DESIGN: CTP2 Bugs and alterations [fix list] - Page 3

Martin Gühmann · August 1, 2003, 10:25

Quote:

Originally posted by Solver
As for reloadslic, we can implement yet another command /clearvar <variable name> to nullify one variable, if need be.

The problem in the /relaodslic example I mentioned is that you have no idea which variable causes the crash. You need also the /reloadslic thing to reload strings or to reset strings, fix the messages if you modified the string files. Well for this there is a chat window command in the ctp2.exe and also some others:

pacman
toe
/rnd
/resetlos
/reloadslic
/exportmap
/importmap
/attach
/detach
/resetstrings

Well know are /reloadslic, /exportmap and /importmap, for the others it is not clear what they should do and if they work.

-Martin

J Bytheway · August 1, 2003, 10:50

In CTP1 /attach and /detach alter whether players are AI or not (In PBEM, and MP, I think... probably hotseat too), and so I guess that's what they're supposed to do in CTP2 too.

The others I would guess:
/rnd: reset the random number seed
/resetlos: reset LOS
/resetstrings: reload strings.txt etc.

But with luck guessing will soon not be necessary

.

Locutus · August 1, 2003, 10:56

/rnd [x] is let the game play for [x] turns without human intervention (in theory, at least; there are some caveats).

/resetlos is as John says and I don't think /resetstrings does anything (nothing I could ever figure out, anyway).

Martin Gühmann · August 1, 2003, 11:02

Quote:

Originally posted by Locutus
/rnd [x] is let the game play for [x] turns without human intervention (in theory, at least; there are some caveats).

That is something usefull, I usually played as Barbarian or let the game play.

Quote:

Originally posted by Locutus
/resetlos: reset LOS

And for what stands LOS?

-Martin

Locutus · August 1, 2003, 11:46

Line of Sight

Barley · August 1, 2003, 17:43

Not read all the posts in full so apologise if this is in the list, plus its not a bug.

For multiplayer games could their be a calculator or something similar that although land is randomised it ensures the starting values of say an inner circle of 60 squares and outer circle of 300 have relatively equal total values re production, food and commerce?

Is this possible?

MrBaggins · August 1, 2003, 22:04

for *just* their starting locality? probably

otherwise? probably not

LDiCesare · August 2, 2003, 04:36

I think fixing crash to desktop bugs should be priority. They moslty happen because of mods (I got one in Cradle where my save games will crash after hitting the next turn, I think there are some related to Diplomod too). Whatever the SLIC'er writes, it shouldn't cause the game to crash.

LDiCesare · August 2, 2003, 06:40

Another thing to think of: Getting rid of CD-protection stuff.

hexagonian · August 2, 2003, 08:34

Quote:

Originally posted by LDiCesare
I think fixing crash to desktop bugs should be priority. They moslty happen because of mods (I got one in Cradle where my save games will crash after hitting the next turn, I think there are some related to Diplomod too). Whatever the SLIC'er writes, it shouldn't cause the game to crash.

I'm guessing many of these crashes may be tied into the SLICS in some way. The last update (1.35) deactivated the Pirate code because of a bug in there - but tracking down that one was like finding a needle in a haystack - and does not solve any additional crashes either.

The whole nature of crashes makes them hard to figure out, because finding out what causes them is so tricky - especially for us non-programmers.

Somebody pointed out that new civs from revolts may cause a crash, which is something (probably) in the source code. Playing on gigantic maps, as well as the increased AI ability to catch slaves in Cradle, and greater happiness problems in Cradle (city caps), greatly increases the chances of this occuring, and might be the cause of current Cradle crashes. I tended to play on smaller maps, so I did not have a lot of crashes.

Another one is the end-game crash...

LDiCesare · August 2, 2003, 10:14

Quote:

The whole nature of crashes makes them hard to figure out, because finding out what causes them is so tricky - especially for us non-programmers.

Actually, with a debugger, a crash is very often easy to track down. That's why I think it should be possible to correct the crashes fairly easily, so that SLICs would be more reliable. I also think it would be fairly easy to track down what happens (many crashes are because of uninitialized memory, or 0 values where something is expected) and thus get a better understanding of the code. But then I'll probably leave that to someone else to check.

