Incorrect default walk speed?

[artoo]

@Stan`

 

Is the documentation on walk speed wrong?

I get the impression the unitmotion walk speed is in km/h, not in m/s as documented?

If I use accurate 1.5 m/s walkspeed, the animations are too slow.

If I use km/h, all is pretty nice?

 

Edit: Looking at actorViewer.cpp l. 401, it looks as if the default value for speed is too high, hardcoded? 7 f? m/s or km/h? The latter would kind of be accurate.

Edited November 18, 2021 by artoo

[Stan`]

Note that actor variants define the animation speed.

Edit projectile speed and turn rates were increased to

1) Make dancing even harder

2) Address the complaints on the forums about the game being too slow.

[artoo]

23 minutes ago, Stan` said:

Note that actor variants define the animation speed.

I was on the hunt for this mysterious factor that seems to apply to walk speeds, ie the physics of unit motion. The Editors's ActorViewer.cpp in my view gave a hint, it looks to me as if the mysterious factor is about 7, which pretty good approximately matches my current 1.5 m/s walk speed times atm factor 4, using accurate acceleration etc. 7 km/h would be a avg walk speed, but isn't it supposed to be m/s?

I think somewhere in the code might be a mismatch between m/s and km/h for walkspeed, and hence all derived speeds like run.

 

Simply put, my hunch is, instead of using default values in m/s, somehow km/h sneaked in. Would be little bug with big effect, like accurate walk speed of 1.5 m/s has to be multiplied by a factor to have fluid motion.

Edited November 18, 2021 by artoo

[Lion.Kanzen]

30 minutes ago, Stan` said:

Note that actor variants define the animation speed.

Edit projectile speed and turn rates were increased to

1) Make dancing even harder

2) Address the complaints on the forums about the game being too slow.

the game is fast with the CS.

Just tried the AoE IV and I'm disappointed.

Our game has more personality and soul.

How AoE is generic feels like EE 2 or 3.

[alre]

I must insist that meters and seconds have no real meaning in the game settings. All soldiers models are about 4m tall. Also speed is clearly not meant to be realistic, starting from training and building times, to walk speed and attack times. It's all about making the game play in the desired way.

[Reyhan]

4 minutes ago, alre said:

I must insist that meters and seconds have no real meaning in the game settings. All soldiers models are about 4m tall. Also speed is clearly not meant to be realistic, starting from training and building times, to walk speed and attack times. It's all about making the game play in the desired way.

Change the numbers, not the units! m/s is S.I basis normalised with respect to many other physical dimensions, making some computations much easier (for example consider acceleration). km/h may be handy in one respect but it is very tedius to convert all other quantities into this unit basis and perform calculations. 

If you really need it, think about the speed in km/h in your head then convert it into m/s before you write it into the code. 

Also I would like to recommend weakening the Iberians for A26. 

[artoo]

13 minutes ago, alre said:

It's all about making the game play in the desired way.

I am basically telling you, there is most probably a bug, inconsistency between m/s and km/h, which is significant. Unless this is an undocumented feature, when all values are in m/s, should be a bug, wrong default/fallback value.

Its a difference if the unit moves with accurate 1.5 m/s in game, or 7 m/s, if probably 7 km/h was in mind of the dev, ie got confused, which can happen.

To make it clear, convert 1.5 m/s to km/h.

Edited November 19, 2021 by artoo

[Reyhan]

8 minutes ago, artoo said:

To make it clear, convert 1.5 m/s to km/h.

Multiply the coefficient of the (m/s) basis by 3.6:

1.5x3.6=5.4km/h

If the units are inconsistent then it must be converted into the same basis. 

[artoo]

2 minutes ago, Reyhan said:

1.5x3.6=5.4km/h

Correct, 1.5 is a conservative walk speed, if you take 1.8 or 1.9 m/s, like maybe jogging, you get ~7 km/h.

So something is off with the walk speed, 7 m/s is like roadrunner, unless 7 km/h were meant to be set, ie ~1.9 m/s.

[alre]

18 minutes ago, artoo said:

there is most probably a bug

there is not. and even if there is, it's irrelevant.

[artoo]

1 minute ago, alre said:

there is not. and even if there is, it's irrelevant.

As a 20 years software developer, this pains me to read.

[Lion.Kanzen]

36 minutes ago, artoo said:

As a 20 years software developer, this pains me to read.

you'll get used to worse things here. lol.

It is only to be patient.

[artoo]

11 minutes ago, Lion.Kanzen said:

It is only to be patient.

