Spring developer crisis

[PicassoCT]

But he has a point.. a real C++ AI Competition, with the monthly winner on the frontpage .. that would gain us alot of people.. and doesent lua have a c/api ?

I do not doubt that lua devs could do a complete and complex AI in lua. But the engine needs c/c++ devs, so there is need for a C API and a constituted competition.

The Rules:

I. On every first of month the competitors shall meet in a Free for All AI-battle to the dev. The winner is the last team standing.

II.The Game of competition is the one with the most players by stats that month.

III.The Engine Version used is the current release. No matter how broken. Any AI can put up a fight, a excellent AI can win with one hand tied to the back.

IV. The language used is C++. No ifs, no buts, no throws, no catches, swallow it switches.

V. A replay has to be released and the winner has to open his source.

VI. The Winners AI and his name shall be put to the Frontpage, throughout the lands he shall be know as the sole Survivor. The Loosers shall wear the [~SHODAN] clantag as sign of there shame.

VII. Why are you not coding yet?http://springrts.com/wiki/AI:Development:Lang:Cpp

VIII. Every post after this is volunteer.

[FLOZi]

The lua API docs are based on a modified BNF, not sure how that comes across as gibberish to anyone with a CS BSc. I did some messing around with templates but couldn't agree on various issues with other key contributors.

I can't speak for the devs but the view I perceive is that devs prefer not to write docs on wiki because they prefer it to be 'close to code' and generate-able, hence automatically updated.

I agree with a need for some clear pages outlining the technical aspects of Spring (Pages on e.g. technical details of algorithms used (ROAM, QTPFS) and the sim model etc) and 'Possible areas for development' (The latter basically exists already but is somewhat outdated)
N.B. The 'About' page mentions "ASSIMP, ... we have a lua environment, that we're OpenGL"

Also not sure how your hatred of fixed width layouts really puts off engine developers (I won't make a stereotypical comment about programmers' art )

I'll agree that the environment is often caustic, but a. we now have an active mod team again and b. welcome to the interwebs.

Also for the sake of it we now have c. A wiki with a sensible category structure, you're welcome.

[enetheru]

Don't listen to anything this man says. AF stands for Alantai FOOLstar. He couldn't program his way out of a paper bag. He brings shame to the good name of Java. If you see the letters A and F in that order, turn around and run the other way.

Reasons why AF is not a very good thing:

• He speaks lies, untruthitudes and falsities.
• He just hangs around to increase his postcount.
• Makes endless bulletpoint lists.

is this serious?

[knorke]

this doesn't mean that native AI makers should be villified, and that's how it's been taken.

imo native AIs vilified themself, you made a thread about it: http://springrts.com/phpbb/viewtopic.php?f=15&t=19524 Stability problems and unclear to use for players (which game is supported etc)

People don't play spring, they play BA/ZK/etc
The tools for individual games to build on for self determination are not enough.
The lobby situation isn't that great, and advancements have always been in the wrong areas ( I should know, for a longtime I was the one making them ). Only ZKLobby made meaningful steps in the right direction, not counting NOTALobby

The problem is that gamedevs feel there is the need to rebuild such things as lobbies, installers etc in the first place. They should not have to do that.
Sadly it is not enought that people make their game.sdz, they must also re-invent all the infrastructure around it like lobby, installer, mappack.
Do you think that will happen for games that people made in a gamejame? It is a serious obstacle that should not be there.
Many players play multiple games, any "branding" does more harm than good.

This infrastructure should be provided by Spring, that is also why zKlobby and NOTAlobby are imo steps in the wrong direction: If Imperial Winter was finished next week should players wait another year until IWlobby is done, too?

[AF]

Before I continue, let it be known that i lost the mantle of highest post count after an almost 9 year streak to the venerable Smoth.

Also, BNF may well be a good way of doing it, but for those unfamiliar a link explaining it would be enormously helpful. When I first looked at it i found it confusing, and although I figured it out in time, it was very intimidating.

I can understand why auto-generation would be useful, so perhaps a discussion on automated generation tied to buildbot and where in the site that would fit should take place?

I also agree regarding lobbies, but I make the distinction between branded and brandable lobbies. ZK Lobby is a step in the right direction for ZK, but it doesn't benefit other people and projects in doing so other than via ZKs popularity by association.

