LteX2.1 - improvements
Fri, 06/19/2009 - 17:06 — maho
New improvement in Ltex2.1. I have added 2 new settings:
/help tlgive
Option: tlgive - chance to loose invention while givint it to someone
Status: changeable
Value: 50, Minimum: 0, Default: 0, Maximum: 100
/help tlrecv
Option: tlrecv - chance to loose invention while receiving it from someone
Status: changeable
Value: 50, Minimum: 0, Default: 0, Maximum: 100
It means, that by the way every transfer of technology (stealing, conquering or trading), both donor and receiver have chance to loose that tech.
I have set tlrecv to 40 and tlgive to 60.
Example what happends when someone steals you tech:
Comments
In my opinion random chance
In my opinion random chance rules make the game less strategic and more frustrating. Consider making it more deterministic somehow.
I also don't see how someone can "make a mistake transferring knowledge" and as a result forget it, but whatever, it doesn't matter really how you justify it. ;)
I did have an idea a while back about "negative science", i.e. a player could get negative science points in a turn and as a result "forget" already know techs. This could be used for example to model the degradation of a civilization when in a fundamentalist or similar government too long.
Another idea I had was to make non-discovery tech transfer only capable of learning the techs that allow some things in the target. For example, if you don't know Laser and you conquer a city with SDI, Laser would be one of the techs you could gain as a result of capturing the city. Similarly for a diplomat/spy, if there is say a submarine in a city and you do not have Combustion, then in stealing tech from that city Combustion could be gained.
one more...
Great ideas. :)
How about this one: units need to have a unobstructed supply line with their home city. I.e. it must be possible to plot a complete connected line of tiles from a unit to its home city with no hostile units on the path (ocean squares are ignored or perhaps require a vessel). If they do not have a supply line or if it is broken, the unit loses some HP every turn (like a helicopter).
That way you can starve lone units which are hiding in mountains etc. Requires you to mind your flanks also.
Of course, the idea comes with complications, since units such as explorers and partisans should be exempt.
And maybe weak units like catapults and canons lose more HP than legions and conscripts. And perhaps units can exploit the food output of the square they are on...
Btw Book, you ought to play next LT to enjoy the benefits of the improvements you made. :)
Implementing supply lines
Implementing supply lines would be a little tricky, and to do it right would probably require adding new packets (e.g. so the server can send the "supply route" to the client so it can display it). Also, you would have to consider that computing paths for hundreds of units even just over the turn change would be rather cpu intensive. I'm not sure about the best way to implement it, maybe there is a way to avoid these problems. I'll think about it for future freeciv versions. :P
supply lines
I definitely would like to play a freeciv game with supply lines !!!!
I think this would be a great improvement for this game who makes warfare too irealistic and with too quick benefits (my opinion).
I see things rules like that (forgetting CPU issues or client compatibility problems, but trying to make them simple) :
- beggining with musketeers (or let's say with units with more than 10 hp), because medieval units were used to live on pillage. If i remember well, supply lines really became mandatory during napoleonian wars. Ah, medieval units on mountains should be supplied too, no ?
- if a unit has no valid supply line at TC, it looses hp. Let's say 1/4 of full hp per turn. According to the scale of time in freeciv (> 1 year), it should be instant death...
- supply line is valid if the unit is able to reach an allied city or fortress within a turn, including ennemy units and zoc (this to be able to cut supply lines).
Would it be too complicated ?
Surely many other things to add sens. For example, supplying is already implemented in the game by the shield taken from the home-base city. So the supply line should go from the unit to its city, no ?
May be introduce another type of unit specialized in supplying on long distance...
Please Book, make supply lines your priority ;)
I gave it a little more
I gave it a little more thinking recently.
I think it is important that we build on the currently used concept that a unit has a home city from where it is supplied. It would in my opinion not make sense that a unit could be supplied from a town different than the one that is paying the shield. So we must let the supply line run from the city that is paying the shield.
But what if a unit is far from cities, and wants to switch supply to a different city? It would not be realistic for that unit to travel a big distance to a city just to select that as home city. It must be made easier. It must be such that a unit can select which city supplies it, but with a penalty to make is not as easy. Options are:
* Only allow to select on cities that have the "Logistics Office" built in a city.
* Or perhaps, ask that a unit does not move and use movement points for 2 whole turns before switching...
* Some Logistics wonder allows to freely select supply cities on all units.
(Of course, it is still possible to change the homecity by travelling to a certain city and home'ing it.)
Also, let it be clear that a unit in his own nation borders does not use/need supply routes (local population helps with cargo, etc.) How can you then use it against an alpine troops in a mountain who is in nation borders? Simple: capture the surrounding cities.
