Suggestions for 0 A.D.

[Argalius]

I'm afraid a couple of the team members are in a funny mood and have decided to be sarcastic with you. We try to teach them not to do that. It's an ugly teaching process, involving batons and electric shocks. You won't believe the things that take place in the Staff Lounge sometimes...

←

Hmm ok... I already didn't get anything about the fourth dimension, but I thought he probably knows what he's talking about, it's probably some programmers slang... but it wasn't

[Ykkrosh]

probable (?) performance hit?

I don't know what the cost would be, but it would be non-zero - it needs more textures on each unit (and so more video memory), as well as doing more calculation per pixel. It'd also need the artists to do more work (e.g. creating high-poly versions of every unit, and using them to generate the normal map) and learn new skills, which takes time. And all the existing art would have to be updated.

We're not aiming to make the best looking game ever, and have limited resources, so unfortunately I don't think we can afford to use shiny new effects throughout the game. (Shiny things are still nice, though )

batons and electric shocks

Not again, not again!

[pyrolink]

Dot3 bump mapping (which uses normal maps) isn't very expensive in small scale use, but we're talking a huge terrain. It really should matter too much anyway. The tiles are very close together, and so per vertex is actually quite accurate in this case. It's not set for water either, and probably isn't even likely. (Trust me). Not only does it require more video memory and per pixel calculation, it also requires lots more calculation to set it up. I'm not BSing you by the way, I'm just trying to get it across that it can be tricky to pull off, and expensive in large scale.

Imagine every vertex on the map. Lot's of data. First, take the light position, multiply it by the inverse modelview matrix, take this and subtract the current vertex which will produce the object light vector. Now we must take the inverse of the tangent space matrix (TBN->TSB) and take the dot product of each vertex's Tangent, Binormal, and Normal which will produce a scalar that will be the equivalent of the X, Y, and Z components of the tangent light vector respectively. This must be done any time the light, a vertex, or the camera moves(for the camera moving meaning rotating). Not to mention you have to calculate the Tangent, Binormal, and normal for each vertex. And then, don't forget to set up a cube map which returns normalized vectors instead of texels. Now you set up the texture environment to take the dot product of the surface (the normal map's current pixel) and the vertex's Tangent Light Vector, not to mention the excessive blending required.

So....That's why we aren't going to have support for normal maps.

I hope that ended the dispute.

[Argalius]

Imagine every vertex on the map.  Lot's of data.  First, take the light position, multiply it by the inverse modelview matrix, take this and subtract the current vertex which will produce the object light vector.  Now we must take the inverse of the tangent space matrix (TBN->TSB) and take the dot product of each vertex's Tangent, Binormal, and Normal which will produce a scalar that will be the equivalent of the X, Y, and Z  components of the tangent light vector respectively.  This must be done any time the light, a vertex, or the camera moves(for the camera moving meaning rotating).  Not to mention you have to calculate the Tangent, Binormal, and normal for each vertex.  And then, don't forget to set up a cube map which returns normalized vectors instead of texels.  Now you set up the texture environment to take the dot product of the surface (the normal map's current pixel) and the vertex's Tangent Light Vector, not to mention the excessive blending required.

←

Ok, I'll keep my mouth shut from now on.

[Zorromorph]

Just wanted to say that I'm very excited about this project, and wish I had the technical knowledge to help out. Also, the one thing that's most important to me is that there is flexibility regarding the game's time scale -- the 30mins. to 1 hour for a game that was mentioned is obscenely short to someone of my potentiality, so I hope there will be the option to slow things down DRASTICALLY via a mod of some sort if desired.

ZM

[Wijitmaker]

I'm sure if you found a player that was your equal the game could go on for 2-4 hours, it will likely depend on the game mode chosen and how equally you are matched.

[Belisarivs]

Hi there.

I have some ideas:

What if Phalanx and other soldierz equipped vith pikes (or Sarissa) had minimal and maximal range equal?

Alone, one pikeman against one swordsman would have no chance, but formations 3 ranks deep would be absolutely destructive as all 3 ranks could attack at once, but would be vulnerable when attacked from flanks or rear.

In fact that was reason why Legion deffeated Phalanx, because while Alexander maintained exellent cavalry to protect flanks of Phalanx (he also put new type of soldiers named Hypaspists, who were something like better manouverable infatry to better connect his cavalry and infantry), later Macedonian generals believed only in power of Phalanx.

Would it be possible to make unit with extra strong first attackk?

It would represent Cavalry charge.

I have also something to say about Legion.

Legionaries were heavy assault troops, for defence they offered low firepower. I'm saying it because I'm afraid that with they ability to switch between pillum and gladius someone would use them preferably for ranged attack to prevent injury, but it is historically inaccurate, as pilum was only supporting weapon.

I also would like to ask whether those tools "shipped" with game will be for win32 only, as I use primarily Linux.

I expect that main tools be 2D and 3D graphics (I hope in implementation of Blender in game) editors and some text editors. Will it be sufficient?

[Dnas]

Well if the tools don't ship for Linux, I'm not going to be too happy with the rest of the team.

