Sayyan

It shouldn't take too long to get to a point where you can make a useful contribution to 0 A.D. by learning JavaScript. This tutorial will give you a good start, although as with most JavaScript tutorials it focuses on scripting for the web. Still, it should give you a good understanding of the language and from there you can start taking a look through the JavaScript files used by 0 A.D. - make small modifications, run the game, see what changes, rinse and repeat.

The summary screen changes introduced a new element. The patch is currently up for review.

Can these be placed somewhere publically accessible? The links come up with a HTTP auth login, but neither forum nor Trac usernames work.

Does there need to be a specific point to it? I assume it was put in simply because it gives you an idea of how much farming someone in comparison to how much use they made of livestock.

Unfortunately the "Share this topic" line underneath the thread doesn't have Google+'s +1 button, so I can't ironically +1 this.

Ah, my mistake.

Wouldn't it make more sense to have all animal corpses convert to Gaia when they die? Seems odd that a corpse would belong to anyone.

Reproduced this on the SVN copy. When the goats got to zero health their health bar dissapeared leaving food resource bar. However they stayed standing and citizens weren't able to gather the food from them. I've created a ticket for the bug on Trac: Enemy goats count towards the victory condition

Ah, haven't taken a look into the formats yet, will have a bash at it sometime this week.

Were there any ideas for the implementation of the deforestation stat? I assume it would be the percentage of the total trees that were cut down; but I imagine it would be non-trivial to get the total number of trees in the map at the beginning of the round.

A patch with all these changes can be found on ticket #686. Unit count split into army (infantry and cavalry), navy (warships) and workers (female citizens and citizen soldiers)Kill/death ratios for army and navyFavourite units for army and navyTemples and special buildings countWorkers trained countsTotal for gathered resourcesLoot counts for all resourcesEntity identities can have an optional <StatisticsGroupName>, to be used for statistics gathering where classes aren't flexible enough (currently only on temples and special buildings.) The summary GUI code has been changed a lot in this patch, so any edits to the summary.js and summary.xml files that aren't on SVN may not merge well. However I'm happy to merge any of these manually.

Okay, that's fixed now. Hopefully I'll add the patch for all this tonight, just testing now.

If I'm correct the soldiers don't get the Worker class, but get the CitizenSoldier class instead. As it is they are't be counted on the "Workers Trained" as it only counts units with the Worker class. Would it be preferable to rename that stat to "Citizens Trained", or include the citizen soldiers in the count?

To continue this discussion, I've implemented the following: Newly implemented: Unit count split into army (infantry and cavalry), navy (warships) and workers (female citizens)Kill/death ratios for army and navyFavourite units for army and navyTemples and special buildings countWorkers trained countsTotal for gathered resourcesLoot counts for all resources

As discussed on IRC, removing the DELAYLOAD on OpenAL fixes the issue for me. I'm running on 64bit Windows 7 as is historic_bruno (I think.) Once I've finished wrestling with the summary screen I'll take a look with Dependency Walker.

Is there anywhere where the 0 A.D.pathfinding is well documented so I can get a good overview of the current implementation?

To my mind it makes more sense if the territory area is shrouded in FoW, until you have units there. Essentially you've mapped your territories, but unless you have units or outposts there you'd be unaware of what's actually happening.

My reasoning for the metal being light grey was to continue the iron metaphor used elsewhere. As there's no legend (and really, should an RTS map need one) it isn't obvious what metal looks like on the map.

Might it not look weird if only some things have pixel art? I'd think if one resource had it (metal) the others should too, for consistency. Civic centres would make sense as the only buildings. Downscaling the resource icons may work (although trees should probably look vaguely like trees, not logs.

Wouldn't garrisoned units make more sense elsewhere on the UI? They don't seem conceptually related to formations.

Having checked in game again, the Gaia colour is currently the orange I've listed by "Animals" (I only found animals at the time, but saw that bushes and grapevines are also that colour.) Unless food sources owned by Gaia are a different colour from other entities owned by Gaia?

In response to historic_bruno's ticket regarding the minimap colours clashing I've created the following proposal for a new colour scheme. Changes: Player one has been altered to tonally match player two, which also helps player one contrast when on water.Player eight has been changed into a much darker grey in order to stick out as much as the other players.The wood colour has been lightened and made a little more neutral so it's less dominating; it may be wise to represent trees with smaller blocks of colour so as not to drown everything else out.Metal has been changed from off-yellow to light grey to differentiate it from player four and to provide it with more contrast.Animals have been lightenedKnown issues: The gaia colour isn't featured here, and would need to change with metal being a light grey colour. Any suggestions on the best colour choice for gaia?Yet to be tested in other biomes, there will likely need to be some tweaks on those.

I'd suggest that items commonly used together should be grouped together. Interface elements commonly used during battle should be physically proximate, but there's little need for them to be near to the construction options. By making the different areas of the interface more clearly defined you help users build up a mental map of the interface quicker and improve the learning curve of the game. This Gamasutra article on RTS user interface design is worth reading: http://www.gamasutra.com/view/feature/1839/too_many_clicks_unitbased_.php

If I update SVN and build the game in Release mode in VS 2008 will this provide an accurate test? If so I'll do this now and let you know of the improvement.

Just to add some more detail to this report. Myself and a friend have been playing multiplayer (internet, not LAN) games against two AI opponents. The slow downs seem to be fairly synchronised between us, lasting the same length of time as the game progresses until the game is fairly unplayable. We're able to finish the game, but it's freezing up every second for half a second by the end. Our machines are both reasonably powerful, an AMD Phenom 9950 and an overclocked Intel i7 each with 4GB of RAM, I can provide a more complete hardware/firmware/driver breakdown if useful.