I also think that the supply routes would have to be optionally visualized. Such as that you can have lines drawn for trade routes, that you can also have lines visible for supply routes (perhaps to avoid clutter, it shows only for the selected unit.)
Sokrat proposes to have supply routes start from musketeers age. I disagree, since even in the Roman times legions were connected to the home front, and a phalanx in a mountain in your terrain can be a big pain in the ass. But you can also consider some units exempt such as a diplo and explorer. Furthermore, you can make sturdy units lose 1/5th HP (such as a legion, riflemen, etc.) and weak units such as catapults and canons 1/2th.
This brings me to a different idea altogether. It is not very realistic that units do not 'eat' but only use a shield. We could make it such that each unit uses 2 food and 1 shield per turn. If it is on a tile that has less than that, it will get the remaining food/shields from the homecity. So keeping units in mountains then is expensive (2 food from home), while walking on plains is 'free'. But then your units are much more vulnerable to the enemy! That gives a totally new prospect on warfare and unit management. (Just a thought that I have right now, not a worked out plan.)
In the summer holiday if I have spare time I will take a look if I can understand the source code and will investigate if I can code supply lines.
This brings me to a
This brings me to a different idea altogether. It is not very realistic that units do not 'eat' but only use a shield. We could make it such that each unit uses 2 food and 1 shield per turn. If it is on a tile that has less than that, it will get the remaining food/shields from the homecity. So keeping units in mountains then is expensive (2 food from home), while walking on plains is 'free'. But then your units are much more vulnerable to the enemy! That gives a totally new prospect on warfare and unit management. (Just a thought that I have right now, not a worked out plan.)
I think it's great idea!.
But I think that unit should take it's upkeep not from homecity, but from nearest city. I have mentioned about it about year ago, it should be somewhere at forum.
Eg. You have army near Berlin, but units are from some city at Siberia or New York. I doubt it that Siberian/New York city pays upkeep, I rather think that villages around Berlin feel pain of having army in the neighbourhood.
So, along with your idea, I think that upkeep should be taken from:
1. terrain
2. if terrain is not enough - from nearest city,
3. if nearest city is not enough, from nearest allied city,
4. if nearest allied city is not enough - from nearest own city.
Re: Btw Book, you ought to
Btw Book, you ought to play next LT to enjoy the benefits of the improvements you made. :)
It's too time consuming. I spend too much time already playing on warclient servers, and not enough time actually developing freeciv. :(
In my opinion random chance
In my opinion random chance rules make the game less strategic and more frustrating. Consider making it more deterministic somehow.
But random chance rules are everywhere in freeciv, even in battle ....
I also don't see how someone can "make a mistake transferring knowledge" and as a result forget it, but whatever, it doesn't matter really how you justify it. ;)
Well, imagine that you invented gunpowder and you give that knowledge to me.
Mistake from your side: you gave me your notes and forgot to make a copy :). You don't remember. Or your lab burned while showing it to me ...
Mistake from my side: I used that knowledge, but suddenly I exploded with my lab, so the only person who knew what is it about - died.
I did have an idea a while back about "negative science", i.e. a player could get negative science points in a turn and as a result "forget" already know techs. This could be used for example to model the degradation of a civilization when in a fundamentalist or similar government too long.
I was thinking about it too, but making it only for fundamentalism would not bring neccesary effect. Note, that I'm trying to prevent situation like in current or in previous ltex game: very strong axis of evil (Elrik, Mir3x et consortes in this game, Durandal in previous) is unbeatable because he can earn more in turn than respawners can weaken them.
But random chance rules are
But random chance rules are everywhere in freeciv, even in battle ....
Yep... and I always try to get rid of them. I am even considering making battle completely deterministic.
Well, imagine that you invented gunpowder and you give that knowledge to me.
Mistake from your side: you gave me your notes and forgot to make a copy :). You don't remember. Or your lab burned while showing it to me ...
Mistake from my side: I used that knowledge, but suddenly I exploded with my lab, so the only person who knew what is it about - died.
Maho is so funny. It sounds like a "civilization" is some poor grad student.
I was thinking about it too, but making it only for fundamentalism would not bring neccesary effect. Note, that I'm trying to prevent situation like in current or in previous ltex game: very strong axis of evil (Elrik, Mir3x et consortes in this game, Durandal in previous) is unbeatable because he can earn more in turn than respawners can weaken them.
I would suggest not allowing huge alliances (maxallies=1 or 2). Also, the bigger the empire the more unit and building upkeep should cost (unfortunately this would require changing code in common/ which means client incompatibility). The "negative science" would probably be best as a penalty for large empires, at least for most governments (maybe excepting the "high science" and high unit upkeep democracy and republic).