But the tools that we will ship with will be, as far I know, format converters and stuff. I believe we will be supplying tools to make editting the xml files easier and such, but they aren't neccessary.

For converting through Blender, at the moment, we only have conversion tools for 3DS MAX, but that will probably change in the future. After all, I'm starting to use Blender to unwrap my models. (Go LSCM Unwrapping! )

[Rinion]

I have a query.

Will the goods produced by the teams differ, like in The Last Alliance? If so, do the Celts get salt?

[Argalius]

I'd like to make another comment on the ships.

If you look at this screenshots, big ships look just awesome. But when you're in normal view they are just too big.

[Belisarivs]

I'd like to make another comment on the ships.

If you look at this screenshots, big ships look just awesome. But when you're in normal view they are just too big.

←

Right you are. They look reasonably big if compared to ground units, but they are too big to manouver with them in naval battle. Landing operations would be really hard to manage due to difference of sizes.

But personally, I'd prefer this style still, because that's more realistic and landing operations should be harder. I don't like style in AoK, when one transport ship carries 20 units and in second unloads all of them to your island. No need to say that it looks weird when transporting ship is slightly bigger than one cavalry unit, but can carry 20 of them.

Edited September 12, 2005 by godlike

[Mythos_Ruler]

Naval battles were very hard to maneuver in real life, so it stands to reason if one were to try to simulate that experience in some way, then yeah, ships would be big and clumsy (not necessarily unresponsive, there's a fine line there). And I agree with godlike; I hated the look of a little transport ship unloading 20 elephants, each ele being bigger than the ship! AOE, AOK, and AOM had this problem. In reality (ancient times), purely transport ships were usually bigger than the fighting ships, not smaller. The Persians actually invented new technologies in order to transport entire cavalry squadrons across the sea in large galleys. These ships weren't the size of a rowboat (AOX) either.

In 0 A.D. we hope to scale up the ships significantly, yet not make them so big as to fill up the screen and not be maneuverable on what would essentially become a puddle of water on most maps (instead of an ocean or large river the bodies of water are supposed to represent).

I think ES finally got their ships scale right. Not too big. Not too small. I think they could easily go about 10-20% bigger and still work fine, but no bigger than that. All I know, is I hope the days of toy boats (AOE-AOK-AOM) are over.

[Aeros]

In 0 A.D. we hope to scale up the ships significantly, yet not make them so big as to fill up the screen and not be maneuverable on what would essentially become a puddle of water on most maps (instead of an ocean or large river the bodies of water are supposed to represent).

A very simple sollution to the scale issue is that if you're controlling a ship on the ocean, the camera could slowly zoom out to transition to a bigger view angle for better control, and once you get close to the shore again the camera will slowly zoom down back to normal view. Very simple thing that I'm sure will easily remove the scale issue and allow you to properly control naval battles. in 0AD you aren't the captain controlling the ship (if you were then the clunkyness and scale of the ship and its vissibility limitations would be a factor), instead you're controlling the entire naval force. Its the same logic that if you were the commander of an army in 0AD you wouldn't be stuck with a ground view of the world. (although that would be something interesting to try in a game).

To clarify on what I said, these are just my opinions and its not necissarily true that 0AD would turn out like this in anyway, I came up with the idea just now - but I'd definately like to see it tried for our ship model. There hasn't been much thought given to that yet since we aren't at that stage in development.

[Belisarivs]

A very simple sollution to the scale issue is that if you're controlling a ship on the ocean, the camera could slowly zoom out to transition to a bigger view angle for better control, and once you get close to the shore again the camera will slowly zoom down back to normal view. Very simple thing that I'm sure will easily remove the scale issue and allow you to properly control naval battles.

I agree, but not fully.

Movement around shore could be terrible as camera would still zoom in and zoom out.

I'd prefer hotkey for switching between zoom in and zoom out.

But that aoutomatic switching would be cool when switching between view on ground and sea, but that automatic switching would be disabled when view is centered on coast and player will choose when to zoom in or out by hotkey (to trevent frustration from permanent switching between zooms when moving with ships or ground units around coast).

Another cool idea would be to to automaticaly change zoom level when commanding ground unit or ship.

Example:

You sail with landing forces to coast, so camera will be zoomed out, but when youunload units from landing vessels and send them out, you'll choose to command ground units, so camera will automatically zoom in.

Are in game any attempts to prevent heavier ships to sail in small rivers? I disliked in AoK Galeons or Cannon Galeons sailing on tiny rivers and shallows and almost on the ground.

[Aeros]

Movement around shore could be terrible as camera would still zoom in and zoom out.

Not if it was done very smoothly - the idea I have in mind is very cinematic-like. It would take a few seconds to full zoom out and it wouldn't do it with the same speed and then stop suddenly.

Additionaly what I have in mind is exactly your example, it only zooms out if you're commanding ships on the shore, or if there is a ship in the view (whichever turns out to work better), but if you go back to clicking units on the battlefield, it'll then do the 5 second fade back to normal view.

