FreeGameDev

Must-have

• bugfixes (as they come)
• new/improved graphics (as they come, 3D and 2D, models by Skorpio and p0ss)
• Keeper speeches (as they come)
• more refactoring (as it comes)
• separating level loading from game start up
• new website header, use forum + MediaWiki RSS add-on as news provider

At risk (one would be good)

• Minimap
• Complex minimap/controll GUI container (with non-rectangular buttons, etc)
• multilanguage support (tinygettext)
• A scripting language

Anything missing?

More refactoring (will make a post about this)

I would also like to see some concept art for the two factions that we have the final list of monsters for, i.e. the Undead and the Humans. We also needs modelers, so a campaign to get those would be nice. I talked to Goat and he said that a Reddit+DeviantArt campaign might be a good idea. He also mentioned that we should have some models from Poss that are not put into the game yet, as far as I remember he made a better version of the game tiles, i.e. those you dig, is that not right? Here is a list of all his posts. I think the tile set will have the best impact on the game, so lets try that out

Seeing as we're short of modelers, we can grab stuff from OpenGameArt and similar projects like DungeonHack, LipsOfSuna, and WARP. Warp even implemented some things Poss made for us that's hosted on OGA before we got round to it, like the archery target.

Lips of Suna has a nice door, when are we going to get round to implementing those?

Also, I haven't seen the Dragon's Bed implemented yet?

Poss made a model for the eatery, when's that planned for?

TheAncientGoat {l Wrote}:Seeing as we're short of modelers, we can grab stuff from OpenGameArt and similar projects like DungeonHack, LipsOfSuna, and WARP. Warp even implemented some things Poss made for us that's hosted on OGA before we got round to it, like the archery target.

Lips of Suna has a nice door, when are we going to get round to implementing those?

Also, I haven't seen the Dragon's Bed implemented yet?

Poss made a model for the eatery, when's that planned for?

I also remember something about a bed for the dragon, but I am not sure it was poss who made it, it is not in any of this posts. Do you have a link for the thread containing it?

Edit: I found it, it was Arcanis who made it.

TheAncientGoat {l Wrote}:Seeing as we're short of modelers, we can grab stuff from OpenGameArt and similar projects like DungeonHack, LipsOfSuna, and WARP. Warp even implemented some things Poss made for us that's hosted on OGA before we got round to it, like the archery target.

I feel the problem right now is not the lack of models available, (though we will need more eventually) but the issue that we don't have anyone who knows blender and how to adjust things and export them to fit in the game.

May i suggest an actual map? not the one that you use for testing, but one in wich you start from 0, lets say in a corner and have to kill 3 other guardians while passing througth rogue creatures in the way

I agree. We should start working against having something playable.

Since neither AI nor network play works, an actual map is not that interesting.

I think we need documentation for the map editor, then we can make a different test map at the same time. But either multiplayer or keeper AI would be greatly appreciated. But I fear that the keeper AI would be hard to make, any ideas?

Also the creature AI needs to be tweaked, when you drop kobolds next to tiles you want to dig out then rarely actually stars digging, they just run away.

Multiplayer is probably easier to get working, (it's half-way working already) than a keeper AI. I think doing some basic scenarios could be done even without, or with a really simple computer AI as well. (Though we first need to implement what's needed for an AI player.)

By the way I have some ideas for multiplayer modes, I will post them when I get home

Edit: they are posted here.

oln {l Wrote}:(Though we first need to implement what's needed for an AI player.)

That's the big problem, putting the needed class and methods system for supporting an AI. Once an dumb (even random) AI will be implemented, it will be easy to improve it progressively.

MCMic {l Wrote}:Once an dumb (even random) AI will be implemented, it will be easy to improve it progressively.

Hmmm that is quite the statement , but I hope your right

svenskmand {l Wrote}:

MCMic {l Wrote}:Once an dumb (even random) AI will be implemented, it will be easy to improve it progressively.

Hmmm that is quite the statement , but I hope your right

Well, at least we'll have a problem where we can try several things and see what it does, experimenting will be harmless, while the work of coding the structure is a one-go thing.

On an other thread some people think the AI should be done in lua (or at least a scripting language), that is a choice that should be done now.

My thought on this is that we could make a base ai player class. Then we could have e.g one subclass that is a c++ ai, and then another subclass can easily be made to be lua-scriptable later.

Yeah, sounds like a good plan!
Once the architecture is done, I'd be happy to give a try in making a simple AI, I enjoy coding AIs! (that does not mean I'm good at it ^^)

Oln and I have talked about a different thread model which will make the game optimally parallel (i.e, we can use an unlimited number of cores, if available). We can put that on the list also, but it will not be done for 0.4.9, but the initial footwork can be started, i.e. rewriting the way we store the game state. See this post for the idea.

