Project 10 in 2010 Brainstorming

[zwzsg]

1)CLIENTS
TASCLIENT [...] We need to eliminate This application from the pool

That's almost funny because it sounds like you want TASClient's head just because you know it's so good it leaves no room for your own pet project client. TASClient is just the best client right now. Other clients lacks the handy features TASClient has, and TASClient GUI is just the cosiest. I tried other clients, they were ugly and impracticable, they made me felt agressed by ankward interface and being forced into their dev's skewed view instead of helping me do what I expect clients to do.

3) LUA
Ive seen some of the lua scripts people use, and some of them look like a 3rd grader wrote it.

That's the downside of having such an easy scripting language. Even if I dislike all that Lua at first, I must admit it was very succesful into making Spring much more customisable, not just on the paper, but in number of actual realisations. Having every 3rd garder able to implement his own feature is way more productive than having to wait for the two or three overburdened experienced C++ coder to agree spending time on other people's ideas and then hard code them in the engine.

7)BUNDLE PACKAGE
A bulk download is completely neccassary, it will probably be in the form of an autodownloader, but with no options. options cause diversity and thats bad in our case. It should come with standard maps and mods, etc This MUST BE STANDARDIZED.

I liked the previous downloader with bunch of maps packs and mods. I disagree about "no options", I think options should still be there, just, they should all be ticked by default, except for maybe the hugest map packs.

Maybe make it impossible to run spring.exe directly, let it need at least one command-line parameter to prevent noobs from installing spring, manually running spring.exe and baaawwwwing "Where the fuck am I??"

I much prefer my idea of letting me customise spring.exe menu with a widget.

[Neddie]

I'm in considerable pain and have been unable to sit up at length for days, so I apologize if I am both sharp and brief.

We've already met this unique player goal, I estimate roughly 1400 different actual people play each month. I think analysis of new account registration will reveal that our player base continues to grow monthly, with between ten and twenty people actually sticking around for repeat play - I suspect the areas with greatest growth are Northern Europe, Eastern Europe and the United States, but there have been recent gains in the Fertile Crescent and Coastal Asia.

The development side of Spring is a matter of merit and time-availability, not a matter of popularity. Our project lead is our project lead because he is first supremely capable and second available. He happens to also be affable and intelligent, so much the better for us. The election of a figurehead is first unnecessary and second unproductive - we have senior moderator administrators and senior administrator developers who represent the community and do it well. We have leaders - Licho, Aegis, Tobi, SwiftSpear, LordMatt. I understand that SwiftSpear and LordMatt are less active than they once were, but while a number of people have asked me to take on their informal responsibilities, as yet there are no plans to change this casual leadership.

Far be it from me to say our moderation is flawless, but it is not as misguided and ineffective as some of the participants in this thread seem to believe. I am available as a moderator roughly six to fourteen hours a day normally, covering a number of undermoderated timezones, though one day I hope to scale that down to three to seven. Few people have complained about my performance, nor would many complain of others such as Quantum and Lion_Heart. While some of our moderators who have served us well have stepped down, and others have scaled back their activity, I am currently working with Tobi/Quantum/Aegis and others on some reforms to streamline our present moderation process. I am also reviewing possible candidates to fill our less moderated time zones, and I implore anybody who has a candidate for moderator or an idea for moderation to bring it up with me in a private message on this forum. I will not be able to respond to many of them but I do promise to read them all.

I find the side notes on GUI in this thread somewhat misguided, and would like to address them as well. We have a highly functional default GUI, we do not need a visually improved default, but rather projects need to take on the task of creating their own custom GUI. It can be done, Star Wars: Imperial Winter has an atmospheric custom GUI, while I am designing a comprehensive GUI for Spring 1944. Even some mods, I refer obliquely to Complete Annihilation, have custom GUI in development. Lua is not a particularly easy tool for creating a GUI, but it does have the power necessary and, given some open examples and documentation, is comparable in resultant function if not wasted time to implementing some unnecessary xml abstraction. GUI is one of the most difficult elements of a modern program, as evidenced by the abundance of obtuse and unattractive GUI in open source as a small world - after years of development, programs like TASClient, The GIMP and Sauerbraten all have substandard GUI.

