New version of the no-limits add-on

[Julius Brenzaida]

I’m working on the second version of my no-limits add-on, for ToT and maybe later for MGE.
The 1st version got rid of the gold (30’000) and population limit(320’000) and extended the maps from 10’000 to 32767.
The second breaks the units limit (1948) and the cities’ limit (255).
It should also later give more barbarians and difficulties level (up to deity+5).
If you want to test, send me a mail : I can’t put the file as an attachement here (too big).

[Julius Brenzaida]

I have broken the units limit and it seems to work, but with the cities I am at lost with the units allocation. City 256 is the home of all None units and city 257 creates units that are located in city 1. 258 in city 2 and so on. Any idea on how I could get rid of that ???
I can try to change all the ff00 (256) in the programm, but it’s going to take monthes, I am afraid.
What’s raging is that the trading works !

[Seeker]

Hmmm the city limit thing interests me the most...

[William Keenan]

The problem is that the size of the city number field is only one byte. There is no way to fit a number lager than FF (255) in a one byte field.

[Mercator]

Isn´t the units limit 2048? And I have to concur with William on the cities limit. Even if you´d be able to change this, that would make the no-limits patch incompatible with all other Civ2 versions.

Oh, yes, and some notes about your current no-limits thingy. The game.txt files you included seem to be changed in more ways than just the 10,000 changed into 32,767 (which is the only change required, right?). For instance, in picking the number of civs in the extended original, it says something like "7 civs (2 aliens race)" and "6 civs (1 alien race)". I assume you made some changes to your rules etc?
And perhaps you could let the different game.txt files already be contained in their subfolders in the ZIP, so extracting the zip will automatically put them in the correct dirs. That way users don´t have to rename them first.
And thirdly, solving the previous two "problems". Perhaps you could have a small program instead of the game.txt files which simply finds all occurences of 10,000 and changes them into 32,767. That way, people´s own changed files wil remain intact and the same version can be used for all different language versions of ToT.

[Julius Brenzaida]

William, it is more tedious than that. Through hard, hard work, I’ve found the number holding the number of cities and it was FF followed by 00, I made it FFFE and now I can build more than 65000 cities. So it looks to me that the number of byte might not be a problem, as far as cities are concerned. What worries me is that units need to know where they are located and it might be a 1 byte stuff. What’s more maybe that the saved file is build for only 255 cities. Strange things happened in the 257th city after I saved and loaded again. Another tricky piece is that city 256 supports all none unit. It should be changed too. So maybe it is workable in theory, but the time required to find the good ff00 to change or the good 0000 is worrysome. I spent 2 whole days changing FF00 by the dozens before finding the good one for the city limit. And I was pretty lucky. I think it should take weeks trying before I maybe get it right. All of this for the city limit and with maybe 50% chance of succeeding, what do you think ???
But anyway, you can build more than 255 cities and the games goes on. Of that I have proof of. If you want to test it, I can send you the new version of the patch. It does work. I’ve not named it civ2.exe because it would be too dangerous to overtake the previous version. But you can click on any renamed civ2.exe program and it still plays correctly as civ2.

[Julius Brenzaida]

Mercator, the limit is 2048 for all civs combined and 1948 for one civ. I am sure of that, after all the testing I did last week. Apparently, I’ve been able to get rid of that limit. And I can build as many units as I want with the cheat mode. It would require more testing, but it is quite long to do. Any takers ???
For the slight mistake of the game.txt of the extended original in the 1st, yes you have well noticed. Of course that it doesn’t change anything in the game. It escaped my attention a year ago. And of course there was lighter ways to change all the games.txt.
Maybe for the second add-on, but if only the units limits is broken is it worth making a new patch ?

[Julius Brenzaida]

Concerrning the level modes, Deity+5 is playable in ToT, but you can’t save and you can’t get to the hall of fame, when it is possible with 2.42 and FW !!! Any idea why ? New barbarians levels work fine, though . (Fine if you like red everywhere).
And 2.42 reminds me of something : was there not something about all units being none after city 128, would it work with more than 255 ??? I should try that too. But it takes so much time. Maybe that waiting to see how much cities civ3 can afford is the best policy.

[Mercator]

I didn´t know about that 1948 per civ limit.Anotehr thing about the 255 cities limit. AFAIK, the city ID in a savegame (and the civ2 memory for that matter) is only one byte. That´s what´s causing that city 257 creates units supported by city 1. Somehow you´d have to change that one byte to two, but I don´t think that will be as "easy" as changing, say, FF00 into FFFF. Apart from that, this change will drastically change the structure of savegames etc., making them incompatible with other ToT versions. And depending on how the executable itself finds the different sections in the savegames even the executable itself might not be able to locate them correctly.
About the Deity +5, is saving only impossible at that level, or does the problem already start at Deity +3? Originally, there are 6 difficulty levels, adding 3 more, makes it 9, which can´t be saved anymore in one byte, this "overflow" may be causing the trouble.
And about the 128 cities, I think that was a problem of version 1.x only, and was solved with the 2.42 patch.

