"Standard LongTurn ruleset"

Planning and discussing Freeciv Longturn gaming
User avatar
Corbeau
Elite
Posts: 508
Joined: Mon Jan 13, 2014 11:13 pm

"Standard LongTurn ruleset"

Postby Corbeau » Thu Apr 13, 2017 11:41 pm

(I have put quotes in the title because I don't think that standard rulesets should be a holy thing, there is also room for variation and improvement and there is never a "best" ruleset so different games and different people will always like different things. Besides, the POINT of a customisable rulesets is that you can customise them and have some variety in life ;) But let's try to find the lowest common denominator.)

The main reason why LongTurn games need to be much, much more carefully tuned is the fact that tehy last for months. If a fast- or medium-paced game that is over in a few hours, or a day at most, you can forgive and forget glitches that influence it. You may even not notice some disbalances because the game progresses pretty fast and there is no time to analyse all the details. Worst case, you'll shrug it off and play another one next day or next week.

However, LongTurn game is influenced by some mis-programmed feature, that is a big deal because it can spoil something that you have been slowly building for months. There are far fewer LT games and this is why they have to be much more carefully prepared. Also, when playing such a game, you don't just do the moves, you also spend some time planning and analysing and you are able to catch much more issues than in a short and fast game.

So these are some LT-specific things that need to be addressed.

1. One thing that is a must in LongTurn is x2 movement. it's not only a matter of the world size, but also of the fact that you have to wait a whole turn for your units to move again. For example, a worker. With x1 movement, you get one day to move, another day to actually start working. With x2 movement it becomes neater and more sensible. Current LT37 is being played with x3 movement which I think is a bit too much.

2. Startup should be quicker because the first week or two of the game tend to be more boring, and you don't remove this boredom by faster turns like in non-LTgames. A quickstart means more Settlers and workers, maybe even more explorers, although I think most LT/GT games started with one explorer. What I'm working on rigt now is a Founder unit (or "Super-Settler") that would move even quicker to speed up the startup, but would be start-only, can't be built unit. What could also be done is get some more Warriors in the beginning so that you can concentrate on growing.

3. This is not my personal issue, but it seems that the LT community has grown weary of trade routes. I find it weird that you can build one unit, send it to another city and receive profit from that until the end of the game, if both cities stand. So that should be looked into, experimented with and analysed some more in oredr to balance the issue.

4. Tech exchange seems to be a big issue in LT. With a lot more emphasis on diplomacy and alliances, this is a very big potential party-pooper. I don't like how this issue has been addressed so far and I'll propose my solution. First, the problems with two extreme options:

a) No tech trade.This is a problem because, once a nation lags behind, it is very difficult to catch up. TechLeak helps only a little bit; a smaller nation will remain backward forever simply because it can't get enough bulbs, period. So, basically, the game is decided too early, there is no real cooperation and the only viable strategy becomes getting as many cities as possible with no space for variety.

Most LT games recently had this setting which I believe is a pity because it removed an important component of the game and made cooperation between players at a very basic level.

b) Tech trade allowed. A huge problem with this is that te game basically becomes a popularity contest. Whoever gets the biggest alliance, wins. Also, a nation that joins this alliance will get all the techs without even trying, while someone who decided he's like to play alone will become hopelessly backward. So, the only way to win is to group up and become a part of a herd.

I think the best solution and middle ground would be to combine two things: Tech upkeep and root_req.

Tech upkeep option requires spending bulbs on maintaining the present level of technological development. the amount of upkeep needed depends on the number of techs you have doscovered, but also on the number of cities. So, a large empire will use more bulbs to upkeep its tech level than a state with one city (Singapore!) This way it will be more difficult to receive an infinite number of techs for free from your friends because whatever you receive, you also have to upkeep.

A possible workaround/abuse could be to share only critical techs, such as Gunpowder, without sharing its prerequisites. For this, let's implement root_req wich dictates that, for example, you can't HAVE Gunpowder without Invention nor Invention without whatever is its requirement. So you must HAVE those requirements and upkeep them all.


All this is a pretty complicated mechanism and it needs to be balanced very well. In one game that I played, tech upkeep level was set to high and it strangled almost everybody so that they lost the wil to play.

----------------

So much for now. I will continue later.

User avatar
AndreasR
Elite
Posts: 655
Joined: Thu May 02, 2013 10:26 pm
Location: Norway

Re: "Standard LongTurn ruleset"

Postby AndreasR » Fri Apr 14, 2017 8:32 am

I hope that you could create a new LongTurn rulesets for Freeciv, using the latest svn trunk version of Freeciv, and propose to get it included in Freeciv. Once the ruleset is included in Freeciv, then it can be included in Freeciv-web also. This was a great description of the ruleset.

