it was building. the unit could also be used to speed up roadbuilding in ones own territory. Furthermore it could be an idea to draw pop directly to the roadbuilder unit, for a shorter or longer time, to speed up roadbuilding.

And finally I do agree that the cost of roads should reflect the size of the map.

[LFS]

regarding men lost in battle - it is an issue i really noticed in the carthage scenario, battles seem to knacker both sides. This is fairly un-historical, most ancient battles were fairly one sided in terms of the overall result - certainly there were battles where the victor took excessive casualities but they were exception not the rule.

(good ol Pyrrhrus comes to mind, but even his 2 major victories cost him only around 7000-8000 - about half that per battle- out of some initial 30000, and that was considered 'pyrrhic'. This game seems to make pyrrhic more desirable than a normal result).

I think the game does need some way of reflecting the generally one-sided nature of victories, and returning lost men back to the victor may be a way of achieving this - the victor recovers his wounded who lives to fight another day.

though the the current model does seem like it may well deal with combat during the period of the rifle and machine gun (circa 1860 - 1918)

[Mark_Everson]

Some good comments!

But please move all the military model issue discussions into the Mil. Model thread. We also had a thread back somewhere for how to handle roads too IIRC, where we should move the road building discussion. I'll try to find it. If you don't hear from me on the topic, someone could just create a new road discussion thread starting with the comments in this thread if they'd like to. . .

I just wanted to post to say that I've updated the first post to make it clear that Dawn is busted and point to Martin's suggested fix (Thanks Martin! ). I reproduce the text here so someone that's just looking at the end of the thread will see it.

Good to see all the suggestions and comments on how to improve things!

[Richard Bruns]

After the first few games, where I do things more or less by the book, my playtesting philosophy is probably best described as "Abuse the game until it breaks." Well, I broke something.

I wanted to play around with the game and the economics model, so I played Jericho as Emar, the civ on the bottom of the list, in order to give me the power to play around. I swiftly used my warrior to take over the civ to the south of me after they left their homeland undefended. I was curious to see what would happen if I colonized some sand dunes, so I took some of their people and marched southeast, and plopped them down in the middle of the desert on turn 17 to see what would happen to their population. I wanted to see if I could use civ funds to build up farms in that new province before they all starved.

The game never reached Turn 18. It kept running the turn, so I figured I had caused a bug, and I quit the game. After checking the economics log, I saw a gazillion lines saying:

ERROR null sector: farm

I then went back and proceeded withthe same strategy again, taking note of what happened before I hit the end turn button. The square said:

Popoulation is 500
No owning civilization

and there was no border around the square.

I was unable to open the province econ window, because there was apparently no province to open. I was also unable to pick up the population again, even though my unit was still standing right on top of them.

The white line around that square never moved, even though I was still able to select other squares. Also, the units box stayed the same even though the pother squares were selected. The escorting unit still had the setter icon as well, which means that the at least the graphical glitches start the instant the people are placed on the farmless square. Disbanding the unit did not add to the population of the square, but it did, oddly enough, put an end to the graphical glitches, allowing the white selection box to be moved as normal.

I am quite sure that this bug is reproducible. I wish you luck in getting rid of it before I manage to find something else to break.

[Martin the Dane]

After reading Richards post I tried setling the sea in Delenda, and well the result is preaty much the same as what Richard described, except there was no graphics glitch, except the missing border.

[Martin the Dane]

Ok I found a bug/cheat: I started Jericho as Emar picked-up all except two pop. This resulted in ~1000 pop at turn 2. I then did the same to a square with ~4500 the result was 5000+ pop at the start the next turn.

I thought that it might have been caused by people moving back from the surrounding land, but this is not the case.

After taking Jericho i counted the pop of the province, there was 4055 in the city, and 4027 in the rest of the province. So I picked up 4053 from the city, the next turn there was 5173 in the city, an extra populated square, and a total of 8794 plus the 4053 held by my unit. Still holding the 4053 the next turn totaled 10445 pop in Jericho.

[Martin the Dane]

Just to se how it worked I tried in Dawn to drop two pop and then move on, this resulted in the game hanging, and all sorts of graphics glitches. And the error logg, showing much the same as with colonizing uninhabitable squares.

If I on the other hand droped all the pop, and picked up all but two I got a pop at the next turn slightly over 1000. I then tried the drop all pick up all-2 move two squares, drop pickup etc. maby 4-6 times. All squares except one got more than 1000 pop from the manuver.

A completely unrelated isue is that the next demo should include the option of attcking a given enemy TF, it's quite anoying to pass the attacker just to watch him take your land while you chase him all over the place. Or the switch squares on each turn problem.

