Clash Demo 8 Download and Comments

[alms66]

Originally posted by Mark_Everson
BTW I was looking for input from Others on the importance of those features. ;-)

I understand wanting input for the transitions (well, not really, but let's say I do ), but the larger tiles needs no discussion whatsoever, it simply needs to be done. The height of a tile image should be 3/4 of the width, not 1/2. That's where the term "3/4 view" comes from!

[Fosse]

I've just gotten around to downloading Demo 8 to my laptop at work.

It's a 1.4 GHz, 256MB running Windows XP.

I've been fiddling with it over the past hour or two, and have played Dawn fast and loose to my defeat, Carthage as Rome until the program froze up about a dozen turns into it, and about five turns of the Ancient Middle East.

The following are my quickest first impressions. It's been ages since I played the last demo version, and I've been checking the forums once a week or so, so I feel like I have a decent idea of what's going on.

Long decimals are mostly gone, but I still see them sometimes. They really make things seem weird. Like I'm running a computer program instead of leading people. Also, in Carthage with giant militaries there should really be comas in the battle reports. Seeing an attack of 4,596,202 is better than 4596202, especially in a box full of numbers.

I don't know if the ruler stuff is working in all the scenarios, but looking at it I'm not sure exactly what "Empire Preferences" are. Is that the effective decision made by the other blocs before the ruler stuff comes into it?

Some panels in the UI open up to be too small. The social panel in Dawn opened so tiny that I didn't even know the pie charts were there until I expanded it again while playing around. Also, what are those pies?

Pull down menus don't open under their button. This is EXTREMELY annoying.

In Dawn the southern barbarians took 2 of the 3 province tiles, but the capital was still mine. I got a report that I had lost control of the province. Is this supposed to happen this way?

In Dawn a little grey box often appears on the Full map. it moves around as I click things.

When you lose a province, it vanishes from the map even though it still exists. My cartographers didn't forget about it, it should still show up, just in the dimmed fog of war.

Before Rome locked up I found myself not always knowing what had happened between turns.

If little icons appeared in squares where battles had happened (and maybe a different icon where they are still happening), this would help out.
Also, since enemy units "pop" from one place to the next, it would be nice if we could see their previous movement path from the last few turns.

So in a turn with many battles on many tiles, I would see the paths the enemy took to get there, and there would be little shields showing where things went down. This saves from having to check every tile in a crowded area when the map gets extremely full.


In the Ancient mideast. When i capture native land it becomes my province, but called Native Land. I currently have two long strings of this province on my north and south. Does the program consider them to be the same province? If not, could they just be called Native land (1) or (2) until I get around to naming them?

Also, I'm on turn 4 and each turn takes around 20 seconds to process, and grinds the system to a halt while it's at it.


Great work guys, I'm looking forward to the other scenarios when I'm not at work and afraid of being caught. :b

[LDiCesare]

Hi Fosse,

Some panels in the UI open up to be too small. The social panel in Dawn opened so tiny that I didn't even know the pie charts were there until I expanded it again while playing around. Also, what are those pies?

Social panel opened too small? I must check that again. At a time, it didn't even show...
(edit) The panel should have a default size of 500x300. If that's what you get and it's too small for you, I can increase the default size. If that's not that size, then we should check the settings file in the jar to see if it's smaller for some reason.(/edit)
The pies show the relative proportion of social classes, ethnicitiies and religions in you empire.
Social classes can help with the ruler panel: If you have 90% power held by the military and a social class aristocracy that makes only 10% of the population, you're likely to get riots if you go the way the military wants you to because they represent a minority of the population. Ethnic and religious pies show ethnicity and religion breakdown, which lets you know whether you should lower the ethnic/religious discrimination policies. Also, ethnic will tell you that you have to beware about a new ethnic group. If they are a small minority, they may riot very locally and you may have to handle them by hand rather than using the button (which TRIES to find the best solution).

Pull down menus don't open under their button. This is EXTREMELY annoying.

You saw that too... I am afraid this needs some serious overhaul of the ui infrastructure in order to fix that stuff si I didn't do it. I'm still annoyed and would probably like them to at least always open in the same place. Or something...

In Dawn the southern barbarians took 2 of the 3 province tiles, but the capital was still mine. I got a report that I had lost control of the province. Is this supposed to happen this way?

No. Only if the capital is tajen you lose the province, but one thing may happen: Barbs move into your square and take it and, in the same turn, you move into the same square and take it back. So you'd get a message. If you never moved your army from the capital, then it's a bug.

