Radagast.

When I committed my fixes today morning, I realized the arising conflicts. Well done nevertheless as you found the correct places and fixed it accordingly (as now each component is registered as global automatically since Sander's r15150 commit!). I then accidentally pushed without pull ... Radagast sometimes is pretty confused, one of them is forgetting to pull possible changes by other people (This was not very likely so far as noone ever bothered until you did today night .). Now the issues are fixed. We can start experimenting with the Capturing algorithm now. Slowly our list gets less depressing, thanks to Stanislas incredible new momentum that makes me accelerate too.

hehe, I like our Pharao in colour. How did you fix the Alpha channel not being rendered player color. Was it somehow black underneath the Alpha channel? I'm sure I checked that and it was white for me. Didn't your template changes show a "template not valid" error for the Pharao for you too?

Atlas view: short tunic with misproportions + prop misalignment: Atlas view: short tunic fixed: Player colour still not showing up for the Pharao's crown cloth. :/

- Fixed proportions. - Exchanged m_dress_cuffs with m_tunic_short body meshes. Lion, are you sure you created the texture for the tunic? There are some glitches, though they are minor. @Stan: Oh don't worry, comrade, I'll try to fix it now.

True, though the commoners may wear the the cloth (without the gold parts). Ok, I'll change it. Wait, it's not black and white. It's player color. No idea why it doesn't show up as such. Isn't player_trans as material correct? Also I checked in Gimp using quantumstate's procedure and it shows correct player color there. Later of course it would be nice, though I don't think anyone will notice it on the cloth but you are right it would look better than clean colours though it might be less universally applicable then but we'll see. (btw. there is no custom texture, it's using bronze_age_materials.png instead) This means I don't have to wait for your UnitAI removal commit anymore and should fix it myself? I would be happy to fix it, if it's ok.

I think the roof is not drawn accurate enough but it might be that it should not resemble straw. Or did the Han dynasty have straw roofs for temples like the Japanese for common houses?

I tried to make it player color but it apparently failed. Gold instead of white is an option but is it historical accurate? Might the crown even be reusable for other common Egyptian units in the white-player-color variant if we use gold-player-colour for the Pharao?

I also though of creating a bamboo door. That might be as simple as a common (already existing) bamboo small wall segment with one extra post to each side and a lot of ropes to bind one side together. We don't necessarily need 3. It's too much work. Han China can have one. It'll be enough. (Creating heroes is a lot of work and 1 per faction we model will suffice. The rest we or other contributors can still be added later.) Liu Bang is enough for finishing the Han Chinese. If anyone wants more heroes, then he might learn blender and create one. I'm against removing existing work of other artists. Create the new fortress as a variant and give the existing one a low frequency to make them appear less often. In reality not even two fortresses looked the same. Why should we artificially create such a scenario? I would love seeing these barracks in and nevertheless keeping as variation.

I would say, we trust Ayakashi and assume they immediately started to build complex temples in 50AD when Buddhism came to China. I'd say, we model this then: Alternatively we can trust Scion Development and model their concept. Don't we have a temple already? http://forum.sciondevelopment.com/viewtopic.php?f=42&t=187

Can be committed soon. The head looks even bigger than in blender, having been scaled down.

Also an interesting variant?

yes please Stan, this thrust is impressive! @2nd Question: Please use the newly created outpost. The default one no fits Han China at all. Your eagle's eye sees it all. @5th: If you want to replace the ship, do so. Though I wonder if we should remove it without being certain. I have no idea if it was historical accurate. Ayakashi might know .. I thought the same. It's almost seleucidic with those han blacksmith building references.

Nice summary of what we have to tackle. Here's my opinion if you are interested: :-) :-) :-) ("wall placement code and gate" mesh has to be added, thx for pointing out stan and leper) New : Fifth Question : Should we keep replace this boat (Seems Inacurrate) ? :-) As a variation of this : :-) (if we get around modelling it, then yes. To honour Ayakashi's work.) :-( (not now and in the near future) :-) if possible :-( perhaps an embassy later, but not in the near future. Anyone knows if a Government Center is essential for Han China, and if then why? :-) If possible:

Did it ever rotate before? No idea how the issue with the flaps intersecting with the body/shoulders looks but I think Enrique's proposal to make the crown's flaps a texture might look no better when the head moves but we can try later if the intersection anomaly turns out to be really awkward. I have still to commit the pharao's head mesh + xml, beard mesh + xml, fixed pharao Ramses II xml, head texture. I have to raise my morale to commit it today.