To elaborate on the last
To elaborate on the last bit, governments like democracy and republic would allow you to amass tech and gold but would have steep penalties for having large numbers of units and cities. The "war governments" monarchy, communism and fundamentalism would cause you to lose techs gradually, but would allow you to support more units and cities (but the larger the empire, the faster it would "decline" of course). This would model the "rise and fall" of empires, and would theoretically permit freeciv games to go on forever, in a series of cycles. :D
I really like those ideas!
I really like those ideas!
About the supply chain: how about if a unit is surrounded by enemy units on all 4 sides it gets killed? It's like in the boardgame Go: when your 'freedoms' are gone you're dead. If 3 sides are covered by enemy units but you put a friendly unit on the last remaining side then they'll have to kill that one first before covering all your sides, so you can protect your flanks.
Probably more difficult to implement: if a city is surrounded by enemy units on all 4 sides (or all 8 sides), all units it supports are dead.
About tech loss to create a cyclical game: a way to implement this would be to give every tech you own an upkeep cost, just like units and buildings. The more advanced the tech, the more bulbs per turn it costs to upkeep it. That way your net research (total research minus upkeep cost) can be positive and negative. If you become advanced and then turn into a raving fundamentalist you loose techs one by one until the net research is at least 0 again. If the upkeep per tech is small (like 1 bulb/turn for simple techs, 10/turn for advanced techs) then people would normally hardly notice it.
It makes some sense: in the real world you also have to train generation after generation of scientists to make sure you don't forget about nuclear fission or so. If 1 generation is skipped then all our real techs are gone and we're cavemen again.
=== 0h Marduk, thou art chiefest among the great gods! To destroy and to create; speak thou the word, and thy command shall be fulfilled. ===
An instant death would
An instant death would probably be a bit dramatic :-)
In warfare a pocket is not killed instantly but occasionally can break out ( wiki ).
I know how to code so I might give it a shot. Probably would first try something more simple to get to know the Freeciv code. although, I have the idea the Freeciv code is kind of arcane and hard to follow... Book is there any decent documentation on the Freeciv code/architecture? :S
Book is there any decent
Book is there any decent documentation on the Freeciv code/architecture? :S
Unfortunately not really. There is the doc directory in the source tree and some random information in the freeciv wiki, but easiest (if you can read and understand c code) is to just trace out some execution paths in the code base using a tool like cscope or whatever your ide has.
Because most parts of the freeciv code are very old (10+ years) there are a few quirks you just have to get used to, and pitfalls due to rotting code you have to avoid.
Just before you actually start changing things, be sure to read:
http://freeciv.wikia.com/wiki/...
http://freeciv.wikia.com/wiki/...
so you start with "correct" habits, especially if you want me (or any other freeciv dev) to examine your patch and possibly add it to main freeciv (i.e. trunk in the freeciv repository on gna.org).
Of course if you just want to make an experimental feature for longturn you do not have to follow the guidelines so strictly (since maho doesn't anyway :/).
Re: I really like those ideas!
About the supply chain: how about if a unit is surrounded by enemy units on all 4 sides it gets killed?
Probably more difficult to implement: if a city is surrounded by enemy units on all 4 sides (or all 8 sides), all units it supports are dead.
Instant death would be too powerful, for example a guy could use delayed goto to suddenly surround the victim in the last second of the turn. Maybe just a loss of hp for surrounded units; for cities they already lose control of worked tiles with enemies on them, allowing a kind of "siege", but this usually takes too long to kill the city in the popular rulesets.
There is the slight detail that a unit has 8 sides in freeciv, so completely surrounding it would be pretty hard. Maybe just if a unit cannot move at all due to enemy zoc, then it would be considered "cut off" from its supply line and would start losing hp (or some other penalty). This would have the added advantage of being simple to understand and simple to implement.
I like the idea of integrating concepts from other games into the freeciv rules. I once thought about making land or sea mines and "detector" units that would only tell you the number of mines directly adjacent. So the players would end up playing minesweeper on the map. :D
About tech loss to create a cyclical game: a way to implement this would be to give every tech you own an upkeep cost, just like units and buildings. The more advanced the tech, the more bulbs per turn it costs to upkeep it. That way your net research (total research minus upkeep cost) can be positive and negative. If you become advanced and then turn into a raving fundamentalist you loose techs one by one until the net research is at least 0 again. If the upkeep per tech is small (like 1 bulb/turn for simple techs, 10/turn for advanced techs) then people would normally hardly notice it.
This is actually an excellent idea and would not be that hard to implement as far as I know (though incompatible with old clients of course). I'm definitely adding it to my plan for future versions.
Hex map
Book's 4 side 8 side comment brings up for me the possibility of playing hex map games. One thing is for sure, city placement is more intuitive in hex than in a square map. Perhaps it would make supply routes possible as well.
There was also a "hex
There was also a "hex amplio" tileset called aphex posted in the freeciv forums a while back that looks damn good.