In Dawn a little grey box often appears on the Full map. it moves around as I click things.

Don't know this one.

When you lose a province, it vanishes from the map even though it still exists. My cartographers didn't forget about it, it should still show up, just in the dimmed fog of war.

I'm (tangentially) working on this, so that will be fixed.

Before Rome locked up I found myself not always knowing what had happened between turns.

If little icons appeared in squares where battles had happened (and maybe a different icon where they are still happening), this would help out.
Also, since enemy units "pop" from one place to the next, it would be nice if we could see their previous movement path from the last few turns.

Icons for battles should be doable. The units don't really pop, they pop 10 times (one per tick) but that's actually often too fast and they do not glide, so it shows badly. Having smoothed movement would be an option, but since the move happens in one tick, it would be hard to show. Coordinates (polygons) would make things easier... I don't know about showing a previous path.

In the Ancient mideast. When i capture native land it becomes my province, but called Native Land. I currently have two long strings of this province on my north and south. Does the program consider them to be the same province? If not, could they just be called Native land (1) or (2) until I get around to naming them?

Yes, that's the same province. You can right click on a square to affect a square to a different (even new) province. Though square per square, that's tedious. This scenario has a huge, single, Native Land province, which not what we'd like in the long run.

Also, I'm on turn 4 and each turn takes around 20 seconds to process, and grinds the system to a halt while it's at it.

Well, will that make it better that it used to be worse? My computer is a laptop with 1.8GHz, 256Mb, XP. I think the problem lies a lot in the econ, but there may still be ai glitches as I didn't replay that scenario after the latest ai changes. I'll check it to see if there's one big thing that can be corrected fast (may very well be the case since you mention turn 4, which may be when some units are built).

[LDiCesare]

When you lose a province, it vanishes from the map even though it still exists. My cartographers didn't forget about it, it should still show up, just in the dimmed fog of war.

Okay, this is now coded: You'll keep seeing the provinces that you know. Note I don't do a civ3-like thing where you know when a city half the world away switches hands. You keep knowing the same info until you send a unit or get knowledge of the map to refresh your info.

[demipomme]

Clash is still alive!

Checking the Clash website periodically I thought that the project was dead as the website hasn't been updated since 2003. On the off chance I took a look at the forum and what do I find, a new demo!

It's looking really good and plays more smoothly now. I particularly like the feature of gaining control of a whole province once the capital is taken. However I would suggest that when capturing enemy territory, squares revert to original owner unless occupied for several turns. This would make tracking down lone enemy units that are stealing your square easier. I also think that a square should only be able to be part of a province if connected by an adjacent square.

To my knowledge, most of the world's land was occupied with nomadic, semi-nomadic and small settled tribes before the advent of civilisation. To model this is the game would require every square to be occupied, with each tribe owning between 1 and 5 squares. These are then assimilated into the growing civilisations by war and/or diplomacy.

What I dislike about the Jericho scenario (and all civilisation style games) is that I have to send units across vast swathes of uninhabited land to fight my enemies. This is unrealistic in two manners, firstly it didn't (to my knowledge) happen and secondly it wouldn't: most wars are fought over territory and if a civilisation has enough territory to expand into, why would it fight?

Even if squares are unoccupied, a civilisation can claim ownership to it. Take Antarctica for example.

I'm not too keen on the colonisation function as it seems to me a little too powerful. However it has an important role in moving ethnicities into new areas, so that loyal ethnicities secure control of area. There are many examples of this throughout the colonial period. However most people moved voluntarily rather than being moved by the military.

I did notice a couple of bugs. Firstly, the Tutorial and Jericho scenarios kept crashing saying "Still running last turn, cannot start a new one". This doesn't annoy me too much, as a windows user I'm used to things crashing, but being able to save would be nice (perhaps next demo?). Secondly, I failed the Tutorial because I wasn't attacked by the Northern Barbarians. This seems a little unfair to me as I had little control over their invasion plans.

Please don't take these negative comments too seriously, because the game is already the best civilisation style game out there. Cheers.

P.S. If you need someone to update the website, I know a little bit of html and would be glad to help.

[LDiCesare]

What I dislike about the Jericho scenario (and all civilisation style games) is that I have to send units across vast swathes of uninhabited land to fight my enemies. This is unrealistic in two manners, firstly it didn't (to my knowledge) happen and secondly it wouldn't: most wars are fought over territory and if a civilisation has enough territory to expand into, why would it fight?

