aeonios

I didn't even notice that, but obviously he took that into account when setting resources to very low and was counting on his opponents not knowing about that, which I didn't and didn't even notice on the replay. No, when I was first attacked it was a cav rush, and I was way behind on units then. He used that to slow me down to ensure that I couldn't recover, while he still had enough units gathering resources that the lost time with the cav didn't matter. By the time he attacked with rams I had just barely passed him with pop, but I was still severely behind in resources which meant that if I had used every unit at my disposal to fight him off I would have gone broke and died anyway. No really? When I spent most of the game 20-30 units behind and panicking you think I could stop to think about countering rams? The only thing I was thinking about was scrambling trying to play catch up in pop count, and even then I just barely managed to succeed at it. Given the huge disadvantage I started out with I think it says a lot that I even managed that much. I've seen this guy do the exact same thing to other people over and over. It's basically all he does. If you'd seen any of the other games I've played you wouldn't be insinuating that skill had anything to do with it. I've beaten stronger people in a fair fight, more so now that I've gotten used to multiplayer. For example: Sophie_Hatter, new but surprisingly strong with ptol: commands.txt CAGD_Lulofun, kush vs kush: commands.txt Wycombe, kush vs brit: commands.txt So yeah, sure I was angry that I lost, but losing is fine as long as you can learn something useful from it. The only thing I learned from that is that this guy has figured out a scheme where he can score free wins by using the map and map settings to gain an unfair advantage. He gets free resources by going straight to the treasures while his opponent gets crippled by the very low resource setting. I mean he wasn't a bad player overall, but can you honestly say that what he did was fair in any meaningful sense? It's basically like having a one-sided resource cheat.

That's a fair point. It really needs to be verified server side, so there will need to be a way for mod devs to declare the rated settings for their mod on the server even if it's defined in the mod's files. Mod.io is still rather new and I don't think infra really supports much of anything as is.

No, but if you're expecting to have 300 resources and then dump all 100 of your wood on a farmstead the game tends not to go in your favor.

Well usually people don't do screwy things with the settings. I've played a bunch of games since then and never had that happen again, at least for rated games. However sometimes people do screwy things when testing or playing nonstandard games and accidentally forget to fix the settings afterwards, and there are also a rare few bottom feeders who try to score free elons by abusing the settings to their advantage. Checking the settings for every game would be a pain, and lots of new players just don't know that this sort of thing can happen.

Point. Capturing siege weapons is pretty useless, especially with siege towers that are basically uncapturable when garrisoned. Vs siege towers your units happily suicide for no good reason. Siege towers also suck (they're useless vs buildings, some siege weapon eh?) and don't really do what they would have been used for historically (capturing fortifications). That said I don't think kush are actually particularly weak. They don't have sword cav but they do have sword infantry, merc sword infantry, and champion sword infantry trainable from temples. Their merc skirms are pretty brutal too. They also have both ranged and melee champions and a hero that gives them a massive buff.

How did you do that? Full support for a glow effect would be relatively expensive (you'd basically have to apply bloom twice, with extra full-screen buffers, plus extra hacky and also expensive shader magic) so even a hacky solution would be preferable. Also vanilla 0ad has no real use for glow effects like that, except maybe for a few rare odd things.

I'm afraid that'd require a lot more than just a shader. You'd have to specially copy the glowy bits into their own bloom buffer and then use a special extra bloom pass just for glows. At the moment there's no way to get an effect like that without modifying the engine itself, which would not be easy to say the least, and people wouldn't be able to run your mod using the vanilla engine like they could with any other mod. What you could do instead possibly though: A: Make the blue parts brighter colored. The glow effect only replaces the lit color with the raw surface color, so if you use a dark surface color that's all you get. B: Emulate the bloom using a sort of particle effect that follows the unit's face around. I don't know anything about pyrogenesis particles so I can't be much help there, and you might be able to do it using some other similar method without particles (ie similar to tree foliage, with glow enabled), but that's the basic gist.

Apparently the shader already supports it, through the USE_SELF_LIGHT define. It uses the alpha channel of the specular texture, so it also requires that USE_SPECULAR_MAP be enabled, and AFAIK it's inverted so that 0 alpha = full glow and 1.0 alpha = no glow.

Hold my beer.

Eventually after the freezes are lifted I intend to look into fixing up multiplayer in a lot of ways, to reduce the lag and the load from spectators and whatever else. Saying it's alpha is totally an excuse. It's been alpha since 2001.

They're not good enough. I went and played my first rated game today, expecting to play vs a more or less even skilled opponent, and instead I got a game that was completely rigged in his favor. He set starting resources to very low on butana steppe and then rushed me repeatedly (which I actually fought off without much trouble) after my build got screwed up from the complete lack of expected starting resources. At every point in the game I was either behind in population or behind in resources, or both, and even though by the end of the game I managed to steadily pass him in population my resource starvation was apparent as I was unable to fight off his rams and repeated attacks and also unable to resupply my pop effectively. He had upgrades which I couldn't afford, and had it been actually fair I'm pretty sure I would have wiped the floor with him. Why is this sort of crap even permissible? Games should be won on skill, not by dirty tricks that people come up with to snare unsuspecting players with an unfavorable situation they didn't count on at game start. IMO all rated games should be locked at low starting resources and 300 pop. There are certainly other issues with rated games, but this one is easy to fix and very, very stupid because it not only allows but even perhaps encourages unscrupulous players to engage in unsportsmanlike behavior. That's really irritating, to say the least. also, the replay, just because: commands.txt

Missing metadata.