The Big Mc · August 2, 2003, 10:43

What about the fact of pillaging a tile imp with no owner surely that would be something to do with a 0 value

Martin Gühmann · August 2, 2003, 15:02

Quote:

Originally posted by hexagonian
Somebody pointed out that new civs from revolts may cause a crash, which is something (probably) in the source code.

Actual this is a problem of Diplomod 3.6 that is not part of Cradle, so it doesn't cause a crash in Cradle.

Quote:

Originally posted by LDiCesare
Actually, with a debugger, a crash is very often easy to track down. That's why I think it should be possible to correct the crashes fairly easily, so that SLICs would be more reliable. I also think it would be fairly easy to track down what happens (many crashes are because of uninitialized memory, or 0 values where something is expected) and thus get a better understanding of the code. But then I'll probably leave that to someone else to check.

With a debugger - well maybe we should also extend the abilities of the slic debugger, so that it is easier to track down bugs in slic.

Quote:

Originally posted by The Big MC
What about the fact of pillaging a tile imp with no owner surely that would be something to do with a 0 value

Could be anything else, maybe the problem is that the pillage function checks if the owner of the tileimp is at war with you and if not it makes the owner to explain war on you, unfortunatly -1 is not a vailid player, but as long as we don't have the source code it is only guesswork what it is. But I think it should be easy to track it down if you have a debugger.

-Martin

Tamerlin · August 13, 2003, 10:02

Fix

Quote:

From the FAQ thread
The player should NEVER disband units from the "Unit" screen within the city display. Yes, there is a button marked "DISBAND", but no points from the old unit are applied to the new one.

Martin Gühmann · August 13, 2003, 12:27

Another little bug, we should fix is the pollution warning message at the begin of the game if you have set MaxPlayers>0.

Well there is a slic workaround, but the message is nevertheless called and you could imagine a mod that comes with high pollution at the beginning of the game.

-Martin

Aias · August 15, 2003, 08:03

Also, give the AI civs the ability to build slavers. I don't know if this is a SLIC thing or not.

Maquiladora · August 15, 2003, 09:04

They already do build slavers

Aias · August 19, 2003, 23:13

I've never seen the AI build slavers in an unmodded game.

Also, if there is there going to be a unit upgrader included in the redone game? While playing CTC, when I upgrade a unit, it's veteran status disappears. I realize veteran status doesn't do anything, but hopefully it'll mean something in the new game.

st_swithin · August 20, 2003, 06:55

Eh, slavery is a standard part of the history of human civilization. Ever seen "The Matrix"?

Maquiladora · August 20, 2003, 07:31

Quote:

I've never seen the AI build slavers in an unmodded game.

Mostly its the aggressive imperialist civs that build slavers.

Quote:

Eh, slavery is a standard part of the history of human civilization. Ever seen "The Matrix"?

Walk into any macdonalds in the world and you can see that :P (edit: not the matrix) shame firaxis caught onto this massive part of history so late.

st_swithin · August 20, 2003, 15:30

I like your avatar Maq!!

I can't eat "food" from McD's - to me, it tastes the way masking tape smells.

Doesn't 3M own McDonald's franchises?

Maquiladora · August 20, 2003, 21:33

Quote:

I like your avatar Maq!!

I had some inspiration

I havent been to a macdonalds in years, i cant really remember chewing into any stationary

plenty of ground up grissel and plasters though.

st_swithin · August 21, 2003, 03:33

Ralph S. Mouse, at your service.

BureauBert · August 26, 2003, 05:41

I apologize for not having particpated in this forum for the last months, mainly because I have been busy with some real-life issues. Since the release of the source-code has been announced I decided to completely stop the work on my "Modern Times"-Mod, which has nevertheless developped pretty well until this point, but testing it became more and more disappointing - not because of my self-created bugs and lacks of balance, but because of growing instability and slowness due to the increased nuber of civs and map-size (testing the mod with the default values, 8 civs and 140 x 70 map, it ran pretty stable but it would obviously not be interesting for anyone to play it this way). I think no 'serious' mod can provide satisfying gameplay on the 'world's scale' of the default game.
This issue often appears in these forums as a 'savegame-bug' or similar, I would consider it as a general 'lack of capacity' somehow concerning the way ctp2 handles its data. I am not even able to guess about the reasons and about how hard it would be to improve ctp2's data handling capacity, but anything that could be done about this would be a great support for any further modding.