I'll also note that I don't have an utter distaste of fixed width at all, my own sites are fixed width. Having said that, this site could well be expanded to 920pixels with minimal issues. I did most of the necessary work myself, and a quick change in the web inspector demonstrates this.

There are people ( such as Smoth ) who worry about the readability of forum posts with the extra width, for which there are solutions, but none were seriously investigated, hence why the width hasn't been changed.

But that would merely be an incremental improvement, which is not what we need.

[FLOZi]

I agree that a page (or template to include in the pages) with a key to the grammar would be a good idea.

RE: site; still doesn't address that imo prospective developers don't too much care.

[raaar]

People don't play spring, they play BA/ZK/etc
The tools for individual games to build on for self determination are not enough.
The lobby situation isn't that great, and advancements have always been in the wrong areas ( I should know, for a longtime I was the one making them ). Only ZKLobby made meaningful steps in the right direction, not counting NOTALobby

The problem is that gamedevs feel there is the need to rebuild such things as lobbies, installers etc in the first place. They should not have to do that.
Sadly it is not enought that people make their game.sdz, they must also re-invent all the infrastructure around it like lobby, installer, mappack.
Do you think that will happen for games that people made in a gamejame? It is a serious obstacle that should not be there.
Many players play multiple games, any "branding" does more harm than good.

This infrastructure should be provided by Spring, that is also why zKlobby and NOTAlobby are imo steps in the wrong direction: If Imperial Winter was finished next week should players wait another year until IWlobby is done, too?

I kinda agree. Game devs made dedicated lobbies, but they didn't have to, did they?

We already have the game-neutral springlobby. Maybe more effort should be put into improving it and making it compatible with game-specific "skins", "plugins", "webservices" or something like that.

In general, some effort should be made to define what are the simplest and most relevant features/bugfixes missing.

[AF]

A more component based setup would be better, as they'd need to do something similar to make an installer, so why have 2 paradigms.

What matters more is if it says "SpringLobby" in the start menu or "Play XYZ", the rest can be setup with hand picked defaults and some modifiable text entries and filters. It's not an ideal situation, but it's an improvement.

Re Site: Indeed increasing the width has advantages but they are small. It's worth doing but I agree it is not a fix or a big one at that. Painting a broken gate doesn't fix it, what we need is a site that sheds a lot of the clutter and mess and lets the tools we already have work better.

[luckywaldo7]

This infrastructure should be provided by Spring

Who is Spring?

[knorke]

This infrastructure should be provided by Spring

Who is Spring?

Not sure what you mean. The infrastructure is sites like springfiles.com, rapid, lobbies, replays.springrts.com, this forum, autohosts etc. It is all run/developed by different people. Who these people are should generally not matter to a gamedev - as long it is not him because that would mean to re-invent all these things.

[luckywaldo7]

The infrastructure is sites like springfiles.com, rapid, lobbies, replays.springrts.com, this forum, autohosts etc. It is all run/developed by different people.

Volunteers do it because they want to. Not because someone on a forum declared that they are Spring and they "should provide it".

[AF]

The likes of springfiles and the spring lobby server and autohosts, are not community infrastructure, they're private infrastructure provided as a community service.

What's important is that one can grab all the tools and deploy their own infrastructure with minimal effort if they want to. If they need to build their own springfiles or spring lobby, or code a new autohost on the other hand....

[SinbadEV]

What's important is that one can grab all the tools and deploy their own infrastructure with minimal effort if they want to. If they need to build their own springfiles or spring lobby, or code a new autohost on the other hand....

Yes, this is one of the more troubling issues... when iamacup went away it took quite some time for someone to reverse engineer the infrastructure his servers had been providing... our bus number is higher now but, ideally, one could get everything they need to set up an entire parallel infrastructure without having to create any new code... and I don't believe this is the case.

[knorke]

The infrastructure is sites like springfiles.com, rapid, lobbies, replays.springrts.com, this forum, autohosts etc. It is all run/developed by different people.

Volunteers do it because they want to. Not because someone on a forum declared that they are Spring and they "should provide it".

Yes, of course it it is volunteeres.
Do you mean something along "providers of infrastructure can do whatever they like, nobody can force them to do xy because it is all done in free time?"
Of course that is true.