This one is a tough decision. The Ancient Middle East scenario (and Clash demo 4 btw) has a fully settled map -warning, it's slow). But then doing some settling is fun, to me at least. And conquering uninhabited lands in historical times did occur: New Zealand had 0 human inhabitant in the first century for instance (one should check various remarks by Gary for details on settling of NZ).

The still running thing (usually) means a crash somewhere, and a stack trace can be displayed that helps debugging. I can't remember how to activate the traces when running the jar though (something like activating the console).

Check with Mark if you want to do some updating (just writing Demo 8 on the web would be great), but I don't know who even knows how to upload the files on the server...

[Mark_Everson]

Thanks for all the input Fosse and demipomme!

Originally posted by Fosse
Carthage as Rome until the program froze up about a dozen turns into it, and about five turns of the Ancient Middle East.

The detailed comments of what you want are great!

As Laurent said, there is a way to get us the debug info. If you run the program using the batch file (described near the end of the D/L post) you will usually get a stack trace put out to the dos window that we can use to debug the crash. If you're game, posting those is very helpful along with descriptive info on anything out of the ordinary you might have been doing (if anything occurs, there may just be nothing obvious.)

Long decimals are mostly gone, but I still see them sometimes. They really make things seem weird. (snip)

I'll see what I can do about that. Some of it is a problem with Java itself I think, where rounding errors at the edge of the precision lead to "2" being shown as 2.0000001 or whatever. Laurent, if you don't object, I've also been bothered by the combat output readability and could take a shot at it. If you would prefer to do that, then that's fine with me. I had a though of at least using K and M for thousands and millions when showing numbers that way at least leaves two significan figures. EG show 43312 as 43K and 1364985 as 1.4M.

In Dawn a little grey box often appears on the Full map. it moves around as I click things.

I know what this is. If you rt-click on an unowned square the little grey square appears as a near-zero-size version of the menu to change province name, province capital etc. This should be easy to fix. I'll add it to the bug list if it doesn't get fixed immediately.

Great work guys, I'm looking forward to the other scenarios when I'm not at work and afraid of being caught. :b

Thanks, Fosse!

[QUOTE] Originally posted by demipomme

I also think that a square should only be able to be part of a province if connected by an adjacent square.

We clearly need some more refinements in this area. Some suggestions are, I believe in the D8.1 planning thread. The reason non-connected squares can be part of the same province is simple. We want it to be possible to maintain province names over time. That means that if the A civ takes over province XXX of the B civ it will still be called XXX (unless the player renames it.) To keep this up, we need to allow weird things during conquest like the issue that you note. Overall I think our province name stability approach is a good one. What do you think now that you know the reason?

I did notice a couple of bugs. Firstly, the Tutorial and Jericho scenarios kept crashing saying "Still running last turn, cannot start a new one".

See my answer to Fosse above about running in batch mode to get us stack traces for errors. I do have to apologize though, we thought the demo was pretty clean in terms of crashes. You guys must be doing things that are out of the ordinary way that all our current regular players are doing things. If you get us the traces, we'll fix those bugs! At least most of 'em

Please don't take these negative comments too seriously, because the game is already the best civilisation style game out there. Cheers.

Thanks for the boost!

P.S. If you need someone to update the website, I know a little bit of html and would be glad to help.

Thanks for the offer! I have contacted che, our prior web guy. We will get him to make the D8 changes if he's willing. If he isn't we can turn it over to you if you're game.

Both Fosse and demipomme, we can also set you up with the "testbed" version that will let you run nearly-current versions of the game. (A lot of changes have already been made since D8.) If you are interested, email me, and I'll get you set up.

Cya,

Mark

[LDiCesare]

Laurent, if you don't object, I've also been bothered by the combat output readability and could take a shot at it. If you would prefer to do that, then that's fine with me. I had a though of at least using K and M for thousands and millions when showing numbers that way at least leaves two significan figures. EG show 43312 as 43K and 1364985 as 1.4M.

Go ahead.

[Mark_Everson]

Here is what the new combat output looks like:

Combat in [4, 1] flat Marsaglia won by Carthaginians

  Roman armies 
   lost 94 health in long-range fights
    Attack: 406K down to 180K
    Defense: 26 down to 24
    Health: 5.0K down to 1.7K
    Cost: 680 down to 267

  Carthaginian armies 
   lost 44 health in long-range fights
    Attack: 2.6M down to 2.5M
    Defense: 44
    Health: 38K down to 37K
    Cost: 7.3K down to 7.1K

Any comments? I could add attack odds at beginning and end with a relatively modest amount of effort. Is it worth it? What else?