[Richard Bruns]

I tried this also, and got slightly different results. I started Dawn, dropped everybody, picked them up except for 2 people, and moved 2 squares NW. The bar at the bottom said that the population was 2, but the square detail frame said that the population was 1914. Apparently there is a gap between the bottom bar and the actual information, because it always skips counting the population of the last square that got a boost.

The Econ Info window showed product numbers that I would expect from a fully populated square, and a Labor of zero for all 4 sectors. Wages were all very low, except for production wages which were negative. Prices for farm and production were also very low, which makes sense because the square was apparently producing at its full capacity but there were very few people to consume the stuff.

On the next try, I picked up all but one person. The next turn, that square had 2189 people, an even bigger growth in population. The third time, I picked up all but one person, and the result was 2098 people, which was odd because the latter square had 12 farm sites and the former had 8, meaning that whatever is behind this bug causes a greater population jump for fewer farm sites. But this may just be random, because the third square I tried it on gave a population of 2392, despite appearing identical to teh square that gave 2189. I assuming population growth has a random factor; I have not seen the mechanism documented anywhere.

I think I saw a symptom of this bug in the "Initial sector Stats for Turn:" section, where it lists labor for farm at 2.8579216, or about 2,858 people, on the first turn. So somehow the full value of all 5000 people is being calculated before the turn ends. My theory is that they produce all that food, but then when the food is applied to the population growth, it uses a base of only 2 people so there is a huge surplus and the population jumps. This would explain why only putting down 2 people causes a bug, since it calculates stuff the instant they get put down, and there is a null error since they can´t make enough food by themselves.

However, it is interesting to note that, later in the log, it lists Labor = 0.0011432171, which is what I would expect for only two people.

Here is a log showing square details and the population the turn after settling. Except for the first turn, all these came form a population of one person left behind, and except for the third turn, they were all on plain squares with 8 farm sites. Around turn 4, I switched to a market economy, and before settling the 14th province I switched back to a traditional economy. The bigger numbers then are especially odd because I was colonizing with about 4500 people instead of 5000. It is easily apparent, however, that a market economy causes a lot less of this population growth for some reason.

Number | Population
1 | 1914
2 | 2189
3 | 2098
4 | 2392
5 | 502
6 | 487
7 | 503
8 | 494
9 | 500
10 | 498
11 | 510
12 | 517
13 | 521
14 | 2272
15 | 2313
16 | 2271

Obviously, with this kind of population base, fending off the barbarians was easy. I could make warriors sprout up everywhere. But they moved around a lot, and it got annoying to have to chase them down all over my civ.

---

By the way, please fix those memory leaks as soon as possible. On many computers it is so bad I can´t even finish a game.

---

I would also like to report something that I have seen many times before but have not seen any comment on. I call this the "1% army supply bug", but it may not really be a bug. I´ll let the econ log do the explaining for me:

----- Army Supplies For Civ------

Army Supplies Needed (existing army) = 1.05
militaryPayRate for Natufian = 100.0
Army Supplies Available/Needed From Economy = 15.18957 / -8.509571
cost20 = 3.9245846
Army Supplies set to 1.0%

How is it possible that a negative number of army supplies are needed? This usually happens when the army size is low. It may also be related to changing the tax rate. A more reasonable calculation fomr the log, when the army was bigger is:

----- Army Supplies For Civ------

Army Supplies Needed (existing army) = 8.719662
militaryPayRate for Natufian = 100.0
Army Supplies Available/Needed From Economy = 14.098911 / 4.8525476
cost20 = 11.982861
Army Supplies set to 8.901634%

[Martin the Dane]

I followed the support costs in Dawn, and every second turn the support cost was 1%. I thought I had reported this, but aparently i forgot.

Could this be related to the mismatch between the reported and actual buildtimes. I have had several units built while the GUI reported less then one unit/turn. What I'm thinking is that to much is allocated and the leftover is then spent on the next turn.

[Mark_Everson]

Hi gents, thanks for all the good testing!

I will fix the settlement problems the next chance I get to code. They are simply issues that were overlooked.

The Army Supplies is working more or less correctly. A negative amt needed from economy simply means that the supplies stockpiled (sometimes there is a stockpile at the scenario start) are more than enough to handle the available troops.

The oscillations Martin reports in the post above are just due to a stupid controller to fix the level of army supplies. Right now it doesn't know the troop builds for the next turn and so it can oscillate when the troop builds in a give turn are a significant fraction of the army that already exists. Eventually I will make it smarter.