User avatar
Corbeau
Elite
Posts: 508
Joined: Mon Jan 13, 2014 11:13 pm

Re: "Standard LongTurn ruleset"

Postby Corbeau » Fri Apr 14, 2017 9:24 am

AndreasR wrote:using the latest svn trunk version of Freeciv

I probably will as soon as I find out what it means :)

User avatar
AndreasR
Elite
Posts: 655
Joined: Thu May 02, 2013 10:26 pm
Location: Norway

Re: "Standard LongTurn ruleset"

Postby AndreasR » Fri Apr 14, 2017 9:31 am

Corbeau wrote:
AndreasR wrote:using the latest svn trunk version of Freeciv

I probably will as soon as I find out what it means :)


The latest version of the Freeciv source code can be found by following this documentation: http://freeciv.wikia.com/wiki/Freeciv_s ... repository

User avatar
Corbeau
Elite
Posts: 508
Joined: Mon Jan 13, 2014 11:13 pm

Re: "Standard LongTurn ruleset"

Postby Corbeau » Fri Apr 14, 2017 10:45 am

That will happen when - and if - I get the time. Right now I barely have the time to do very minor changes to a ruleset every other day and, since I'm not a programmer, it would take me days, maybe even weeks of concentrated effort just to be able to understand everything that is written on that page. And I just don't have the time for that right now. I'm just a player with the tendency to tinker with stuff and devise new systems while sitting in a bus to and from work.

User avatar
JTN
Elite
Posts: 373
Joined: Wed Jan 30, 2013 12:15 am

Re: "Standard LongTurn ruleset"

Postby JTN » Sat Apr 15, 2017 9:24 am

To get a 2.5 ruleset to work on trunk you'd need to do several rounds of updates:
http://www.freeciv.org/wiki/How_to_upda ... 2.5_to_2.6
http://www.freeciv.org/wiki/How_to_upda ... 2.6_to_3.0
...and a 3.0 to 3.1 (trunk) step that isn't documented yet; don't know if any changes are actually needed yet, 3.1 only branched recently.
There is some mechanical ruleset format translation available for the 2.6->3.0 steps onward, in the form of the "freeciv-ruleup" tool.

(yes, our stable release is now several major versions behind the head of development; my bad)

Which makes me think: trunk is bleeding-edge code which is the developers' playground and formats and behaviour can change incompatibly. Freeciv-web is based off trunk and occasionally takes updates, and stops being able to load old savegames and so on. That sort of churn seems incompatible with the stability that a Longturn game would need; I guess you'd at least need to freeze a particular trunk version, if not have a version of freeciv-web based on a less churny branch.

(And this all sounds like more than Corbeau has effort available for.)

User avatar
AndreasR
Elite
Posts: 655
Joined: Thu May 02, 2013 10:26 pm
Location: Norway

Re: "Standard LongTurn ruleset"

Postby AndreasR » Sat Apr 15, 2017 10:58 am

I think LongTurn in Freeciv-web would be great. It has been requested multiple times by end-users. The current LongTurn test game with 10 players filled up very quickly.

> Freeciv-web is based off trunk and occasionally takes updates, and stops being able to load old savegames and so on. That sort of churn seems incompatible with the stability that a Longturn game would need; I guess you'd at least need to freeze a particular trunk version, if not have a version of freeciv-web based on a less churny branch.

I am open to suggestions about how to improve the process of keeping Freeciv and Freeciv-web updated and in sync. For LongTurn it is also possible to freeze the Freeciv server on a particular trunk version.

cazfi
Elite
Posts: 1292
Joined: Tue Jan 29, 2013 6:54 pm

Re: "Standard LongTurn ruleset"

Postby cazfi » Sat Apr 15, 2017 9:34 pm

Corbeau wrote:you can't HAVE Gunpowder without Invention nor Invention without whatever is its requirement. So you must HAVE those requirements and upkeep them all.

Whether holes in the tech tree are allowed or not is controlled by just couple of entries in the game.ruleset in the latest branches. Using lots of root_reqs (setting them to all techs and not only to create a couple of separate tech trees) is known to cause performance issues.

User avatar
Corbeau
Elite
Posts: 508
Joined: Mon Jan 13, 2014 11:13 pm

Re: "Standard LongTurn ruleset"

Postby Corbeau » Sun Apr 16, 2017 9:00 am

Can you clarify the first sentences, please? Does "latest branches" mean "not 2.5" and does it relate to Freeciv Web?

As for performance issues, what were they? Morphles (I think) made such ruleset (every tech had a root_tech entry, so he says) and we played one or two games admined by him and I don't remember any issues. If they were lag-related or something like that, it probably didn't have much effect in longturn (because plenty of time to make a move, running speed at TC not really relevant and so on).

cazfi
Elite
Posts: 1292
Joined: Tue Jan 29, 2013 6:54 pm

Re: "Standard LongTurn ruleset"

Postby cazfi » Sun Apr 16, 2017 9:25 am

Corbeau wrote:Can you clarify the first sentences, please? Does "latest branches" mean "not 2.5"
Yes, I think the fature was introduced in 2.6.

Practically unusable slow client when using experimental ruleset with all techs having root_req has been reported many times. We had to revert to not using so many root_reqs in some branch (S2_5?). It might be that problems occur only in conjunction with some other ruleset property.