historic_bruno

There's some info in the manual: http://trac.wildfiregames.com/wiki/Manual_SettingUpAGame#HostingJoiningaMultiplayergame That should be enough unless you encounter problems.

Hmm, it would still be nice to know why the hardware cursor is failing to display. Setting nohwcursor to true will cause cursor lag because it's tied into the renderer instead of the OS.

Yeah - don't depend on order, it's not just a bit dirty, that would be a filthy hack There's no reason why we can't change how actors are defined, we're still in alpha. Most artists don't actually mess with the actor XML, they use the actor editor GUI instead. Why not add a name/id attribute to the actor <texture>s that matches the ones in the material <sampler>s? Explicit naming and behavior is nice, if we do add such an attribute it shouldn't be optional, we can update all our actors and actor editor to comply - it's easy since they only reference one type of texture currently. I think good names would be like "base" (default), "normal", "specular", "ambient" (I was thinking "diffuse" instead of "base" but that might technically include AO too). The best part of that is the type of texture map is independent of both XML element ordering and filename.

Sometimes I get a slightly different error, again while placing the 5th waypoint flag: Function call failed: return value was -120003 (Handle tag mismatch (stale reference?)) Location: h_mgr.cpp:245 (h_data_tag) With this call stack (thread 1 is the relevant bit):

Hmm, what happens when we need to adjust the height of those shaded bits or change their spacing if the font needs more room? Then we not only need a new texture but all the code has to be updated as well. I agree with Brian, it really needs to be split into separate pieces (preferably a minimal set of reusable pieces).

OK I just reproduced this crash on Ubuntu, looking into it. For some reason it doesn't occur in Windows, perhaps a GCC/Linux only bug?

Hmm, I just placed about 50 of them on a civ centre with no errors... which OS are you using? Is there a chance you can run the game in a debuger (e.g. gdb) and get the real call stack as described here?

I tried the new patch on Windows. There was one conflict due to a recent change, I think your fork/branches might need an update (probably good practice anyway). It compiled successfully in VC2010 This time without looking at the code or technical aspects, I wanted to get a feel for how it works in practice. Music works well, it faded at the right times and even after finishing a game I could hear the main menu theme again and the next game's music with no glitches. Very nice It's about time our music was heard properly. The "Enable Music" in-game option doesn't work anymore; to disable music for testing other sounds, I set sound.musicgain = 0.0. I notice there are also options for master, action and ambient gain. There should probably be a separate channel for UI effects and maybe others - would voices need their own channel? I don't know how difficult it is to add new channels like that, hopefully not too hard? A few things I noticed about voice / order acknowledgements: The volume is very uneven, some are too loud (Greek females) or quiet (almost all others), but I assume that's easy to tweak Sometimes I can't hear an acknowledgement when I order an off-screen unit to do something. A good way to test this is to play Greeks, select your starting females, move the camera into SoD and then task the selection to move. Most of the time there's silence. You should always hear the acknowledgement independent of location on their own/UI channel. Another test: move the camera well into SoD and find an idle worker (.), you won't hear them answer like they do when already visible on screen. Training/construction sounds: I can hear training and construction completion sounds for any structure in my camera+LOS, whether it was my own structures, an ally or enemy. I think those should be considered a UI sound and only heard by the owner (except maybe in cooperative play we'd share those sounds) but from anywhere on the map. I imagine we could filter out certain classes of sounds (UI) based on the viewing player's ID and the entity's owner or the player who triggered the action - no need to get into diplomacy. Battle sounds: No more out of memory error on the huge combat demo in Atlas! I believe this is related to other out of memory errors people encountered on longer games. Overall I think it's a big improvement over the previous behavior Death growls are too loud and possibly too frequent, in large battles they dominate the rest of the sounds What annoyed me more was zoom and screen position having no effect on volume. I wanted to zoom out from a battle to make it quieter, but unfortunately it remained just as loud on any part of the screen. I think we should take zoom into account IMO battle should be the current volume when zoomed fully in and much quieter at "default zoom", barely audible as you zoom fully out I like Philip's idea of hearing nothing but ambient sound when fully zoomed out. We could make ambient sounds louder when zooming out to accentuate the quieter action sounds. There seems to be an error in calculating which sounds are off-screen, consider this screenshot from the Death Canyon Invasion force map: Clearly there are no units visible on screen, but I can hear most or all of the battle in the upper-right corner. At most I should hear very faint sounds of battle from a few units or more strictly, nothing but ambience. The minimap camera quad is also way off on that map as you can see, maybe both sound and minimap are using a flawed calculation? Another example of buggy sound cutoff, looking at just the minimap (Punjab 3): I marked in green a few camera locations where I heard a group of female's order acknowledgements and marked in red where I couldn't hear them (you can see the females as a group of blue dots in the middle of the revealed LOS). It shouldn't be too hard to reproduce. While acknowledgements are a special case that IMO should be fixed as described above, they are useful for testing cutoff too. Resource gathering seems to be handled differently, it always stopped for off-screen units. It seems the endlessly repeating bird in Atlas has been fixed, scenario designers will be happy about that Minor annoyance: sound really needs to stop immediately when I close the game. Especially with user feedback enabled, music and other sounds can linger for several seconds which is weird considering the game window has disappeared. Even without user feedback enabled, I heard a few bird chirps once after exiting the game. There may be similar issues if the game crashes and continues running in the background but I couldn't trigger that. Needs more testing, if anyone thinks of more good test cases, post them