Richard, we'll see what we can do about the memory leaks or whatever the issue is, and give it a top priority.

Programmers, is there anyone who feels competent to figure out why Clash grinds to a halt on Richard's systems? Richard, can you describe the systems?

[Richard Bruns]

I am not a programmer, and I am competent enough to figure out why Clash runs at the speed of molasses. These things have 32 megs of RAM, most of which is of course eaten by the OS and a gazillion other extra things that the administrators load them up with.

I can usually run about 30 to 50 turns before the game starts to rely on the hard drive for virtual memory. However, the number of turns it takes to reach this bottleneck seems directly related to the number of provinces in the game. It ran a lot slower when I tested the settlement population jump described in the post above. This seems to indicate that economy model calculations are at least partly to blame. And as mentioned before, the event popup boxes and several other GUI elements are memory hogs.

[Richard Bruns]

After getting on one of the better computers, with 256 megs of RAM, I just finished winning Jericho as Jericho. The game ran fine. I followed the basic strategy of swift northern expansion, taking over my northern neighbor so that all fighting took place in that province while Jericho churned out warriors. I also used the tactic of sending my weak and badly damaged units on suicide missions to ravage their territory, picking up their people while our armies were busy fighting me in the south. They usually had a garrison that prevented me from taking the main city, but I could still do damage in the outskirts of their civ.

I soon started noticing bugs relating to these zero population squares. For some reason, they always remained mine even though enemies would walk over them, as shown by the attached picture. I also got combat notices saying that they won fights in those squares, even though nobody was there. Here is one of these fights, from the log:

==== Beginning of combat ====
Combat in [7, 8] plain
Scout output phase: 0.0
Manoeuvre output = 0.0
Manoeuver output phase: 0.0

Fight won by Emar
==== End of combat ====

Also, any square with zero population was never re-colonized, even after I took over the capitol of that province. The people in that capitol expanded diagonally, refusing to send anyone to the zero population squares. I had to move people there manually, and then growth continued as normal. It seems that there is no migration from one part of the province to another, and that all squares must grow their population from their own people. I realize, of course, that this behavior is designed to fix the cheat that I discovered earlier, but if I own both the empty square and a square capable of colonizing it, than colonization should continue as normal.

I would also like to note that I have never, at any time while playing any of the demo scenarios, seen militia come up in any fight.

[Mark_Everson]

Originally posted by Richard Bruns
I just finished winning Jericho as Jericho.
(snip
I soon started noticing bugs relating to these zero population squares. For some reason, they always remained mine even though enemies would walk over them, as shown by the attached picture. I also got combat notices saying that they won fights in those squares, even though nobody was there. Here is one of these fights, from the log

Hi Richard, congrats on winning Jericho. But as I said in the Jericho thread, Picking up people that aren't your original folk is IMO a cheat, and shouldn't be used. YMMV

On the bug, its already in the list, sorry you didn't see it. Its a good idea to check the bug list for new entries every once in a while. . .

-164 (Mark) When population of a square is taken away so that square pop is now 0, square doesn’t change possession when conquered. At least it appears to take special circumstances for it to be conquered, although it can happen eventually. It appears the code to change square possession is ignored if pop = 0, suggest the test become if the square has an administration it can change hands. I looked for the code, but couldn’t find it after about 15 min of looking around, so I quit. [Med D7.3]
Workaround: never remove all population from a square, just leave a few and its ok.

I didn't know about some of the other issues with 0 pop though, so your effort wasn't entirely wasted. Just use the workaround immediately above and it'll be fine till we fix it in the next demo.

On militia, I think the unit power/population size ratio in Jericho is such that they would never appear. Militia only show when they might win the fight. Hopefully Laurent will correct me if I'm wrong.

[Richard Bruns]

I didn´t pick them up to use them, I picked them up to lay waste to the enemy´s territory. THat is IMO a perfectly acceptable and historical way to do things. The unit escorting them was usually shaughtered eventually, meaning that they vanished into thin air.

[Richard Bruns]

Oops, I should be working with known bugs. I´ll give a general report:

Bug 145 has been fixed, as far as I can tell. I am not seeing those problems anymore, even on the old systems.

I have not seen anything as egregious as bug 156 recently, but I sometimes get new units built even though I had not allocated any military funds for about a dozen turns. These units are also in recently conquered provinces.

I´ve never seen 157 again.

I´ve noticed that units only seem to use roads if both the starting square and destination are connected by a road. (160)

f135b has already been fixed.

One feature I would like is for the units box not to always appear on top. If I minimize the game and bring it back up again, the units box always appears on top of any event boxes tehre might be.