What? Most games last 30 minutes to an hour, sometimes more. How much slower can you get? Offensive towers are a completely legitimate tactic, just like offensive fortresses which is pretty standard. The issue with lag is complicated, and is probably a result of bad design in terms of the network architecture the game uses. Lag should be fixed, rather than worked around, and it especially should not be a motivation for changing the gameplay. If some specific element of the gameplay were to blame for the lag then ditching it might be understandable, but as far as I can tell there's no good reason why the game should lag as bad as it does. What? Offensive towers are bad but making it so that capturing an enemy building allows you to build a tower in their base is good? That's insane. Ranged siege weapons already have an operator as part of their model set. Given that they also cost a significant amount of population that would just make the game more of a pain and more turtle friendly. Feeding the turtles isn't a good way to improve anything. There have been plans for implementing siege-equippable ships for a long time, but who knows when that will ever be implemented. I agree that sea is kind of crappy, but fixing it isn't really simple. You mean no suicide? Meh. Suicide is kind of awkward, but you still need some way to deal with buildings blocking your units if you happen to screw up. People would then need to be able to attack their own buildings, or maybe have an un-build command to reclaim some of the resources. Also that'd be a significant nerf to capturing. Why? Commerce is painfully slow already even if you spam traders (which in addition to eating up population are very expensive) and the resources you get are random, which means that mostly trade doesn't produce the resources you actually need, and upgrading it takes resources away from actual combat-related things. Heroes aren't that op, and they're also very expensive and slow to build. Promoting units isn't necessarily a bad idea but it'd be complicated to implement cleanly and would be just one more complicated game mechanic to give players more pain.

That's way too complex imo. I mean allowing standard capture mechanics for ships wouldn't be a big deal, but an attack that had a 100% sink rate with one click would be op, and adding overcomplicated mechanics to counter it would be awful. An attack that did a fixed DoT is one thing, but a DoT with infinite damage that requires a special procedure to remove it is ridiculous. Fully loaded warships already kill each other reasonably quickly anyway, and really the biggest problem with sea is just that ship pathing sucks and ships are really dumb when it comes to garrisoning units.

NPOT doesn't work on ATI. Really it doesn't work well in general, and is not something we should encourage. Packing into an atlas does fix that though.

I just tested this and it does nothing. No matter how many times you hit the bell the citizen soldiers just stand there. I tried control clicking or right-clicking the bell too, no difference.

Yes, but you have to use control+right-click to get them to garrison. Unfortunately town bell does not do anything to citizen soldiers so if you want them to garrison you have to do it manually.

Alright, I reduced the blur scaling to 1.0. Any less than that would defeat the purpose, but 1.0 seems to work fine as far as I can tell. As they say, the perfect is the enemy of the good. Mapgen turned out to be somewhat more involved than I expected, but nonetheless I've gotten some excellent results. I was intending to work on ardennes forest but ended up modifying the alpine valley generator to produce maps in the style of the handcrafted version. I needed the mountain range generator for ardennes anyway and it ended up producing some really stunning maps with really interesting divisions of space.

Meh, it's difficult to get something that works well in all situations, but imo it's better to have too much than too little. Too little and unit shadows and towers don't get any penumbra at all.

I think you guys are misunderstanding what "soft shadows" means. In the context of graphics, 'soft shadows' means 'variable penumbra shadows'. In real life the size of a shadow's penumbra increases based on the size of the light and the distance between the shadow caster and the shadow receiver. So for example a tall tree casts a blurrier shadow than a short one: Or a giraffe's head casts a blurrier shadow than its legs do: And you can see here how the shadow smoothly becomes blurrier from the base of the tower to its top: The shadows are still sharp, but they look dramatically more realistic in terms of creating the impression of distance.

Welp, I ended up implementing percentage closer soft shadows. I wasn't originally planning to because I thought it'd be unworkable in openGL, but I got bored and made it work anyway. Sadly none of this will be in the game until a24, but at least it makes pretty screenshots. There are still plenty of things to do for graphics (mainly performance improvements) but for now I'm considering diving into the random map generator and improving the quality of the maps it produces. Right now the maps it creates tend to be boring or ugly, have bad lighting/environment settings, are often imbalanced, and so on. I think it could definitely produce some stunning maps that are well balanced and fun to play, it's just a matter of tinkering with the settings and generation scripts. Some screenshots of PCSS+DOF:

That actually sounds cool but I don't know that the engine supports stackable auras. Really they should stack defense though, not attack.

I did not know that. It's now on my to-do list for a24.

Increasing unit costs based on how many you already have would be bad. Artificial limits like that just encourage random unit spam and make the game worse by removing real strategic decisions. Stronger role-based balance should ensure that every unit has at least one hard counter and at least one soft counter, so that unit choice is more important without being inflexible. One problem is that not all factions have all unit types available, ie some don't have slingers or archers or javs, some have lots of cavalry choices while others don't, which makes role based balance more difficult. If a faction doesn't have units for a given role, then balancing for it means either buffing or nerfing the entire faction or creating a situation where some factions may hard counter others. IMO the current damage/armor types are overcomplicated as well.

What? Eh.. maybe? As an aside I ran a few simple tests and it turns out that most of the old trees except for pines and palms really don't use back faces. I suspect that all of enrique's trees do however (as he showed doing in his tutorial video). None of the bushes seem to use backfaces either. I'm curious to see how they'd look if backface drawing was enabled.