Locutus · October 28, 2003, 16:09

A little *bump* now the source is out.

player1 · October 28, 2003, 19:47

I think very important thing to fix is BUG in pathing algorithm.

When AI has two large stacks and one of them is fortified on the road or in the city, order group's goto function would go through the first group tile, but since both armies are so big that they can't stand at same place second army will just get stuck near the city/other unit (it will try to go though tile of fist stack, which is impossibile).

This is one of the main reasons why AI sucked a lot in scenarios like Alexander the Great.

P.S.
Other thing I would like is return of *.fli files from CTP1 (was that the extension I don't remember). Bascily in CTP1 you had much more control in modifing AI due to these files (like ability to control which government they choose, how to value specialists, when and how to change PW sliders, etc...)

P.P.S.
I don't browse these foums much now (as before 2 years), so i don't know is AI to AI diplomacy fianlly fixed as it should (in original game it was non-existant).

J Bytheway · October 30, 2003, 16:10

Error messages which don't minimise the game, and that allow you to deactivate DebugSlic on the fly...

Admiral PJ · November 1, 2003, 00:32

I would love to see CTP2 have an advanced resource system, expanding on Civ3's nice system.
You would have different resources required in different ages for units/buildings etc.
A New GREAT thing would be to have each unit /building need a different amount of resource quantity of each type. They could also need to spend resources for support(this could vary in quantity per unit too).
Resource quantities would vary greatly depending on unit type, and some special civ ability or wonder bonus factors.

Therefore you would need more Iron to build a battleship than a swordsman, and also more support iron for the ship (repairs/maintenance).

Resources such as fuels, ammo - i.e oil , saltpetre could be the only ones that are needed for support.. so that these resources become more important and to make them different from construction resources. This would also simplify gameplay.

It shouldn't be too hard to code the AI for this.. i'm familiar with some simple Fuzzy Logic AI , this could be implemented in part for new AI such as this.

Mutiple Unit Rankings would be great.. as well as the advantages they give.. how about the highest Elite units get 2 attacks as opposed to 1 at regular.. or have half attacks.. which are done at half strength, so 1.5 attacks at veteran gives a unit one attack north, and a attack east at half strength.. no ditch the fractions.. just have integers, although Civ 2 had fast attacks when road attacks only gave 1/3 a point to attack in , these rushed attacks were cool though annoying to plan for ingame.

Canals and different levels of road /railway... steam railway which can be upgraded to diesel railway (its the trains that are upgraded not the track so much, although electric track would be a different kettle of fried fish - this would be new track and trains so much more expensive to make).

I think we should be ambitious.. if we find there are major problems implementing things in code we can always cut back on plans.

Admiral PJ · November 1, 2003, 00:53

Maquiladora, Civ3 had slavery from the start, with rushbuilds using up City population points.. which is about using slavery to use up your people to work on fields (inthe manual it states the unhappiness of the people mean some also flee in disgust at the ruler) ,
but I agree CTP did slavery the way all good slavery is done.

I hope democratic civilisations never ally with slaving civs in CTP.. its very un- civilised to put up with the slave trade.

Speaking of Civ3, there where some things that were missing from this.. taken out from the old civ versions. If CTP Source could put all of the civ series and ctp features together, and some extra stuff.. it would beat the competition.

Solver · November 1, 2003, 08:38

True, and we want CtP2 to beat the competition

. Indeed, the natural resources idea from Civ 3 is, IMO, one of the best concepts to date in a Civ game. And we could build more on those...

Although, personally, I don't quite like the idea of also using resources to support units. I think support is fine the way it is (production percent in CtP, gold in Civ 3). I wouldn't like this to turn into a game like Rise of Nations, where you gather resources and use them for troops.