Yves

Finally! The Code::Blocks developers are heading towards a new stable realese. C::B is a very promising cross platform open source IDE but the last "stable" release was in 2010 and at least on my system it was anything but stable. The nightly builds were much better but still quite bugy. RC1 discussion thread: http://forums.codebl...ic,17070.0.html I'm going to test it and write some feedback. Maybe someone else here could also test it, we support auto-generation of workspaces for C::B with our build system. Development tools (especially IDE's) on Linux are still a bit behind Microsoft's Visual Studio but hopefully that will change in the next few years.

I don't know how to do it the best way, but I can tell you what I would try first. In StatisticsTracker.js it already counts killed units and buildings. This needs to be extended to count the resources these units cost. The cost for units can be found in the unit templates: 0ad/binaries/data/mods/public/simulation/templates But maybe there's already a function that can be used for that. Probably you can find it in the AI scripts if they try to get enough resources to get a certain building or unit. The question is if it should always count the original cost of units or if it should count what the enemy actually paid for the unit. This could differ if we get technologies that reduce costs for units.

That sounds interesting but needs to be explained a bit more if it should be comprehensible for most people. It just says that some part(CCmpRangeManager) of the game has been made up to 80% faster. It doesn't state how mucht influence it has on the overall performance. Also it's most likely not very accurate because I'm sure there are some parameters that influence the procentual performance benefit like number of units moving, mapsize and hardware. I know it's probably too early to tell that.

Today there are a lot of services on the internet that are designed to follow every step we take and document it in databases for companies to make money with. Google's add-sense is an example. They can track you on every website that uses it and that's a lot. I supspect it works the same way with facebook, twitter and G+ buttons. And this is just what happens if you are not even participating in any of these networks. That's the reason I prefer if it's completely removed from my build rather than just disabled. If it's not very well implemented, "disabled" could mean that the buttons are just invisible but everything else still works. Also I like software that is modular and has some optinal parts that can be disabled. Some other arguments against this feature have been identified earlier in this thread.

I don't like social media integration. So if anyone implements it: please add an option (or better a build-switch) to completely disable it or remove it from the build. I'm sorry if it sounds harsh, but I don't like how social media is present everywhere.

Probably we need it for Windows to deliver securty updates, that's true. What I mean is that we should still support the way how Linux distributions update their software. The auto updating component should be separated in a way that you can distribute the game without it. Also we shouldn't release non-security updates too often and make the game incompatible to users who don't like using auto-updaters.

I don't like auto-updating software for practical, technical and security reasons. Unfortunately It has become a common bad practice on Windows which is probably because Windows doesn't offer any viable alternatives. The security problem is that you have to trust a lot of different companies, organizations and people that they implement secure auto-updating. Remember that even Microsoft has had security issues with Windows Update, so It's not trivial. It's also less user friendly compared to a central package management application because all these tools try to update at a different time and bother you with pop-ups, downloads, installation procedures etc... I have no problem with optional auto-updating for development versions but I'd prefer to avoid it in a final release. Maybe we could implement auto-downloading of mod content such as maps, new civs etc.. but we shouldn't auto-update the engine. Engine updates should't be released too often to maintain compatibility. I know that a lot of people don't care much about security issues and Windows-only users don't know anything better anyway... so there will most likely be some dispute.

You asked... The basic problem is that you can have more than one computer in your local network which all access the internet through your router. If you go to www.whatsmyip.org from different computers in your network you will get the same address (the router's ip-address, not your computer's address). Now if you host the game there's no way how the other player can address your computer. Your computer's IP address is not accessible from the outside (the internet). The router can be addressed but it doesn't know which computer in your local network is meant to receive the information. Port-forwarding just tells the router that all data coming from the outside on port 20595 (in this case) should be sent to your computer. As long as there's no application on your computer that is listening on this port there's no risk. When 0ad or another application listens to that port, an attacker can try to send information the application doesn't expect. This can be a danger if the application doesn't handle it properly. It could crash or even execute code from the attacker in the worst case. However the attacker needs to know which application listens on that port, he must know your IP address, he needs to spend a lot of time finding an error in the application that can be exploited and he needs to attack you just at the time the application is running and in the state he expects.

Look here.