Here is my addition to the original post.
Must have

• Working team colors for tiles. Currently you cannot tell where you can drop you creatures or imps as all team colors are the same. I think this is THE most important bug to fix in 0.4.9

Nice to have

• A working multiplayer mode say the "Take what you can and give nothing back"-mode of this thread.
• Health/level flowers or some other indication of this.

So what can we cross out here and should we make a 0.4.9 release soon?
Must have

• bugfixes (as they come)
• new/improved graphics (as they come, 3D and 2D, models by p0ss for example)
• Keeper speeches (as they come)
• more refactoring (as it comes)
• separating level loading from game start up
• new website header, use forum + MediaWiki RSS add-on as news provider
• Working team colors for tiles. Currently you cannot tell where you can drop you creatures or imps as all team colors are the same. I think this is THE most important bug to fix in 0.4.9

At risk (one would be good)

• Minimap
• Complex minimap/controll GUI container (with non-rectangular buttons, etc)
• multilanguage support (tinygettext)
• A working multiplayer mode say the "Take what you can and give nothing back"-mode of this thread.
• Health/level flowers or some other indication of this.

The shader/graphics system is in a need of some work yes, both team colours and other systems. I agree this is something we must fix for 0.4.9.
Going to have a look at this once I get back home. (If someone can get their head around the shader system, team colours should be really simple to implement.)

Keeper speeches should be very simple to implement in-game (might already be in there), so just waiting for venn to have the time to record them.
(We need a config-system for specifying filenames at some point, but this can be hard-coded for now.)

Refactoring we do as we go, I don't know if we want to tie it too much to releases.

Separating level-loading from start-up shouldn't take too long. (I think stefan was working on this at some point.)

The website is separate from the game itself, but I agree that we need to do some improvements before a new release. (Or atleast the release that has either multiplayer or AI sufficiently working to make the game playable.)

--

I don't think we should concentrate too much on the multilanguage support at this stage, though it should be in mind when working on other things. (Might want to start wrapping text strings and make sure visible text is specified in unicode etc. )

I think multiplayer is partly working, but some things are not sent over the network yet, so some extra network-messages has to be added. (I know that room building and traps are missing at least.)

Don't know if we should aim for health flowers in this release. We do however need some design suggestions on these.


---
Besides the graphics I think our main focus should be on having something playable for releases after this one. (Once we have either working multiplayer or AI, we should bump the version number to 0.5 to indicate this.)
I started working on making a framework for keeper AI's, and now that we have angelscript added to the game, that can be used to script the AI once the class the AI is to call through is done.

I only added some todos about the separating of the leven loading. This is something I wanted to do, when there is a good class that holds all the loading/saving stuff (currently they are global functions), maybe directly in the GameMap or the ResourceManager. The gameMap should also be changed somehow as the instanciated object of the gameMap is currently stored globally. I couldn't make it a singleton however because it conflicts with the ProtectedObject class. I'll need to give this a more indepth look.
A way to possibly solve this problem could be to store the gameMap in the GameState(any better suggestions?) class and construct and destruct the map every time a new level loads instead of just "overwriting" the one global instance.
But since this would a rather big change (after all it's affecting probably the most important object of the game) I'd like to hear some opinions first.

Yeah, bugs, refactoring, new sounds, etc should be done when they come, this should not be a "fixed" goal.

About the scripting:
I currently don't have to much time to work on the code (again ), at least not for bigger changes, but locally on my harddrive the AngelScript wrapper is slowly growing bigger. Up until now I mostly read their docs and tutorials and intensly documented the script wrapper code for learning purposes, because the whole scripting world is new to me. But I think in around two or three weeks or so I'll have some more time again, then I'll do some bigger changes at once (my plan is to have the script system working by the latest of end of August).

About releasing: Do when you like, I'll just code my way.

StefanP.MUC {l Wrote}:I only added some todos about the separating of the leven loading. This is something I wanted to do, when there is a good class that holds all the loading/saving stuff (currently they are global functions), maybe directly in the GameMap or the ResourceManager. The gameMap should also be changed somehow as the instanciated object of the gameMap is currently stored globally. I couldn't make it a singleton however because it conflicts with the ProtectedObject class. I'll need to give this a more indepth look.
A way to possibly solve this problem could be to store the gameMap in the GameState(any better suggestions?) class and construct and destruct the map every time a new level loads instead of just "overwriting" the one global instance.
But since this would a rather big change (after all it's affecting probably the most important object of the game) I'd like to hear some opinions first.

I agree with making the game map an instanceiated class. I don't know where it would be best to store it either.
Also, the GameState class isn't really finished yet. Currently it's just wrapping the call used to check if the game was single-/multiplayer, server/client, game/editor. (By checking if sockets exist, and who the local player is.) The state should be set when changed instead. (We don't want the socket objects being global.)

Ok sounds good. I will try and make a deb source package tomorrow if I am successful I will make a ppa for OD

Nice.

I might have an extra machine when I get back from holidays, so I could set it up to do some daily builds. (I guess windows is sufficient if we can do linux-builds in a PPA?)