A bundle package complete with frontend is, once again, the concern of the games and mods, rather than the engine. Spring is an engine, attempting to offer comprehensive downloads of an engine with a variety of games is counterproductive. I suggest game and mod developers look at Spring: 1944 and Kernel Panic for simple bulk installers - we are presently refining our 1944 offering, and I believe zwzsg is looking into porting his frontend to linux. It would be nice to host bulk installers of fully free games on this site as options on the download page, along with the present Lobby + Engine installer.

While I feel it is often wasteful in some ways, I do agree that multiple clients - and to a lesser extent, autohosts - are, in the long-term, good for the community. They foster an environment of competition and thus self-improvement. I presently provide graphical elements for Winter, and am willing to help with other projects that require icons.

I am fading fast, so I will say little on the subject of Marketing and Advertisement. In September I will be able to take on market research and project advertisement once more, though I will not advertise the engine, but rather some games - 1944, Kernel Panic, Gundam RTS. Going out and talking to people remains the most efficient way to develop the community for most. The players hailing from Finland, Great Britain and Italy - all fairly well represented nations - achieved their lasting growth primarily though person to person transmission. If you plan to advertise on a wider scale or through another form of communication, I advise you to advertise particular games on the engine, rather than the engine itself - it is roughly three times as likely to attract somebody into a download. I would also not stress that the product is free - in many developed nations, "free" applied to a product implies a lack of quality, if you only mention the free status as necessary you should pull roughly twice as many people as you would if you opened with "Free RTS Spring". I also request you refer to the engine as Spring, RTS Spring or Spring RTS - this scans better with people and they're about twenty five percent more likely to respond to a post or engage in discussion.

I would like to discuss all of this with you further, but I need to lie down.

Addendum: Tutorials and missions, once again, the responsibility of the game and mod developers. Some projects are working on campaigns but they will be some time in coming. I can help a little KaiserJ, mostly in writing and voice acting, and I recommend poking [XHC]DavetheBrave on lobby - he might be willing to help with a CA campaign. As for the wiki, I'm looking for some people willing to take on that task - I do not wish to over commit myself.

[zwzsg]

zwzsg is looking into porting his frontend to linux.

Not exactly. I'm looking into a dev coding me the one command that I lack for my widget port of my frontend to work again. Having my front end ported to linux would be a waste of precious coder time. Though if someone feels like to, I'd be happy. Until I become sad from finding out he disappeared leaving his links down, like the last one.

[Doleo]

Even with all the criticism, I think this is a good thing to work towards. I just hope it gets acted out upon and not just talked about. Good luck.

[AF]

It appears my epic troll and permaban had its influences, and now that I'm back, I have certain recent advances to share (last time I promise)

Singlesign on:

OpenID

Frontend:

Until someone addresses the problem of the intro movie and shiny menu in full screen we're not going anywhere fast. A major UK PC gaming magazine mentioned starwars and mentioned this problem as a block that if you can get past gives you starwars RTS. Did that not ring alarm bells?

It doesn't need an integrated lobby either. Make an executable called frontend.exe that does single player and options. Give it a multiplayer button that launches the preferred lobby and make it fullscreen and UI shiny, then make it launch spring.exe with a script under missions/singleplayer.

glou thingy:

What is this? I keep seeing cryptic discussion about it and links to cryptic pages. I want to see a nice plain and simple explanation of the protocol. If this cannot be done then all hope is lost.

Prizes and bounties:

Bad idea, how do I know? Experience

I managed to weasel $30 out of caydr (hooray) for implementing features when he put up the first bounty. It caused a tonne of controversy and I myself quit the community for a short period not long after. There was a big where's AF thread and arguing between developers and community members.

The developers don't like the idea of bounties, since they've done all their work for free, and anyone who does come along working for bounties sort of has a taint to them. In the end the bounty caused more harm than good.