[Mythos_Ruler]

Sounds nice and cinematic!

[Belisarivs]

Not if it was done very smoothly - the idea I have in mind is very cinematic-like. It would take a few seconds to full zoom out and it wouldn't do it with the same speed and then stop suddenly.

Additionaly what I have in mind is exactly your example, it only zooms out if you're commanding ships on the shore, or if there is a ship in the view (whichever turns out to work better), but if you go back to clicking units on the battlefield, it'll then do the 5 second fade back to normal view.

←

But what if you are commanding naval and ground battle at once?

What if I command battle on the gound and switch view to naval battle. I have to wait for 5 seconds before I can command navy effectively. That is too much. Wouldn't be better to switch zoom immediately?

[Titus Ultor]

But, at the same time, that would make the gameplay choppy. I'd prefer about two seconds, because five seconds is really a long time during RTS battles.

[Aeros]

The zoom wouldn't be dramatic, maybe a 33% increase in view. The ships aren't going to be designed to take up the entire screen.

If you're battling on shore, then you would simply continue battling, you would just have a slightly larger view due to a number of ships being near the screen as well.

2 seconds, 4 seconds, 5 seconds, that all counts as " a few seconds" or however long it takes for the effect to be good and not anoying. Eitherway its a concept I said would be worth investigating, what is the point argueing about it now especially since I don't think anyone is quite getting what I'm saying.

I have to wait for 5 seconds before I can command navy effectively. That is too much

Why? Gameplay would never stop. Think of cinematics in AoM playing without the letterboxes on, you still see the UI and have full control of everything, think of that but instead of playing a track it simply gradually zooms out. Whats the problem there, it wouldn't be necissary to be zoomed out to control the navy, it just makes it more comfortable.

[Belisarivs]

I meant 5 seconds too much, because if I command naval and ground battle at ince, I switch views quickly. This means, that I wouldn't wait 5 sec. untill zoomed in or out. This means for me. that I will never have reasonable zoom to command units in time, when it is really needed, as I don't want to loose any battle.

My idea is, that if you are zoomed to ground units and switch view to the sea, you will be zoomed out immediatelly and vice versa.

But cinematic zooming on shores would be OK.

[Mythos_Ruler]

godlike - why would you have to "wait"? Like CheeZy said, the zoom is real-time. You can still command units and ships while the zoom is occuring.

[Belisarivs]

godlike - why would you have to "wait"? Like CheeZy said, the zoom is real-time. You can still command units and ships while the zoom is occuring.

←

I understand that it is real-time and can still commnd units, but I won't be able to command units effectively as my zoom will be too close or far. I know that 5 seconds isn't that much, but battles are usually quick and those few seconds could play at my disadvantage.

Just imagine, you command naval and ground battle at once, so you keep switching views quite quickly, but you have to wait for zoom in or out, to command units with maximal effectiveness.

My idea is, that when you look at ground, you are zoomed in, and when you switch view at sea you are zoomed out immediatelly (and vice versa) and when you move your view through the shore you are in real-time slowly zoomed out or in (if you move from the ground to the sea and vice versa).

That's all.

Anyway good luck with development.

[Mythos_Ruler]

Actually, the zoom need only take roughly 1-2 secs.

[thedrunkyak]

Any programmers out there that want to join a great team that has made a lot of progress and still has work to be done that one could hone their skills on... that isn't going to be a flake and bug out when finds that great coding and documentation is a lot of WORK to get this thing well made?  If so, please apply.. see the 0ad webpage for where and how.

←

I recently applied after finding the project. It seems really interesting, you guys have put in a bunch of work and have great results so far. If you want some examples of what work I have done, shoot me an e-mail.

I do have a question: Do you plan to implement (or leave the folder open to allow for) user-added AI for computer players, CP? (For more experienced modders to create new strategies for armies). Also, are you going to set it up so you can watch the Computer play against itself (Creating a movie of the game, and watching it instead of interacting)?

Right now I'm catching up on all of the different forums. I would recommend checking out Bolo as a simple networked game where resources are limited and replenish themselves (trees grow back), you can demolish ground to create rivers, set traps, etc. It is different in that it's only a 2-d, simple game, whereas 3-d scales the complexity exponentially when all of this is required.

Good job, I can't wait for the game to come out. I also hope to be able to help where I can.

[Wijitmaker]

Hey Daniel, good to see you found your way to the forums. We got your application in on the 9th (looked good;)). Hopefully you got a reply back, but currently the plan is to start doing some interviews over the next few weeks. We are still waiting for applications to come in.

In reply to your query about AI, yes we plan to fully support AI scripters that would like to do work on the scripts after the release. We plan to have lots of documentation, access to the AI developers right here on the forum to answer questions, and of course the multiple scripts made for release with the game availible for examples. AI vs. AI is something that will be availible to do also. We have some guys from the Age of Kings community that really enjoy doing this in AoK. They have AI tournaments and stuff - if your interested check them out here: http://www.aiscripters.com/

Bolo? Haven't heard of that before. Is that an independant game?