I am, I tracked down the bug.

 

source/simulation2/components/CCmpUnitMotion.h, line 248, it sets a walk speed of 7.0.

Code explicitly documents m/s, so 7 m/s is wrong, should be 1.9 and all will be fine in terms of yet to achieve balance. I would call this a bug, unless deliberately chosen for mysterious reasons.

 

[Lion.Kanzen]

Just now, artoo said:

I am, I tracked down the bug.

 

source/simulation2/components/CCmpUnitMotion.h, line 248, it sets a walk speed of 7.0.

Code explicitly documents m/s, so 7 m/s is wrong, should be 1.9 and all will be fine in terms of yet to achieve balance. I would call this a bug, unless deliberately chosen for mysterious reasons.

 

it would be nice to make a new topic.

[artoo]

4 minutes ago, Lion.Kanzen said:

it would be nice to make a new topic.

Would be nice if I found myself proper time to make a phab acc, go through the process, to open a review request for patch, along with the balancing I have in a branch on GH.

[Stan`]

5 hours ago, artoo said:

source/simulation2/components/CCmpUnitMotion.h, line 248, it sets a walk speed of 7.0.

That's the XML schema template example... It's not even used.

It's just so the docs can provide examples of usage here https://docs.wildfiregames.com/entity-docs/a25.html#component.UnitMotion

[Stan`]

Split.

[Freagarach]

7 hours ago, alre said:

All soldiers models are about 4m tall.

Here is your answer, @artoo. If one would scale the units down to ~1.5 m (so roughly divide by three), the walk speed can be reduced without looking strange.

I'm not an expert and maybe @bb_ has more input on this.

[ChronA]

This is the same **** that makes people keep changing the projectile gravity to 9.81! The distance units in the game are not meters! They are their own thing with no consistent relation to real world distances. The numbers have to be fudged for anything feel right or make sense in gameplay. Yes, units walk insanely fast, projectile ranges are absurdly short, most buildings are comically tiny, and in the real world it takes months to plant a farm. 

[artoo]

3 hours ago, Freagarach said:

Here is your answer, @artoo. If one would scale the units down to ~1.5 m (so roughly divide by three), the walk speed can be reduced without looking strange.

Yes, thx, I arrive at a factor 3 roughly, and I was wondering where and why this is set.

Hence my attempt balancing has a 1:3 range approach, ie max bow range is 1/3 of real range.

 

Anyhow, I think the simulation is wrongly calibrated so to speak, depending if 7 m/s was deliberately picked, or a mistake made using actually 7 km/h without conversion to m/s.

The editor has 7 and 12 set for walk and run speed, same there, km/h or m/s?

12 km/h would be a proper run speed too, but 12 m/s, well, that's a Cheetah.

 

Edited November 19, 2021 by artoo

[Player of 0AD]

For buildings, a meter might be a meter, but for units a meter is clearly rather a food. So m/s is rather km/h then

[artoo]

14 minutes ago, ChronA said:

This is the same **** that makes people keep changing the projectile gravity to 9.81!

Correct, as upstream did with a patch that's probably not included yet.

You should not be having like 5x the G to make projectiles look ballistic, and that's not so much related to size and scaling. This is a RTS, real time strategy simulation, the engine has parameters, in SI units. Other RTS have an entire physics engine, so what G is set as default matters for example.

[Dakara]

Bigger units would be nice (prettier gameplay and graphics) even if it is not realistic in terms of scale. and especially that the units are not packed when they move and fight (like before what).

This is not the subject i know

[artoo]

5 hours ago, Freagarach said:

I'm not an expert and maybe @bb_ has more input on this.

This is what I am also interested in.

If indeed the default speeds are km/h, but the engine expects m/s, and editor has also km/h set in the game interface, would this little inconsistency trigger the redoing of say animation sequences, so the engine runs on a real walk speed?

The factor should be about 3.6 atm I guess, is this deliberate choice or mistake?

If it is choice, then unit motion values in the entity template should probably in km/h then?

It would explain much if it is a mistake, why the balancing of movement seems like mission impossible when using m/s in the templates. Despite the fact, it would be possible in theory to nicely simulate human and horse movement, including speeds and acceleration.

[LetswaveaBook]

15 hours ago, artoo said:

As a 20 years software developer, this pains me to read.

I am not sure about this, but I feel that once you get 21 years of experience you will realize that these numbers are just in-game numbers and they aren't related to reality. The numbers probably have been chosen such that they make the game feel nice, such as addressing the complaints on the forums about the game being too slow.