I don't think the UI supports that yet, but it might be simple to add since it uses the shader/material system? For now, I was just thinking a simple flat colored bar with some alpha and an optional second sprite for borders, highlights, or whatever enhancements we want. IMO the player assignment boxes in game setup already look OK on a similar background:

It's mentioned in the post right above yours Also referred to as "A History of Indian Shipping". It's freely available on Google books.

I tried a different Git client and the problems have seemingly disappeared *fingers crossed* Which is strange, I kept the same local repo. Apparently GitHub for Windows is still experimental and buggy. I'm using TortoiseGit+msysgit and that works much better. I did my first fetch/merge from upstream last night and I'm adding local branches for testing patches

Thanks for catching that

We can dynamically create a sprite with the player color (guiObject.sprite = "colour: r g b a"), it's something we already do for game setup and tech pairs, to name a few. It may not look as snappy as Pureon's image, some tweaking of alpha is needed and possibly a second overlaid sprite to get that shininess.

I seem to recall it was broken when Philip moved and upgraded the Trac server. We decided it was worth breaking, since the website is being totally revamped anyway. There will be a new, better looking revision log on the new website. In the meantime, you can use something like http://trac.wildfiregames.com/log/ps , there's also an RSS feed link at the bottom.

Hello and

As I said in IRC, we should keep the armor/attack icon for that and maybe design a new icon It's not intuitive to hover the portrait, I mean people might do that on accident, but what makes more sense is to put the unit description there with possibly the name again. Whereas if there's a distinct icon people are bound to be curious what it is and try hovering it, then they learn what it is.

I agree with Erik that it's likely a driver problem. I'm not very familiar with Linux driver options so perhaps someone else can give more specific advice. Did you install the Intel GPU driver yourself or was it automatically installed by Lubuntu?

Only the host should need to port forward. Did you make sure to forward port 20595 for UDP (not TCP) on your router, and that it's forwarded to your computer's private IP? Also you may have a software firewall, some OSes like Windows have one built in and that needs to be configured (it can be tricky on Windows 7). Which OS are each of you using? To connect you'll need to use external IPs, unless you're on the same local network. You can find your external IP either in your router's settings or more easily from a website like http://whatismyip.org/.

We plan to have battle detection for UI notification, sounds, and also special battle-themed music as opposed to the normal peace music, which our very talented musician Omri has and will continue to compose So the obvious question is: what is a battle? It shouldn't be a battle every time one of your units is attacked, or it would be almost constant. By contrast, a skirmish might be a scout or female being attacked by an enemy archer or even a lion, of course you want to be notified of that, but in a different way. A battle would be something different, a more significant event for your civilization, and we can make it dramatic. Should it be based solely on the number of units engaged in combat in a given area, if so how many? Is an actual attack required or merely the presence of a large number of enemy units? And when does the "battle" end? I can imagine maybe one or two attacks continuing after the major battle has been decided. The sooner we answer these questions, the sooner it can be implemented

This led me to War in Ancient India which has info about ships and naval warfare including a few images, references a number of sources. There are some potentially useful images in "A History of Indian Shipping", pages 40, 42, 46, 48 and it has some content specifically about shipping in the Mauryan period, in addition to earlier and later periods.

Updated mockup: In this case "Y" is the icon for carried resource with the amount listed to the right. These icons might be too small though Another option is to shrink the portrait more or enlarge the whole panel.

I forgot about carried resources What if we shrank the stats a bit more and put it above them? Or even above the health/stamina bars. That might be making the stats a bit too small, but it seems we need some compromise on stats (they are surely too prominent in the current design).

Maybe not for Part 1, since stamina is not even essential for Part 1, but running and charging are essential. So we might end up with stamina only affecting movement speed, then extending it later if that's something we want to do.

What do people think of this? Civ: an emblem for the unit's civ with tooltip name R: rank insignia D/A/H/P/C: icons for armour/attack stats, basically same as in the above mockups One issue with displaying both generic and specific name is for resources, they are always English names and it would look silly to see "Large Oak Tree (Tree)". For every other unit, I think it's a great idea.

Hmm I don't see a .gitattributes in the 0ad repo, so that's worth investigating. That link says the text settings override the global core.autocrlf, but that setting doesn't do anything for me Nor do the "re-normalizing" instructions or various other hacks I've found mentioned.