Messages

It's not something you plan on supporting though, right?

I could use Mandate in its current form (if it works over mp), but it depends what's there right now.

If you really want multiplayer MegaGlest is a better choice than Mandate or GAE. It has pretty solid cross platform stable multiplayer.

Well in theory Mandate has the same multiplayer capability as GAE since its a fork. Although I haven't actually tested it to see if I broke anything.

Most of this is available in my fork of GAE. While the game I am working on is not finished, and has been further delayed by my plan to hit 1trillion isk a year in income in EVE Online, individual features are functioning properly.

As far as "squads" go, I don't use the GAE pet behavior. Its quite limited. Rather you will specify behaviors for units in a special AI file. For instance the defend behavior causes units to, assuming it has the highest priority, attack units which are attacking their owner. The rest behavior causes units to load themselves into their owner, where they then heal, which although not appropriate for squads of humans, could work for a carrier, or a spaceship with drones and so forth. There is not currently a follow behavior to be paired with the defend behavior or a behavior which causes units to attack their owners target, but this would be trivial to implement. Note that these are not commands, and you cannot control a unit that follows AI routines. Well, you can because I haven't added the flag to block hide commands and block right click orders, but that is a small change I am waiting to make after I finish some stuff.

Units can be assigned owners, based on a flag in the produce command, and owners have specific caps for how much of each unit type they control. Automatic generation of units and manual produce commands will not function if the control limit is capped. Note also that units can be set to automatically create other units, which they can own and to which their cap applies. Spawn of controlled units is based on a timer which you can set.

IIRC Hailstone is possibly working on multiplayer for GAE, I don't think there are any plans to add other features in the mean time.

Finally did some work after a month plus of slacking. Not a whole lot. Added wizards, fixed a bug.

Video:
http://www.youtube.com/watch?v=lIUUJCtSKpI&feature=youtu.be

This is just two wizards attacking some enemies. They did fast but you can sorta see them using their attacks smartly.

They look for the most damaging attack to their selected target, check for cooldowns, then mana, then attack. I might try using units with much higher health so its more obvious whats happening.

Not sure how many times I have to say it. "The New Animations Will Not Effect ANYTHING at all"
all they do is make the game look beter. NOTHING ELSE.
...This has nothing to do with the FPS at all it will effect nothing...

:o If you you really believe in what you wrote, then maybe you shouldn't to do that.  :-\
A lot of things have an impact on the amount of FPS and on 101% everything related with animation.

...I would also like to mention the obvious thing, that the acceleration of animation results in greater demand for "PC resources" so you have to do it in moderation...

...but maybe if you can write it with larger font and more red then I'll believe it.  :P

For me it is completely incomprehensible with whom you are trying to argue.
Did so far someone wrote that it is unnecessary and you shouldn't to do that?  :confused:
Until now, you could did prepare a patch and paste it here, which allowing it to tests for others,
and probably most of testers will write that you have done a good job.  :thumbup:

At the moment I have a feeling that discussion in this thread will continue until at least the appearance of stable version 3.18.0  :-X

It seems like you are the one he is talking to. He isn't "accelerating" the animation, so your comment is irrelevant. Realistic animations are not so much about the number of frames or speeding them up as creating movement that properly follows real human movement. He is saying he will make animations where the body parts move in a more correct manner, which will have no effect on performance since each animation is the same number of frames and has an almost identical amount of data to process as the current animations.

there is almost no video game ever that shows that info. if you are so new to a content pack that you don't know what stuff does, you probably can't make effective choices based on that info anyways.

It's not really a feature request so much as a question of whether it can be done.

How does Mandate stack up to GAE and MG? What's is its goal as far as development?

This is certainly possible, depending on how efficiently and cleverly its implemented and whether your computer is strong enough to handle it once it works.

MegaGlest is, as far as I am aware, focused primarily on a stable cross platform multiplayer experience.

GAE is designed to function as a sort of engine for a broad range of standard RTS games. It may be getting decent multiplayer soon.

Mandate is designed to support exploratory strategy games rather than RTS games. It can or will support various kinds of ai, automated unit control and so forth. As well as a very broad range of other nonstandard features. It should allow for RTS games of many base templates, such as Battlecry like games, Majesty likes, starcraft like games and more. It eventually will also support MOBAs, city builders, and such. It may also support a rough form of RPG.

It also should allow for various features from a broad range of strategy genres such as King of Dragon Pass style events/stories and other strange things. Faction anchored AI is also possible where one faction can play like Majesty, one like an RTS, one like an RPG, and one like other things. Such as being able to do as you asked, allow for a sort of commander type scenario where you create general units who can control a wide variety of things cleverly.

