Merits / practicalities of releasing a mod (/pack) as a standalone?

[Mr War]

Hi,

Annex got me thinking. We haven't released the scifi pack because it needs the alpha to run etc. But we could package the alpha as a standalone game, which also allows us to limit tile sets and maps to those that fit the scifi factions.

How easy is that to do? And what are the issues - conflicts, user data folders etc?

[titi]

I am also thinking about this now.
The main question is should these kind of "mods" be mods or their own game?

Even when they come with a complete new menu and all this stuff its I think its still possible to make some kind of mod management for them which gives them their complete own environment!

Regarding the standalone game:
Maybe more easy at the moment is the "own game" thing, but this will devide the communities and believe me, there are really NOT enough players for all of them.
What I think(not shure I will check for it when at home) can easily be done at the moment ( MG3.5.2) is that each game has its own userpath. By this you have a directory .megaglest and .annex which are seperate from each other. So you don't see tilesets/techs/maps of MegaGlest anymore in ANNEX and vice versa and all settings are seperate from each other by this.
The only conflict will remain in multiplayer at the moment, where other players don't have the techtree/maps/tilesets and so on and if you allow downloads form a server things start to get mixed up again. But as their sadly are not so many players in multiplayer at the moment, this might not be a too big problem.

As I said, I just started thinking about this problem and I think others too. So if you have good ideas please post them here!

[John.d.h]

Personally, I prefer to have a single game with lots of mods.  It just seems a little silly and redundant to have the same binaries, config files, etc., installed multiple times when you don't need to.  Plus, if I wanted to play a game with the Megapack and then one with Annex, I don't want to have to start the engine again.  I think the main advantage with making a standalone game is that you can customize the menus, loading screens, etc., but could this be easily solved by porting over GAE's addon system (maybe waiting for a stable 0.4 release)?  Mods can customize everything with that.  Plus, downloading mods through an in-game menu is very convenient.  That makes the game very modular and easy to customize.  I think a standalone is most appropriate only if someone wants to make a commercial game.

[Ishmaru]

Why would it divide community?

[Omega]

Having actually distributed a mod as a standalone game for quite a long time, I can fully say with confidence that it's a bad idea. As a quick sum-up: Apocalyptic Dawn was released as a full standalone mod all the way up to version 2.4, and only the (beta) 2.5 and the yet to come version 3 will move away from that path, with the next version to be an addon that will no longer attempt to modify the menu model, etc).

Firstly, it's incredibly difficult to maintain the versions. If your compiled version of the engine had a bug, and a fix was released, you have to reupload everything. You also have to release both Windows and Linux versions, which means you would have to compile for each, and if using installers, you'd also have to compile those installers, and finally upload them online. Since bugs crop up regularly and most new versions of engines include bug fixes of one kind or another, you pretty much have to keep updating this standalone mod, too.

Secondly, if you decide to modify the source code in any way, you must upload and maintain that source code online (as per the GPL), and with each new version of your engine, you'd have to apply the changes and if any conflict with yours, it becomes more of a hassle to fix.

Additionally, the file size becomes much larger. People have to download a large number of extra files (eg, tilesets which may already be included with the engine), causing the download to be larger (not to mention more to upload). Plus, pretty much everyone who tries your mod probably has installed at least one of the Glest engines on their computer, effectively making duplicates. There's no real way to prevent those duplicates, since overwriting the main menu model would take effect for all mods installed, and could cause conflicts.

All in all, standalone releases are a bad idea. Having been one of the first to create and distribute a full conversion mod (a choice I ultimately regretted), I know how tempting it is to replace everything from the logo to showing off an impressively unique main menu model, but it just doesn't work so well with the players, and maintaining it is unnecessarily difficult.

[ultifd]

One of the more obvious things, is that you'll get less downloads. A good example is Annex...would probably have at least 100 more people playing it if it was in the MDC.

What I think(not shure I will check for it when at home) can easily be done at the moment ( MG3.5.2) is that each game has its own userpath. By this you have a directory .megaglest and .annex which are seperate from each other. So you don't see tilesets/techs/maps of MegaGlest anymore in ANNEX and vice versa and all settings are seperate from each other by this.

Yeah, but this should be done by default by the modders...

Great! If only this could have been a bit earlier though...

Well, the one reason not to use it is that if it's an addon, the menu has to be the same as MG or GAE right? I'd imagine the G3D that ishmaru made is pretty cool.  As for GAE, I'll just say non-solid makes a noticeable difference in megabytes...

No, you can still replace the menus models. Just replicate the exact folder structure from the installation directory. Files that appear in a different location in GAE would need a duplicate, but would still work. The MG only files would be ignored by GAE, and vice versa.

Huh?

[Omega]

Great! If only this could have been a bit earlier though...

Well, the one reason not to use it is that if it's an addon, the menu has to be the same as MG or GAE right? I'd imagine the G3D that ishmaru made is pretty cool.  As for GAE, I'll just say non-solid makes a noticeable difference in megabytes...

No, you can still replace the menus models. Just replicate the exact folder structure from the installation directory. Files that appear in a different location in GAE would need a duplicate, but would still work. The MG only files would be ignored by GAE, and vice versa.

Huh?

Let's use an example (of the part I presume you're "huh"ing about). In GAE, scenarios are stored in ./gae/scenarios, while MG stores them in ./scenarios. If you had a GAE only scenario and a MegaGlest only scenario, you could bundle both in the same addon without breaking the opposing engines, because GAE won't use the ./scenarios folder, and vice versa. As well, if the scenario works on both engines (eg, using legacy Glest 3.2.2 code), you could have the exact same scenario (cloned) in both folder locations, and both engines would be able to use it.

[ultifd]

Well, yes, but I just don't understand how that's related to customizing the menu g3d specifically for MG...

[Omega]

You mentioned that "the menu has to be the same as MG or GAE", so I assumed MG has changed something?