Leaderboard

I think this suggestion would make the end-game just another grind. If you've lost all your principle buildings, then that should be clue-1 that you've lost the match. If anything, I think we should hasten the end-game with a Sudden Death option. Lose all CCs and Fortresses and you get 5 mins to rebuild a CC or you lose.

I could make the planner somewhat faster, but the main function is still a recursive one. So, I still have to convert it into a while loop, because FF's profiler doesn't help with recursive calls. However, now it is ready to optimize for either cost or time and it is more robust now. I have tested it on all units and structures of athen, hele and mace found in the triple store, which gives a cost report like this: depth iter msec ops Name time food wood stone metal 78 79 13 20 units.hele.ship.trireme 535 850 1900 150 108 109 14 24 structures.mace.wonder 2035 850 3000 2400 2000The costs are total, as if you start with a CC only. Time is the minimum with one unit building and excludes gathering resources. I didn't manage to make nice tables here, shots are attached. The report generator is also online. If someone has a machine planning a units.athen.champion.marine in less than 5 msecs, I'd like to know...

Hi everyone, I'm coming to this burning issue, which has been discussed so many times. In a few time you'll be able to see alpha 16, and if you haven't played on svn, you will find a HUGE improvement on techs. I would say by eye that the number of techs were doubled from a15 to a16. And that is good, I love all the news stuff. Now, here are 5 numbers : 33 minutes of game, 20k of food, 20k of wood, 10k of stone and 10k of metal. That's the average of an expansionist player. And of course, all techs researched. Only the paired one didn't allow me to have really all of them. So there is, I think, a problem. Someone shouldn't be able to have all techs so fast in the game (30 minutes), at least in my opinion. Why that ? Because you can't really specialize in one way of playing. You have the eco guy, the rush guy, the defensive one... (eco != defensive, be carefull !) And everyone should be able to get only one fully direction of techs. Let's say that you have 10 military (m), 10 eco (e), 10 defensive (d), 10 find-a-something (f) different techs, and that currently in alpha16, beeing able to research everything takes you 10k of ressources. That's no far from reality, and remember I gather 60k in 33minutes, so that's 17% of my ressources to e able to upgrade everything. My proposition is the following one : go from 17% of all ressources to 100%. What does that mean ? That instead of being able of researching while expanding and fighting, you make a choice between both. For that, 90% of the techs should be as expensive or more than the phase technologies. From the other hand, the effects would be way more powerfull. Want some examples ? Currently : 100wood, +25% in farming rate Epicness : 500wood + 300 food, +50 % farming rate Currently : 100stone, +1 pop per house Epicness : 400wood, 300 stone ,+10 pop per house Currently : 300stone, +2 pop per house Epicness : 1000 wood, 1000 stone, +15% max pop Currently : 100wood, 100 metal, +2 hack attack on skirmishers Epicness : 400 wood, 500 metal, +8 hack attack on skirmishers Currently : 200 wood (I don't remember), +25% of speed for merchants Epicness : 600 foos, 500 wood, +60% of income for merchants Of course, this way it wouldn't be possible to have all technologies. That's a way to introduce some strategy, to know if in this situation you are allowed to spend your ressources on economic researches of you should do more soldiers, meanwhile military techs would give you a huge advantage. There's a need too to have cheap techs as currently, but not so much. Thank you for reading all of this @#$%, And have a A1A day !

The first more complex 3D model. The wagon prop still is missing. The reference of epicness was created by Ayakashi: (all the following research by Ayakashi too) Another storehouse + wagon reference: The roof references: And here my poor try to not let his work be for nothing: This time it's an especially awkward result, probably I have to add colours. .

Not sure if it should go here or with the ROTE threads but here goes anyways... This topic is just to lump in all the art shizzle that needed to be done to help finish this mod. I'll help as much as I can with the concept art department, and if anybody want me to draw anything, gimme a shout! I suppose we still need (correct me if I'm wrong or missing anything): Stable Gate Fortress* Blacksmith Wonder Any possible special buildings Champion Units? More Cavalry? And here I got the Storehouse to get things started: (tried to keep it consistent with existing structures but dunno if I done a good job of it lol) EDIT: A couple of references that might help with modelling:

Oh right, I have SVN too! I'll try to get on it tomorrow evening, european evening that is. ^^ By the way, Skirm cav and archer cav, just ranged cav in general, need to be removed from age 1. Currently you can do this broken strategy which consists of making 8 ranged cav then going to attack and the only way to hold that off is making ranged cav of your own or investing HEAVILY into slower, ranged units (archers and skirms). The problem with investing into slower ranged units is that ranged cav has the range and mobility to keep harassing and picking off units early game. Back home you just continue making villagers normally and you just get a huge, insurmountable lead. This can be easily changed, imo, by replace age 1 ranged cav with spear or sword cav. you can then rely on the cc for defense, as well as spearmen. it's much less broken than ranged cav which can do LOTS of damage and prevent units from gathering resources through their mere presence. spear and swordcav can still hunt and scout. See this game:

We should ™ have triggers in A17. With triggers, map-custom scripts can mark players as won or lost at any time they want.