But still most modders can not run an autohost or even lobby server on their home PC, develop a lobby or provide enough bandwidth for fast downloads. And I think it is important that they do not have to.
Obviously saying "blabla needs better autohost, needs better lobby" does not solve it but maybe it motivates others to create these things if they see there is a need?

[luckywaldo7]

Obviously saying "blabla needs better autohost, needs better lobby" does not solve it but maybe it motivates others to create these things if they see there is a need?

This is where I fundamentally disagree. After all, generic complains like this are frequent and responses are non-existent. You aren't serving to inspire anyone, just let them know that if they tried, they will probably not live up to your standards because no one else has so far. So I think it really does more to prevent action than anything else.

And when talking about particulars like lobby and autohosts anyway, I can't help but think of this: http://www.joelonsoftware.com/articles/ ... 00069.html

[knorke]

well for example the answers.springlobby page was taken offline because owner thought there was no more interesst in it. Someone else stopped running his relayhost because he thought it was not needed anymore. (atm there are no working relayhosts left)

[AF]

Springlobby was born out of the motivation for a better lobby, specifically a cross platform lobby built in C++ rather than a windows only setup like tasclients delphi, or something that needed a Java VM.

Suffice to say it ran its course, and some may not like the trajectory it took, but it took it, and we're better off with it than without.

Anyway you can't just say we need a better lobby or autohost, without defining "better at what?". 1 lobby fits all isn't and has never been the optimal strategy, as different setups provide different experiences. One dev may want a steam style community setup, be it because they did a modjam, or a research experiment or a fun side project (e.g. KDR's THIS ). Another may want a simplified lobby with branding that does the bare minimum for a game they're ploughing time and effort into ( Forboding Angel/EvolutionRTS fits this bill ).


Eitherway they all needed better docs and easier access. You can find details on the relay system if you dig deep enough, but we can do better.

Lobby wise I see these avenues and competing end goals:

• Brandable simple lobby for repackaging
• Bells and whistles generic community lobby client
• Easy deployment autohost and server with instructions

Our community has almost entirely focused on the second goal, while not quite getting the third goal, resulting in a lot of unused projects and uncertainty.

I'd also suggest any further lobby discussion be split into a separate thread

[knorke]

Anyway you can't just say we need a better lobby or autohost, without defining "better at what?".

Anyone who has ever played a spring game knows "what." The forum/bug trackers etc have lot of feedback and feature requets. Seemy you know it too:

Eitherway they all needed better docs and easier access.

One dev may want a steam style community setup, be it because they did a modjam, or a research experiment or a fun side project (e.g. KDR's THIS ). Another may want a simplified lobby with branding that does the bare minimum for a game they're ploughing time and effort into

How do you do decide what is a "fun project" and what is a serious effort? Going by playernumbers the only serious projects would be zK and BA while everything else is side projects. In the end all games have so similiar demands from a lobby that it does not justify "branding" or a completly new lobby.

[AF]

How do you do decide what is a "fun project" and what is a serious effort?

Simple, you ask. A game dev will have an idea of what they want, and if they don't, perhaps they're still deciding or haven't thought that far ahead? In which case they probably want the community lobby rather than a game specific one. Not that it's relevant. We can see evidence of demand for both already.

Anyone who has ever played a spring game knows "what." The forum/bug trackers etc have lot of feedback and feature requets. Seemy you know it too:

Indeed I have ideas, but that would result in a lobby that is the ideal for me. You are not me, and not everybody has the same goals in mind. Mish-mashing every request together is going to result in a lobby that benefits yourself specifically, or a mudball that tries to do everything and doesn't work that well.

Perhaps people wanting X added to Y is not because X is a good idea, but because Y is a broken one, and the correct answer is actually Z?

[hoijui]

i always wanted Z!!!
it was not really clear to me, but now.. i can see it!
i gonna have myself some kids with at least one Z chromosomes.
and if i have to screw some half ass'ed japanese robot for it, i gonna get there!
and if i have to eat untested nano-bender-robots that re-bend my own chromosomes, i gonna get there!
and if i have to donate my whole body, instantly, to be able to take part in the alpha testing of such a science project, i gonna get there!
signed by
a spring developer in crisis