MetalSucker's lobby simplification thread

[raaar]

i don't think we need a new lobby. The more lobbies there are, the harder it is for someone to help new players due to lobby differences.

What's wrong with the existing lobbies? we should fix that.

Allowing people to join games they aren't yet ready to play can enable them to get feedback from users inside the game, but may also prevent it from starting. It can become troublesome if many people keep joining the game on those conditions (happened with Evo).

The example you presented is interesting. We could use a better system to signal intent to play a specific game type on existing lobbies. A way to send to the main chat channels messages with a link other players could click on to join the game, something like:

<GAME_VERSION> | <DESCRIPTION> on <MAP> ... /<TOTAL>)!

In my opinion, automatic queues would only work if we had at least a hundred players for each game.

The current long waiting times to find games is partly due to social issues. People don't seem to like 1v1 too much, they join large team games and many spectate instead of joining separate rooms and running several games simultaneously. It may also be due to the difficulty in setting up the game, selecting spawn locations, etc.

[MetalSucker]

I'm not stopping anyone from making their lobby better. In fact most of what I'm describing can be implemented in any lobby. I will probably do some bug squashing on current lobbies too should I have the time (because after all, time + skill is the issue)

But the thing is, nobody stopped anybody from improving the lobbies until now. I see ZK has some friendlier lobby features, but I don't see any public conversation about its implementation, it's pretty much a closed development circle - not bitching since there's source to be seen, but it mostly is zk-specific at the moment.

I have taken the point that thin clients are better - it's not hard to replace a chatroom with server logic and a few extra commands.

What I am looking for is feedback from the spring community, and feedback I am getting, that's great. I really don't wanna start work on something that will be frowned upon, not used, not useful, that has no chance of becoming a part of the games. So right now I'm about simple things.

This thread will probably be the basis of my todo list / public bug tracker at one point. Please understand that as far as the actual spring + spring tools development goes, I am new to the community and I've been digging around a lot to get to my current level of understanding, while most devs here go back 5 years or more.

I also don't write a single damned line of code until I have something that kinda-remotely-resembles a technical design thing, that more experienced people here say it's at least in the right direction.

[Silentwings]

But the thing is, nobody stopped anybody from improving the lobbies until now. I see ZK has some friendlier lobby features, but I don't see any public conversation about its implementation, it's pretty much a closed development circle - not bitching since there's source to be seen, but it mostly is zk-specific at the moment.

Sorry, but this is just plain wrong. (1) Nobody is stopped anyone from improving lobbies now, either. (2) There is active lobby development and discussion of it in the usual channels. (3) Lobby development is not a closed community - afaics the only noticeable barrier to entry is having to read the lobby protocol.

[Licho]

Btw ZKL is hardly a closed cycle.I believe it has more contributors than SL and stuff is discussed, just in #zkdev and not here..
For example original proposal is here https://code.google.com/p/zero-k/wiki/MatchProposal
It has been discussed meanwhile in #zkdev extensively

[PicassoCT]

Is there a queud people only chatroom?
3 mins can seem like a long time, if you have no clue whats going on.

[Licho]

It will say stuff like "Game will start in 0:40"
or "Need 2 more to start".

And because queue is actually "fake" battle room (as in open battle) - it has its own chat.

[MetalSucker]

Discussing a thing once in a chatroom does nothing for new developers.

....

Hey, at least I made my own thread to criticize everything instead of spamming threads all over the forum.

// evil rant edited out

[Silentwings]

I don't see any public conversation about its implementation

Discussing a thing once in a chatroom does nothing for new developers

It certainly doesn't if they don't join the chatroom and then claim no discussion takes place.

[MetalSucker]

Silentwings, chat is a realtime system that presumes all devs are there at the same time -> bad idea. It's really not the way to develop anything.

Nobody can be asked to go over years of previous chat logs - if they even do exists. You are restricting development to just the few people that are online at that time.