[demipomme]

Originally posted by LDiCesare

And conquering uninhabited lands in historical times did occur: New Zealand had 0 human inhabitant in the first century for instance (one should check various remarks by Gary for details on settling of NZ).

Although NZ was settled in recent history, the mode of its colonisation was the same as the tribal level spread of humans throughout the world before the dawn of civilisation i.e. the settlers did not remain in contact with the people from which they originated (I'm guessing a bit here as my history of NZ is a little patchy, so correct me if I'm wrong). But I do agree that settling is fun!

Originally posted by Mark_Everson

Overall I think our province name stability approach is a good one. What do you think now that you know the reason?

Seems like a good idea. However if invading armies were forced to occupy a square for several turns before they could claim ownership, then the scattered province square problem would be less likely to happen in the first place.

I don't know whether this thread is the best place for a question, but I'll ask anyway: How realistic are you trying to make the game? The models listed on the website are so detailed it would seem you're going for a high level of realism (at least compared to the civ series), but the scale in the game seems out of whack. In Carthago delenda est, Italy is two squares wide. Italy is about 100 miles wide, so each square is 50 miles squared. Units travel one square a turn and a turn is 5 years. Doesn't add up.

Also, siege equipment was often built on the site by skilled craftsmen rather than transported from a city (esp. one hundreds of miles away).

[Mark_Everson]

Originally posted by demipomme
I don't know whether this thread is the best place for a question, but I'll ask anyway: How realistic are you trying to make the game? The models listed on the website are so detailed it would seem you're going for a high level of realism (at least compared to the civ series), but the scale in the game seems out of whack. In Carthago delenda est, Italy is two squares wide. Italy is about 100 miles wide, so each square is 50 miles squared. Units travel one square a turn and a turn is 5 years. Doesn't add up.

We aim to be fairly realistic, but gameplay trumps realism when the two collide. We have discussed different time and distance scales and how they interrelate quite a bit. I can't find the most in-depth thread with search, it's probably way back there. However here's something I said in a forum discussion quite a while ago:

If accuracy of military movement scales and economic time scales is of paramount importance to you, then set the econ time scale to 1 month and everything will work perfectly as in the real world. But there will be this little problem with playability since you'll need about 100k turns for a game starting in 6000BC...

That sums up the problem. The discussion of related topics (not the best one, but the one that I could find with search) is here .

The comments about the conquest of provinces and squares would probably be better discussed in the current military thread. Your idea seems reasonable, but would require significant code and gameplay changes which at this point are probably not worth the trip. We do know that we need to fix these things at some point.

-Mark

[Max Sinister]

Originally posted by LDiCesare
And conquering uninhabited lands in historical times did occur: New Zealand had 0 human inhabitant in the first century for instance (one should check various remarks by Gary for details on settling of NZ).

According to Jared Diamond, that often happened that way with the big islands. Hawaii, Madagaskar and Iceland also were settled by humans relatively late in history.

[demipomme]

Originally posted by Max Sinister


According to Jared Diamond, that often happened that way with the big islands. Hawaii, Madagaskar and Iceland also were settled by humans relatively late in history.

But was this ever done by what we would call a civilisation?

[demipomme]

Originally posted by Mark_Everson


The long term plan is to go to a continuous coordinate system.

What's a continous coordinate system? Isn't it just making the squares smaller?

Is it possible to map a spherical world with squares? Don't you have to start loosing squares when you get to the poles? I suppose this is really a maths question, but I was never very good at maths. What polygons can be used to model a sphere? I know pentagons work cause it would just be a scaled up docedecahedron. I know triangles work. And footballs are made up of pentagons and hexagons. But neither of these are particularly good for a game.

[alms66]

Originally posted by demipomme
What's a continous coordinate system? Isn't it just making the squares smaller?

A CCS is a method of unit movement which does not rely on tiles. It is also known as a pixel perfect movement system. Basically you calculate the distance in pixels, rather than tiles, from one point to another. In essence, I guess each pixel can be seen as a tiny square, so in that sense, it is just making the squares smaller.

Personally, I have no desire to ever see a version of Clash with Polygons and pixel perfect movement. (awaits the wrath of Gary) I'll take a tile-based system over that rubbish any day of the week. I certainly hope everyone is willing to complete Clash 100%, then implement that afterwards, otherwise I'm basically just wasting my time here, because I'll very rarely, if ever, play that end product.