great. @Spahbod: What you proposed, was discussed on IRC into detail by wraitii, MythosRuler, sanderd, rada, ... all aspects were being discussed. Introduction of loyalty attribute for units+ buildings.loyalty trickle if inside own territory. (somewhat opposite to outpost's health decay, which would then be probably replaced by loyalty decay)Units garrisoned in a building increase the loyalty trickle.Converter units can convince other units + buildings to change side (once loyalty reaches 0, it starts over at 20% or 10% percent to avoid immediate recapture though I would love to not add this as the loyalty trickle boost by garrisoned units leads already to the same effect.).garrisoned units die (or later become POW) once the building has changed side (when loyalty has reached 0).Neutral buildings can start at 0.Converting a building is safer for the attacking units, but takes takes a long time in comparison to capturing (storming with high losses). Think of besieging a city. You have the choice: Either destroy it quickly or besiege the enemy and tear them down by time (desperation).Garrisoning units in a temple increases the territory wide loyalty trickle. => If your citizens pray a lot, then they get hope and hold out longer when being besieged by an enemy army.3 Choices: A building is captured once all garrisoned units are dead or fled (assumes all garissoned units have a prop/turret representation).converted (sacked by besieging) when loyalty is depleted.destroyed when health reaches 0. This levelling of a building also takes a lot of time, but less than besieging (read: converting) a building. Levelling a building is impossible without siege units (infantry weapons have no effect, can be simulated defining high hack armor for the building). => The first hits should have big impact. Later it's getting harder and harder to damage it further.The link between damaging a building and conversion/besieging is to be added later when introducing the morale attribute: Garrisoned units' morale will suffer when the building they are garrisoned in is damaged. (morale decay) The local loyalty trickle bonus that garrisoned units give (for this building only; unlike the global loyalty trickle that can be influence by units garrisoned in a temple), depends on the unit morale (a simple factor).=> Damage a building to decrease morale to increase your chances to sack the building.

hehe

To be split into parts/props to allow for the tech we want to add to every building: upgrading (for the storehouse we add the second level of the main building).

I recall that chinese common house reference where there indeed were such spiky but way more curved / twisted elements. (I based our new roofset on that). He also figured the roof texture makes the tiles look too square (they have a buckle). Also seen is the temple's roof (which is white-bluish). http://www.wildfiregames.com/forum/index.php?showtopic=18689&p=292488 The good point of stanislas' roofs is that they have an additionaly playercolour line below the roof pediment. I have left that one out for and had to add another loop for it to show below that rounded upper cylinder.

Great list. Fewer buildings makes the vikings unique. Wasn't a shipwreck planned as dock. Or a ship unpackable as CC. (=> Jarl's Hold for inland CC, Ship for shoreline CC) How big should each building be? It looks like we have few historical references for the footprint size of the outpost/tower?

Stan, do we have the lampions player coloured? Also should I add some to the storehouse?

I wonder if we can now finally convert all our 3DS exported animations to blender. http://code.blender.org/index.php/2014/06/supporting-game-developers-with-blender-2-71/

what a fortress brilliant player colour placement.

Sounds like what you want to do is what was discussed here on the forum already. I call it Hybrid AI, i.e. player takes control of one unit, the AI controls the rest. Others may call it RPG. You may call it taking leadership. What you want to achieve is what modders try to surrect anyway. Though the Roman timeframe is too rigid + limited + low a goal in my opinion. Supply lines is what also was planned. I have some frontline work done that could be used to calculate the best / safest + quickest (optimum) paths for supply lines. We want to give resources a mesh representation, units shall get a more flexible + even better visualized inventory than it already is currently (e.g. when bearing fruits, a basket is currently seen, but we want inventory to have a prop representation, then a building could have ladder attached as inventory/prop which could be grabbed by units or destroyed by enemies ... or lay around as loot when the building is destroyed. .. ). We might dream a bit much but well ... nothing is impossible. At least not in 0AD. My recommendation would be to develop the engine further, to create that Hybrid AI functionality, layout supply lines + conditional production chains, .. then the rest will follow with ease. (as most required artwork already exists, at least for your Roman setting)

Note: Topological sort should be read as auto-arrange function. Fixed some column misalignment issues. Alternative Enable button placement:

Topological sort now works functionally correct. We found a solution for the horizontal space problem: now the description is shown at the bottom of the available mods list. Already suggested that yesterday, and I hope Radagast beats me to it. Updated some mod info files (Aristeia + Convert attack type + other following) to the scheme you proposed (using =). It is indeed better readable in the mod dependencies column (because there we join all dependencies on " " (space)).