I get it, you don't like what I'm saying, you think I'm doing it wrong and I don't understand it. Noted.

[Silentwings]

chat is a realtime system that presumes all devs are there at the same time -> bad idea.

Bollocks. There is the option to record (I use it), as well as the option to keep a near-perma-present IRC client (which many do). Better still, there is the option to ask stuff.

you think I'm doing it wrong and I don't understand it. Noted.

Said literally none of this. I don't know if you've done anything at all. But if you are going to do something, the info you want won't get served up to you on a plate. And you will only cause offence by confidently claiming that "nothing is happened and there is no info/help anywhere" when you just haven't looked very hard.

[MetalSucker]

You really don't like me, I get it.

[Silentwings]

I have nothing personal against you. I only dislike your confident claims of things that are not true.

[MetalSucker]

Yes, please stay out of this thread. Thank you.

[Jools]

I think what you are discussing is a good idea. You should be aware of the fact that some people here are quite young, and therefore may speak with a language that sounds overly confident and authoritative, but is actually not based on any facts or is out of perspective.

One more issue with this forum here is that most people discussing here are devs or moderators, not actual people playing. Therefore you will get an over-emphasis on discussing the code and not the actual product as the end user sees it.

I think the visual part is important: try to make the lobby look as something serious, white background and not black, and if you want to include nice rank icons, I think there is concensus that the tasclient ones are the best looking, see this topic: http://springrts.com/phpbb/viewtopic.php?f=9&t=21934. The devil is in the details.

Also note this topic: http://springrts.com/phpbb/viewtopic.php?f=10&t=29896

[Silentwings]

Yes, please stay out of this thread. Thank you.

Fine, best of luck.

[Google_Frog]

MetalSucker your idea looks good. It looks a lot like what I have been trying to steer Licho towards with ZK. He seems to want a bit of a more complicated system with spectators and chat. The queue battleroom idea is sound though as it is effectively your channel system, both will be hidden from users.

There are various things that have annoyed some people in this thread. Some seem to be annoyed that you aren't going to support all games and lobbies while others are annoyed that you are making 'yet another lobby'. Hopefully it is clear that these people can be ignored. Game developers (that refers to the part of the game experience that runs on the Spring engine, not the whole game experience) around here seem to take lobbies for granted and get annoyed when someone makes a lobby without their game in mind. There is an entrenched idea that games and lobbies should be modular which mostly results in one-size-fits-nobody. The current lobbies that I am aware of are all a complicated mix of automatic downloader, IRC and generic battleroom viewer and none of that is particularly user friendly.

So do whatever you want and support whichever game you want to support. There is a chance of people incorporating your ideas if they are shown to work. But in this case there is a decent chance that game developers will sit back and wait for you to do their work for them. There have been modular ZK features in the past that people wanted for other games but never got around to configuring the systems required to run them.

[BaNa]

If you do need a lobby to build on, I would recommend this one as my favorite past lightweight lobby client.

Hasnt been in development for a while, so I'm sure it would need some sprucing up, but it was simple and clean and got the job done when it was working. It reminded me of utorrent in the torrent client world, remember all the bloated horrid torrent clients before that?

[enetheru]

what you have describes is exactly what I want to use to play spring.

I dont want all the features of the other lobbies, I dont want any fancy UI. I just want to play a game arcade style ASAP with another human.

so thank you for this effort.

[abma]

a great simplification would be if the engine itself could be a lobby client. small efforts are done into this direction, imo this should be prioritized. some basics already exist to create a lobby client in the current development version of spring.

also fyi: zero-k currently implemented sth. like you suggested as licho alread wrote: https://code.google.com/p/zero-k/wiki/MatchProposal

[abma]

a great simplification would be if the engine itself could be a lobby client

it can: http://springrts.com/phpbb/viewtopic.ph ... 99#p561799

-> https://www.youtube.com/watch?v=C67RwpT_FwY

kudos to gajop!