TASClient and multiple lobbies:

We know tasclient is fail, it always has been, satirik has been prolonging its life though. That he's managed it is a testament after speaking with the likes of altaric and other forkers of the project who've complained at the instability and hodge podge that si the tasclient code.

When tasclient finally dies it'll be a sad day but in the meantime having a healthy lobby development ecosystem demands that more than one lobby client exists. The one downside to this is that no lobby developer should modify the server protocol themselves unless it has been discussed beforehand since it affects other projects. Only the likes of aegis whose in charge of server development should be able to do these things and even then after consulting with lobby devs and being able to justify changes.

Moderation:

Shhhh!!!!!!!!

Advertisement

This is the final step after everything else has been finished, but people keep falling into the trap LETS ADVERTISE SPRING even though this would not help spring, it would actually DAMAGE spring. You don't play spring, you play BA, you play Kernel panic, etc etc, so don't advertise spring, advertise BA, don't advertise spring advertise kernel panic!! etc etc.. As long as people are advertising spring as a game not an engine and showing all these games thrown together, they are reducing spring value to that of a 500 games for a $1 cd, theyre confusing the end user as to what they're actually getting, and they're setting spring up for failure, and horrific retention rates.

Installer

A downloader is not an installer. If I download an installer, everything that installer needs should be inside that file. The only thing it should ask the internet for is updates and patches that may have been released since the installer was built.

I still think an official BA installer is a good idea. Most of the complaints about maps etc are coming from BA players, and its my opinion that the official installer isn't the place for this. These maps are best for BA and belong in a BA installer controlled by BA players, they know what's best for their own game, and by doing this they are helping the spring community as a whole by clearing up the greatest conundrum of this community which is How do I play TA?

Now really you guys make such a huge fuss out of nothing, but your never willing to actually address these points, so your doomed to stay standing still, by only working on the bits you don't need to work on.

[SinbadEV]

I swears, if you could figure out something that would fix my in-operative motivation control center, I would be totally on all these problems... wait, no, if that part of me worked I'd probably be a Type-A personality or at the very least unwilling to help any of you unless it directly benefited me.

Sorry, everything about Spring was doomed from the start and it is a testament to generous people "working on the bits they care about but don't really matter" that any of us have anything to play with.

Also AF, Glou is a pipedream... basically, all those ideas you had for AFLobby back in the day except with less gumption to acctually accomplish anything.

[el_matarife]

glou thingy:

What is this? I keep seeing cryptic discussion about it and links to cryptic pages. I want to see a nice plain and simple explanation of the protocol. If this cannot be done then all hope is lost.

Here's the draft protocol: http://code.inrain.org/browse/glou/standards/main/files
If you've got comments or concerns, there's people hanging out at #Glou on Freenode all the time who can answer most of them. You can also take them to the mailing list available here: https://lists.sourceforge.net/lists/listinfo/glou-devel

Contributions welcome, there's work in progress on the two MIT licensed libraries designed to make implementing GLOU in other programs trivial, plus a reference lobby client that's being built in Java but isn't in the repository yet.

There's also a wiki that you can read to get a better idea of what the goals are: http://sourceforge.net/apps/mediawiki/g ... =Main_Page

[AF]

Is MIT GPL/LGPL compatible? And don't link to a file in a repository! V.Unproffesional looking!!! Me wants a pretyt document

*goes to look*

[el_matarife]

Also AF, Glou is a pipedream... basically, all those ideas you had for AFLobby back in the day except with less gumption to acctually accomplish anything.

Yep, four months with a bunch of part time coders and all they've managed to accomplish is WRITE AN ENTIRE DRAFT PROTOCOL and BEGIN IMPLEMENTING IT IN LIBRARIES THAT YOU CAN PLUG IN TO YOUR LOBBY. You guys really need to get a little more patience, if they hadn't started by defining a protocol, GLOU would have ended up as an unworkable mess. Prior Planning and Preparation Prevents Piss Poor Performance.