On the downside is that I have no plans for a stable multiplayer, indeed given the features I plan for and the kind of games I like its quite impossible to maintain stable multiplayer. It also lacks precise focus I suppose as well as me being, as far as I know, the least traditionally educated or experienced developer. I am also somewhat unstable development wise. I fail to work consistently and I flit around from feature to feature leaving several unfinished or unpolished.

Mandate has this. Well its only about half implemented. I am focusing on the individual agent AI for my current project mostly.

Its unlikely that either GAE or MG would add this. They are more focused on being traditional RTS games.

I implemented some code in Unit::kill() where if a unit dies any units it owns now owns itself. That will solve the error and has to happen for every unit anyways. Later I'll add in some so that units who really need to have a proper owner will seek the most viable new owner or if there isn't one behave accordingly till one exists.

The dependencies are whatever GAE has. I didn't add anything.

At first I had no idea how you could get an error, but what appears to have happened is that you killed the owner of the unit, in this case the structure that auto spawned it, before you killed the unit, this making its owner pointer invalid so that on death it tried to access a garbage pointer to change some data.

I'm not really sure how to handle that quite yet because ownership is used for different things sometimes. Presumably I'll have to clear the owner pointer somehow. Then assign a new owner based on what that unit type uses ownership for. Autogenned units will want to just reset their owner to themselves, while units using the new AI will want to set to themselves until they can find a valid new owner. And service units like resource and item haulers will want to kill themselves.

Pretty sure all Glest derivatives are licensed under the GPL.

Despite of the progress indicator turning into negative percentages it does load and initialize fine. There are selection issues (due to relying on GlSelectBuf, I assume) with structures and units but I think that's a known bug.

I played around for like 15 minutes during which all of my units and structures got destroyed and all I had left was this unit which looks hero like, a guy with a sword and cape who has a lot of points on everything and can create structures (and workers?). So I tinkered for a while, destroyed all the enemy buildings until I finally attacked one of the enemy daemons and ran into:

Code Select Expand


Crash
Version: Advanced Engine 0.3.93
Built: Thu Oct 17 19:29:00 CEST 2013
Time: Fri Oct 18 11:12:37 2013
Description: SIGSEGV: unrecognized si_code
Address: 0
Backtrace:
./build/source/game/glestadv(_ZN6Shared8Platform24PlatformExceptionHandler7handlerEiP9siginfo_tPv+0x1aa) [0x7c796a]
/lib/x86_64-linux-gnu/libpthread.so.0(+0xfbd0) [0x7f55c970abd0]
./build/source/game/glestadv(_ZN5Glest8Entities4Unit4killEv+0x2b6) [0x5a4ce6]
./build/source/game/glestadv(_ZN5Glest3Sim5World11updateUnitsEPKNS_8Entities7FactionE+0x44) [0x78d614]
./build/source/game/glestadv(_ZN5Glest3Sim5World12processFrameEv+0x8a) [0x79360a]
./build/source/game/glestadv(_ZN5Glest3Sim19SimulationInterface11updateWorldEv+0x46f) [0x7a164f]
./build/source/game/glestadv(_ZN5Glest3Gui9GameState6updateEv+0xbd) [0x6011dd]
./build/source/game/glestadv(_ZN5Glest4Main7Program4loopEv+0x20e) [0x676c9e]
./build/source/game/glestadv(_ZN5Glest4Main9glestMainEiPPc+0x2b6) [0x673d86]
./build/source/game/glestadv(main+0x2d) [0x50fb3d]
/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf5) [0x7f55c78e8ea5]
./build/source/game/glestadv() [0x512f21]


Pretty sure the Ruler can't create workers or units. However you can create units at all times, assuming you have the resources, from the faction build menu. But only particular ones that have the "faction" tag in XML.

As far as the error, can you get the debug data? You should be able to get line numbers. I can see its an error in Unit::kill() but not what in that function causes the error. The crash file in glestadv should at least have line numbers if your IDE doesn't.

My terminal window does not provide a green rebuild button. ;-) I was running the same build I was previously reporting about then. Sorry - I should just have rebuilt and tried that.

Which I just did (e57dfd920d3b303f24d1b370e0113066d9c39c58), and the character creator menu came up fine and loaded much more swiftly than before. So that's looking good.

Okay, so its mostly working without problems.

Can you try and load the scenario? Morning it's called. You can't actually do much, its just what I use for testing purposes but it would be good to know that it runs properly on Linux.