And as for whether you should release a new patch if it only lifts the unit limit, I don´t know. I´m not making scenarios and I´ve never run into the limit yet... But I´m sure some scenario makers will appreciate it.

By the way, William, how´s the Hobbit?

[William Keenan]

By the way, William, how´s the Hobbit?

The Hobbit is on the back burner. I lack time and inspiration.

However, I did see the new LotR scenario today and it is beautiful! All new graphics; inovative design; great maps. Stunning!

[Rasputin]

i sure hope you do break that city limit, in our current diplo game on the gigaworld map we hit the limit aobut 1100AD become quite silly seeing so much good land uninhabioted due to city limit...


you are doign well keep pu the godd workl

[Mercator]

Originally posted by William Keenan


The Hobbit is on the back burner. I lack time and inspiration.

Boy, does that sound familiar!

However, I did see the new LotR scenario today and it is beautiful! All new graphics; inovative design; great maps. Stunning!

Great, I´m looking forward to it. So it´ll be done before the movie then?

[Angelo Scotto]

Hello Julius, and compliments for your CSPL! (William, have a look at cradle utilities page...)

I can try to change all the ff00 (256) in the programm, but it’s going to take monthes, I am afraid.

As always, i have to agree with Mercator and William:
the real limit here is not 256 but 8, more precisely its due to data structure of units and not of cities;
The problem is that Civ2.exe reads a single byte (8 bit) from memory when searching for city number in unit data structure, if you want to make it work with 256 cities you should force it to read two bytes instead of one;

in other words

Unit data structure has a particular field (1 byte in size)

00 means unit is supported by city 0
..
..
..
FE means unit is supported by city 255
FF means unit is supported by NONE

This explains why Civ2 acts in the strange way you described:
If you simply break the limit of 255 cities you allow Civ2 to create city 256, 256 is FF in hex,
obviously when Civ2 examine a NONE unit it notice that it is supported by FF, the problem is that a city FF EXISTS in your game and so Civ2 happily decides that it is the home of all NONE units.
The same applies for city 257, 257 = 01FF (FF01 in reverse notation), this means that it uses 2 bytes to store the correct value, BUT Civ2 is programmed to read only a single byte, so from FF01 it reads only 01 and so it assigns units built in city FF01 to city 01, units built in city FF02 to city 02 and so on.
A correct approach requires to find each instruction which reads from unit data structure the home city number and change it so that it can read 2 bytes instead of one but this requires a disassembler, a debugger and a lot of rainy days outside;
It's the same approach i'm following to break the 7-civ limit in Tot, but, even if the approach theorically works it requires a HUGE number of weeks of boring work and i'll quit soon.

What’s raging is that the trading works !

The trading works because programmers of city data section were less greedy than programmers of unit section:
Look at Allard Hex Edit document:

8. CITIES INFORMATION
(source Andrew Livings and Xin Yu)
...
...
69-74 = 1st, 2nd and 3rd trade route partner city number (city number is 85th-86th position)
..

from byte 69 to byte 74, this means bytes 69,70,71,72,73,74: 6 bytes for 3 cities index, this means that city index was supposed 2 bytes long (and Allard confirms this writing that city number is in bytes 85,86).

The problem is not there, the problem is that no one warned unit section programmer which allocated just 1 byte for city number as explains always by Allard:

7. UNITS INFORMATION
(source Allard Höfelt)
..
..
17 home city with number is city number (FF is none)
..

byte 17, just one byte for city number.

Do you understand the problem now? programmers of unit section has condemned the whole game to 256 cities even if the game itself can support 'till FFFF-1 (65534) cities (THIS makes me raging! )

Greetings,
Angelo

[Julius Brenzaida]

Angelo, I told William a few days ago of the authorship mistake he made I'm sure it will be fixed soon If it is not already

You've convinced me than the 255 cities is too tough to break
I had intuitions about what you told, but what I could not fathom was that differents programmers were not on the same number of bytes for the cities. But it explains everything

[Julius Brenzaida]

Mercator, as the levels are concerned, it is hard to explain.
Already Deity+1 crashes. And until civ2 FW 2.72, it works
It is raging, too
I think the problem comes when the programm wants to calculate score, it doens't know what multiplier come with the new levels.
So my theory is the following. It was implemented but not used for civ 2.42, then forgotten, and we're only picking remains of it with ToT and MGE (where they did modify many things)
Maybe only one number change is required but I don't even know it
And maybe it has something to do with the save file, as saved files from 2.42 are different from MGE and ToT.
Anyway, I have no new idea and the solution certainly is not to be found only in the .txt files as it is for 2,42