Try building the game without atlas (the scenario editor). 1. Go to 0ad/build/workspaces 2. Run clean-workspaces.sh 3. Run update-workspaces.sh --disable-atlas 4. Go to 0ad/build/workspaces/gcc and run make And post any error messages you get when running make a second time. Btw. you don't get an exe file, that's just the prebuilt binary which is included. If there's no Linux binary it means that the building failed.

If someone expects my help I expect that he at least tries to give me enough information to help. As Erik said it would be a good start to tell which steps you did, what worked and what didn't work. Basically if you run "make" a second time you should only get a few lines like these and nothing else. ==== Building gui (release) ==== ==== Building lowlevel (release) ==== ==== Building mongoose (release) ==== ==== Building pyrogenesis (release) ==== ==== Building mocks_test (release) ==== ==== Building AtlasObject (release) ====

"patch" (not "path") is a program to apply patches in a diff-format to sourcecode or other kind of text files.

Since the plans about migrating to git are becoming more concrete I'm going to wait with that toipic for a while.

It looks like the "patch" command is missing on your system. Probably there's a package called "patch" you can install or maybe this command is contained in another package.

Were there any errors in the previous step? Update-workspaces.sh should create a directory named gcc in build/workspaces if it's successful.

Use Alt+Enter. If that doesn't work, you probably have to change the Alt-Key to something else for "toggle fullscreen": http://trac.wildfire...Manual_Settings

It's currently only possible by editing Player.js javascript file. public/simulation/components/Player.js at line 14: this.maxPop = 300; // maximum population.

You should also make sure that you have have a graphics driver with 3D acceleration installed and running.

How about a simplification: ([Total number of resources ever gathered] + [enemy units and buildings destroyed in resources] * Multiplier) / Divisor Thoughts behind it: Giving a value to certain techs/buildings and units and multiplying them separately by arbitrary values is very difficult to balance and should be avoided. Insead we should just calculate the resources and should not care about how they are spent. Everything is broken down to a unit (resource) and simplified in this approach The multiplier is to give more value to damage given to the enemy compared to gathering resources, training units and building buildings. One example: Imagine a 2vs2 where player A in team 1 defeats both players in team two and player B in team 1 just stays in his base and collects resources. Player A should get the most points in this case. The divisor is to take Mythos_Ruler's input about keeping scores short into account Each type of resources counts the same Enemy buildings and units are counted as resources (costs 20 wood, 30 food = 50 resouces worth)

Yes, maybe the towers are currently a bit overpowered. It's like a machinegun compared to a javeliner trying to hit a sheep from one meter distance

No answer. Does that mean that nobody with access to the svn config has seen this or did you just not have time to look into it/ didn't prioritize it? Or maybe you don't agree? It's definitely nothing urgent, but it would help to save some time if it works as expected.

I've had a problem applying this patch here: http://www.wildfireg...l=&fromsearch=1 It failed to patch "utility_functions.js" without giving me any hint about the reason. I finally found out that it's caused by the CR LF line endings of that file. After converting the file to LF line endings the patch applied successfully. SVN has a property called svn:eol-style which can be set to native. This means you get LF line endings on a Unix System and CR LF on a Windows system and it should work on both systems. This has been done manually for some files in the past: http://trac.wildfire...cket=on&wiki=on I think doing it manually is not optimal because people will forget to do it most of the time. However it's possible to configure subversion to do it automatically for files matching a certain pattern (e.g. ending with .xml, .cpp, .h, .js etc...). http://svnbook.red-b...nced.props.auto Has this ever been considered? I don't have much experience with svn properties, but I think that seems to be a good idea.

http://de.wikipedia.org/wiki/Wikipedia:L%C3%B6schkandidaten/6._Oktober_2010#0_A.D._.28gel.C3.B6scht.29 That discussion doesn't seem to be very objective. Even the tone of some comments is quite insulting. 382295 downloads between July 2010 and now only on sourceforge (not counting all the installations via package management on Linux and downloads from other sources) confutes both the argument that it's unreleased and that it's not widespread enough.

We got OOS Errors very often with alpha 10 when one was playing on Ubuntu and the other on Windows. I checked the difference between the revisions and the OOS errors made sense, but I didn't check if the data is actually a later revision. Whatever, I think it doesn't matter any more because It should be solved now.

Check this thread. The Windows build uses an older revision. http://www.wildfiregames.com/forum/index.php?showtopic=16226&hl=&fromsearch=1