You need to rebuild. Your version is setting the ListBox to faction 3 which is now faction 2 and there is no 3, because I removed Humans. Sorry. How would you even start the program without rebuilding? Don't you have to hit the green debug button which builds before running the program?

I apparently edited rather than quoted your post and this forum doesn't have history so I just deleted that and made this post.

I removed the Human faction from the repo so that should theoretically clear up all the errors.

Based on my test there should no longer be any errors opening the Character Creator, but it isn't completely polished and would probably be unusable for you just because certain parts aren't enabled and there isn't currently much data to work with for the RPG part anyways.

Thanks for explaining, Ygg, and for commenting out problematic code for now, MoLAoS.

So I'll ignore the override controls warnings but I'm afraid I still run into the temporary error:

Code Select Expand


source/game/world/world.cpp:797:140: error: taking address of temporary [-fpermissive]
source/game/world/world.cpp:798:133: error: taking address of temporary [-fpermissive]
make[2]: *** [source/game/CMakeFiles/glestadv.dir/world/world.cpp.o] Error 1
make[1]: *** [source/game/CMakeFiles/glestadv.dir/all] Error 2
make: *** [all] Error 2


So basically that's unchanged?

Full log

I actually forgot to push. Sorry. I just did it now.

I went ahead commented out the code causing a problem, I'll need to redo a bunch of stuff later once I figure out how I want things to work precisely for BonusPowers.

Given Ygg's response and that, all the issues you posted about should be resolved.

I went ahead and fixed the stupid pointer to temporary thing in a way that shouldn't bug your compiler.

As far as the feature of C++11, I am using the 2008 version of VC++ so I am not sure how that works. Did MS update it to have that feature?

I'll push the current version to the repo as soon as it finishes compiling.

Well that's a shame,
I think people would buy a great quality Glest based game, if priced right.

Well, its technically possible to sell a game using an open source engine by using commercially licensed art. But I'm still pretty sure you can't sell it on most distribution channels and you have to send out the source or make it easily available to anyone who buys your game. Mainly the fact that you can't use Steam or Impulse will cripple sales.

I think maybe Desura can sell open source based products?

Well, that doesn't seem like a problem to me, VC++ didn't care at all and the temporary is only used for two things and all of them are in the same scope so it shouldn't cause any trouble, I'll look at some simple ways to fix it but maybe you have a flag set that catches things which aren't really a problem?

Okay, it looks like I fixed it and I pushed that updated code through, although I can't be sure since I don't have a Linux compiler. According to my research I didn't actually have a problem but it is probably easier to just change the code than argue with any Linux users who try to compile the program. Basically its just a policy issue where some compilers are whinier than others just in case the programmer is stupid and doesn't know what they are doing.

warning: override controls (override/final) only available with -std=c++11 or -std=gnu++11 [enabled by default]

These warnings appear to be based on some flag? And they are all in files I don't recall changing. I'll google it.

I pushed my latest files to Github. I'm not sure why a version that produced that error got sent. Should compile now.

I recall that a lot of people who wanted to do for profit Glest projects were modders rather than programmers, so they were simply incapable of realizing their ideas due to engine limitations. Also the art quality of Glest is simply not comparable to a commercial game.

RTS games are considered one of the most difficult kinds to create as an indie by game developers even if you already have a working engine.

Glest derivatives also lack any sort of quality tool. No unit creator, the map creator is not at all comparable to most commercial map creators, the graphics pipeline is not the greatest and so forth.

Steam doesn't even allow games written with anything open source on their platform, and no publisher means no access to brick and mortar stores. There is pretty much no market for a Glest based game.

Spring and Wesnoth are similarly not profitable either for themselves or for people building projects off them. Hell 0AD couldn't even break 50k on their crowdfunding campaign. Of course they ran it on IndieGoGo which was dumb.

It was confusing at first but I saw that one forum I was the last post and assumed that icon meant no new posts. Not sure that will work for anyone else?

I get that its a theme thing but its just not very clear.

Oh, I forgot that MegaGlest didn't abstract all that stuff. In GAE the enhancement class is the only one that says attack-strength/attack-strenght and upgrades and skills and units have an enhancement member inside them. So you only need the one change.

What are the .pl files for? I don't recognize those from GAE.

AFAIK, only the one XML defining thing has the misspelling. The others were probably fixed already because only that single instance actually affected modders.

As far as changing mods, that's a simple search-replace run on each thing, so one for source, assuming you actually still need to change something there, one for megapack, and then one for each other mod you are fixing. Shouldn't take more than 20 minutes.