Spring RTS Engine

After the next release I want to focus on cleaning things up internally and exporting anything non essential into external modules that the user can pick and choose and that can be updated independently and replaced.

However I mainly want a pluggable extension system so that other developers can extend the lobby without having to develop AFLobby as a whole.

So before I finalize anything and go implementing things, what do people 'as developers not users' want?

Currently My main idea is to allow the initial splash screen to have buttons added to or be replaced, and panels added to the tabpanes in the settings page, the main logged in view, the battle window, the lower half of the main view splitpane, and the player info window.

Extensions should imo hold at least the following info:

- An icon, format unspecified, the plugin would return a java Image object with it so it could be generated from an svg or a gif or procedurally generated etc..
- A name, short name
- html formatted description in long and short versions
- author and group/site/clan
- version number (float)
- version string (XE9.79 haha, NTai)
- development or stable flag
- internal name (abc123-_ no spaces newlines commas etc)
- dependencies (internal names and their versions)

Well more to the point, is anybody bothered about lobby widgets and firefox style extensions?

It'd be a good way of implementing extra features without having to deal with aflobbies internals or having to fully integrate anything.

Well, I'm not entirely a "User" and I can code to a certain extent so I'll put myself in the "allowed to post" category.

From my perspective as long as the core is solid and designed in a way that extensions will be unable to break it in a permanent way (aka, you can always choose to disable them) I like the idea.

I'm not sure what kinds of things I would add but I can see the potential.

well a disable enable setup is definately required, its all about user choice at the end of the day, this way someone can spend an hour on a feature and deliver it to the forums there and then with immediate results without having to mess around with svn or trying to get a hold of me or waiting/syncing with AFLobby releases.

And an eventual SDK in the form of a drag and drop netbeans project compelte with examples would be available to make it even easier for a java rookie to get somethign up and running.

However I'm not sure right now that if I made an amazing interface the likes of which nobody had ever designed before, that anybody here would bother to use it, as nobodies posted much =(

as a user...wil you release a version that will download maps/mods outomaticly? and allow to update mods automaticly as well?
plus the lobby should allow to make ladder games and show all laddr iptions including rankings in the lobby itself.plus it should emphesize all important stuff like help and tutorials even more than other things so that new players will be almost forced to go read help or the tutorial..

Ladder itnegration was added to AFLobby a few weeks ago and is sitting their untested with no work being done on it because nobody would test it.

However this si about implementing plugins not feature requests. Those requests belong in my other thread.

I'd say don't bother now unless making an interface will take you about a day. You'll need users who can and will make extensions and you'll need users who want them; currently there are issues with both groups, and that's not only your particular lobby problem, but the whole spring.

Well actually, I can make it in a day.

I can also figure out in 2 or 3 days howto let people type them in directly into aflobby and have them compiled and ran there and then turning it into a scripting language of sorts.

I have had various people ask me things about developing various features using aflobby, asking about wether I'd mind it, what things would be involved and so on, and there are various features that're small enough that they could be done in an hour or even 5-10 minutes fi the correct tools are there.

And most of all, users will continue to stay around tasclient unless they have a good reason to switch even if the competition is always one step ahead. Innovation is necessary.