[el_matarife]

Is MIT GPL/LGPL compatible? And don't link to a file in a repository! V.Unproffesional looking!!! Me wants a pretyt document

*goes to look*

Well, if we put it in the Wiki we might have people making changes to it which isn't a great idea. MIT is basically the BSD license saying "Do whatever you want with this code including shove it in a proprietary program just don't sue us if it doesn't work" http://en.wikipedia.org/wiki/MIT_License

[zwzsg]

Frontend:

Until someone addresses the problem of the intro movie and shiny menu in full screen we're not going anywhere fast.

I can make full screen menu, and have a little intro show of AI battle, with my Lua things. Only issue is that I can't have a multiplayer button...

Make an executable called frontend.exe that does single player and options. Give it a multiplayer button that launches the preferred lobby and make it fullscreen and UI shiny, then make it launch spring.exe with a script under missions/singleplayer.

Satirik has done that, it just need to be used more.

glou thingy:

What is this? I keep seeing cryptic discussion about it and links to cryptic pages. I want to see a nice plain and simple explanation of the protocol. If this cannot be done then all hope is lost.

I think it's that project about an unified lobby for all open source games. Not sure.

We know tasclient is fail

TASClient is win, you're just jealous. No other client combine as many feature with such a convenient GUI.

having a healthy lobby development ecosystem demands that more than one lobby client exists

Not sure where you got this idea that we absolutly need several competing lobbies. One good lobby would be preferable to five lobbies each with its cons.

Installer

The question with installer is, does Tobi allow us to link to custom insallers in Spring's website download wiki page.

[AF]

Before AFLobby and SpringLobby there was little if any innovation in tasclient and lobby as a whole outside of autohosts and bots.

Once AFlobby came on the scene and later spring lobby, things quickly got moving. Numerous features where nabbed from AFlobby and new features where added. Bugs started getting squashed and server development moved from the usual standstill.

Competition drives things forward.

[el_matarife]

Frontend:

Until someone addresses the problem of the intro movie and shiny menu in full screen we're not going anywhere fast.

I can make full screen menu, and have a little intro show of AI battle, with my Lua things. Only issue is that I can't have a multiplayer button...

I suggest just having the multiplayer button quit the game and start a lobby, like it does in Supreme Commander.

[Auswaschbar]

Frontend:

Until someone addresses the problem of the intro movie and shiny menu in full screen we're not going anywhere fast.

I can make full screen menu, and have a little intro show of AI battle, with my Lua things. Only issue is that I can't have a multiplayer button...

I suggest just having the multiplayer button quit the game and start a lobby, like it does in Supreme Commander.

Liek this?

[Argh]

Yes, like that. Preferably driven by some Lua that's hopefully not too difficult to plug new buttons into or move / resize existing stuff.

The thing about a front-end is that it can be really simple, so long as the method of adding sub-menus to a given button is easy, and triggering a sub-menu also triggers a new context in terms of layout. Talk to Pxtl, he really got that part right with AppLauncher, it's so easy to write new menus with that it only took me an hour or so to do the code for the last release of P.U.R.E.

With that kind of design... even fairly complex menu systems (say, complete how-to-play guides all on the engine, unit viewers, campaigns, that sort of thing) are all right there, in one place, and guys like me will find the time to write things like a complete replica of SpringSettings, so that we gradually lose the separate applications which make adoption harder.

[el_matarife]

I suggest just having the multiplayer button quit the game and start a lobby, like it does in Supreme Commander.

Liek this?

If you click multiplayer on the front menu of Supreme Commander, it literally quits and launches GPGNet. You then join a game from GPGNet's browser, and it starts Supreme Commander for the "battle room". The Supreme Commander method is totally inefficient and crazy, but it does have the benefit of being really simple to implement. The only difference would be that Spring wouldn't start when you join a battle room, which is a good thing anyway, since it makes it easier to multitask while you wait for a game to start.