Nothing is more irritating than having to spend 20+ minutes hunting down random females on a vast map once I have defeated the enemy and virtually destroyed every building. So so so irritating. A 5 minute sudden death option seems nice

That is because the enemy AI or the player cannot respond. They are defenseless when they can't build another CC having 499 of the last 3 resources. Too expensive for a potential recovery. Again. Age of empires gives the computer AI and the player the option to recover from almost nothing, I'm suggesting a possible option for recovery having the territory style of gameplay which has 0.a.d. And most players like challenge. If your companion can't recover from a huge assault, this game is getting worse. And I like the gentleman style of gameplay where you can let your enemy recover and atack him again 1 hour later. I can't do this with this game in its current state. I suggest time limit as an option and the resource alternative building as a feature.

Something like this would work.

I like the Stone vs. Metal pairing because it's a soft choice between Defenses and Units. Defenses use stone, while strong units use Metal. So what you are actually doing with this tech pair is making a soft strategic choice between the two.

Yes, each individual unit could be enhanced to an officer for that type of unit (spearman to spearman-officer, archer to archer-officer), this will cost some resource, take some time and it is only possible near an officers-tent (see previous post), so you cant get a new officer easily in a battle directy after you lost one.

Yes, pikemen are strong from the front in formation, individualy and from the side/rear they are very weak. Officers are a great idea, I think. With moving an officers you should move the whole formation so you can move formations more easy. Officers should be trained by a special building, sort of officers tent. For every tent 1 officer can be trained (until the officers dies, then another 1 can be trained and so on) and there should be a build limit for the tents, so there is a maximum of officers. These tents are a sort of house, but only for officers. Its not possible to have more then one officer in a formation.

I feel guilty that it took me so long to commit. So I've commited the Merchant, the fishing boat and the Snekkja. I was thinking of resizing the Snekkja a little, can I do so without messing up the armature?

Edit: It's not the complete diff, it's only the diff for summary.js of public. Add this to your charts.js: function showMetric(metric){ var m = Object.keys(metrics)[metric]; // Mark heading as active, deactivate others (TODO Use childnodes access instead for a much more elegant solution): selectPanel(metric, Object.keys(metrics) , {/*'sprite_background': 'bkBorderBlack',*/ 'textcolor': '200 200 255' }/*if active*/ , {/*'sprite': '',*/ 'textcolor': '255 255 255'/*TODO Better inherit => leave blank.*/ }/*if inactive*/ , 'chartMenuLabel'/*if sth. to prepend to bgToBeChangedObject*/ , ''/*if sth. to append to object where style to be changed*/ , false /* Is adjusting tab spacer desired? (Attention: Conflicts with main summary panel. Don't readjust!) */ ); // print(fmt("showMetric: %s, %s\n", metric, m)); curMetric = metric; // ... Warning: The approach I used for the highlighting is not my preference. I'd prefer to wrap the 'panelNames'Elements within a parent element and execute a function after Charts.action(xy, yz). e.g. <action on="press">Charts.action(panelIndex); unhighlightAllButThis(this.parent.children, this);</action> Where the function looked like follows: function unhighlightAllButThis(children, exception) { for (var childNode in children) { if (childNode != exception) { childNode.textcolor = 'r g b'; } else { childNode.textcolor = '255 255 255'; //<--reset } } }Here we had not to hardcode the panelNames lists/arrays.Edit: Forgot the line: childNode.textcolor = '255 255 255'; /* default */

Well if you buyed a PC by us and we build it for you we put a warranty sticker on it and if you break your warranty sticker because you build in a SSD then you lost your warranty (And that is with most shops) So its better to buy your parts and build it yourself. And thats also how it works with Acer, Intel, and all the other companies + some companies like HP/DELL you cant always use different components. But after 2/3 years your warranty is over and you can chance what you want.

I think the doors and stairstep sizes need scaled up. Other than that, the shapes look cool. I think the footprint should fill out a square better though, like the reference seems to do.