[Auswaschbar]

It starts springlobby and quits itself when clicking multiplayer. After that, you get normal lobby behavior.

[el_matarife]

It starts springlobby and quits itself when clicking multiplayer. After that, you get normal lobby behavior.

That's perfect then. Does SpringLobby or TASClient or QTLobby have commandline options we can feed it? If so, make a button that launches the default lobby on the replay tab for replays. I think there should also be a "Settings" button to launch Settings.exe and a "Help" button that launches the default web browser and goes to the Spring wiki.

What's a good list of buttons with their functions? Help me write up a good list. (This isn't in order)
[*]Multiplayer - Launches lobby's battle tab
[*]Singleplayer - list of installed Lua scripts or launches singleplayer screen on the lobby client
[*]Settings - launches settings.exe
[*]Keybinds or Interface - some sort of button that will launch the mouse selection interface and keybinding editor. I think the two probably need to be combined into one program and with a user interface clean up. I think the selection editor is pretty user hostile right now.
[*]Help - should open a web browser and go to either the wiki or a new dedicated help page
[*]Credits - list of all the Spring developers, plus the GPL license and credit to any libraries used like SDL
[*]Lua UI - Turn on and off LuaUI scripts without starting a battle
(Sidenote: anyone know why the

control doesn't work?)

[Argh]

• Multiplayer - Launches lobby's battle tab
• Singleplayer - list of installed Lua scripts or launches singleplayer screen on the lobby client
• Settings - launches settings.exe
• Keybinds or Interface - some sort of button that will launch the mouse selection interface and keybinding editor. I think the two probably need to be combined into one program and with a user interface clean up. I think the selection editor is pretty user hostile right now.
• Help - should open a web browser and go to either the wiki or a new dedicated help page
• Credits - list of all the Spring developers, plus the GPL license and credit to any libraries used like SDL
• Lua UI - Turn on and off LuaUI scripts without starting a battle

That's why I'm pointing out that adding / moving / resizing the buttons and adding sub-menus is a must-have... that's your list, most projects will want credits for their designers / contributors, etc. and maybe explain their backstories, gameplay, etc., etc., which is all pretty easy if we can write simple button behaviors, i.e.:

pageZero = {
buttonOne = {"bitmaps/buttonOne.png",30,30,130,60,NavLocation = "pageTwo",},
}

To get there, we just need a few basics- ComboBox / Checkbox in addition to Navigation buttons, to facilitate writing a Settings presentation, a way to display a list of items with a scrollbar (using bitmaps specified), and a way to display bitmaps that aren't buttons (or they could be buttons, but buttons that don't do anything).

Add behaviors like Pxtl did with AppLauncher, where there is an alternative button graphic that's shown briefly before it navigates (minor visual polish) and it'd be good for the vast majority of uses, and flexible for weird ones.

In the end, players will launch a game, using parameters set- i.e., this Mission --> Map, SP vs. whatever AI / Lua, so there has to be a way to set up those parameters so that when players hit "Launch game" we can set up the proper ReHost and launch Spring with the desired behaviors.

The real challenge is going to be writing a Lua'd keybind application (I think somebody did that already, but that needs to be elegant and allow people to put in custom commands into a list- user clicks command, selects a key / keys to bind to, and it's now bound, etc.).

[el_matarife]

That's why I'm pointing out that adding / moving / resizing the buttons and adding sub-menus is a must-have... that's your list, most projects will want credits for their designers / contributors, etc. and maybe explain their backstories, gameplay, etc., etc., which is all pretty easy if we can write simple button behaviors, i.e.:

Yeah, excellent point this should all be editable in some easy config file for all the mods building custom installers.

The real challenge is going to be writing a Lua'd keybind application (I think somebody did that already, but that needs to be elegant and allow people to put in custom commands into a list- user clicks command, selects a key / keys to bind to, and it's now bound, etc.).

I'm reaching out to Sefidel to ask him to knock out a quick .NET application to do keybinds and selection editing. I think it should be integrated in the lobby or somewhat standalone like the settings++ is right now.