How would that be "much better" than a real full-functioning ceasefire/no-rush feature?

Honestly, it would be nice if the fonts included some metadata about size and style, then they could be linked to other variations. That would make dynamically increasing or decreasing font size (which is pretty essential in games, or UIs generally) easier and much better quality. It's a wish for the font system, one of many

For one thing, extern_libs4.lua doesn't have any OS X-specific handling for iconv, which looks problematic. I think when I tested, it at least had line 373 uncommented (probably commented out to not break Linux builds), but it needs more. We build iconv as part of build-osx-libs.sh, so extern_libs4.lua should be calling add_default_include_paths and add_default_lib_paths on OS X, similarly to on Windows (I don't know if the definitions are also needed). See how most other libs are being handled in extern_libs4.lua. Getting that straightened out should fix the iconv missing symbols, what the linker probably means is we didn't even bother to link iconv on OS X (it just can't be so explicit, since it doesn't know where the symbols might come from) That could be confirmed by doing "make verbose=1" or something and checking the linker flags when it fails - if there's no -L/path/to/iconv and -liconv, we know the problem.

Font sizes are pretty hard-coded. You can change a font file, so texts will look bigger, but this will change all texts that use those fonts (including the buttons), so it will lead to texts being too big. Maybe we can introduce new tags (like [h1], [h2], [emph], ...) that will be replaced with user-configurable fonts. Not sure how good this will work. In your case, you could also edit the resolution of the game in the config file. Normally it should be stretched when playing in fullscreen. So your fonts will appear bigger.

Just want to share some fun I'm having with a new group for Hannibal: Scouts. Their goal is to explore the whole map and record geographic features like water, impassable terrain, resources, shores etc. I hope I get them clever enough to detect they are on an island and build a dock + ship to continue exploring the rest of the archipelago. The concept I try to fulfil is: Bots don't cheat. However clouded the map is a bot can easily analyse any part of the map and locate the opponents head-quarter. Hannibal is not like that So here's a first result: If you look closely the dots mark a spot where the scout took a snapshot. While testing I gave the scout three points to visit. If detected idle he moves to the next and so on. Which made another group appear on my list: Patrol. Now that the grainpickers can deal with buildings and scouts with the terrain, the next group is close: miners. After that I have utilized nearly the whole API and will make progress very fast. Here the map to compare: PS: If you think now winning against Hannibal will be easy, just kill the scouts and he's blind - no chance, they reproduce in growing numbers and avoid any violence.

That's exactly my point. Being able to choose only TWO very powerfull techs in 15 instead of having 30 not-so-much-impactant techs. So doing this kind of stuff, you really will change your way of doing with the opponent, since you don't know which tech you should use for now, knowing that in this case, you won't be able to have the one you truly would like to have.

Flavius : having faction techs doesn't mean what I'm saying is irrelevant. You can have as much faction techs as you want, what I am saying is that is they are so cheap as currently, there is no real choice to do. There is an example that would cause a enormous difference between two games : - you are under an early rush, so you want to defend, so you do defensive techs. So you can't do eco techs and booming, for long time from now you will have to be defensive - the ennemy doesn't do any pressure on you, so you have time to boom economically, so you will prefer economics techs. The problem is that if you are under attack just after that, you won't be able to defend well yourself - you want to be faster than your ennemy → military tech → poor eco, you HAVE to succeed, that's a gamble. Lion : oh yeah *_* There's a part of responsibility in the maps that aren't designed for that purpose. Some maps where ther is place between the CC and the border of the map, which allows you to back cc the other.

It's not enabled in our engine currently, but I think this should be enough to enable it:

Thanks for the info. My proposed sqrt function doesn't actually work with 64 bit integers as input, so for now it's off the table. Still, I found that for a 10k sample of input values (logged from actual gameplay) the std::sqrt function gave exactly the same results as isqrt64 (and is 4.5x faster). I had my test code calculate sum of squared errors and it gave 0 for the current isqrt64 implementation versus std::sqrt. If that outcome equality holds true for all other relevant platforms it may be better to use std::sqrt instead, assuming its implementation is faster on all relevant platforms. I won't be able to test this myself (as I only have access to 64bit Intel Macs), but I attach my code and input samples file for anyone willing to do so. If the sum of squared errors for isqrt64 is higher than 0 anywhere (I set std::sqrt as the baseline), it can't be used due to potential out-of-sync issues. At the very least compiling for 64bit and 32bit gives the same results on my system. Edit: perhaps this is better fitted in a trac ticket, so it's not lost. I'll do that later. sqrt_test_with_samples.zip