The Bug Fixes Thread!

[sapi]

Although it's been great to see a patch emerge from the dusty basement of CA, and although that patch has fixed many of the issues with the game, it obviously hasn't fixed them all.

What it has done, however, is provide us with a means to do so by giving us access to the game files via the unpacker.

As such, it's been decided that a thread like this could help the community get together and work on fixing the bugs that remain both for our benefit and to help CA release an official patch which incorporates these changes.

In essence, we want to create a repository of the Org's knowledge about the workings of bugs currently in the game, and, more importantly, how to fix them. This certainly isn't an easy task, but it is achievable, and I for one am sure that we'll make a significant amount of headway in the time before the next patch.

To help this thread work smoothly, several things are needed.

By far the most important of those is a unified submission method that is based on fact - it's fine to make assisted guesses as to the cause of a bug, but please say that you are doing so when you post.

Two bits of information are important - details about bugs that are currently in the game, set out like this:...
Description of Bug and how to replicate it .......
Probable Cause ........
Suggested Fix .......

...and information about methods that you know work to fix a bug, set out like this:
Bug .......
Cause ....... (this information is to pass on to CA)
Fix ..... (Make sure you include documentation - what file you've edited and what you've done - so that it can easily be merged with other fixes)
Author .....
Download location: ......


Thanks for your help - hopefully together we can make the game, and more importantly our experience of it, better.

[Kraxis]

Technically this isn't really a bug, but more likely just an oversight, that had resulted in something not working as it should.

Here I'm talking about the Forlorn Hope unit. They are half size and have similar stats to Zweihanders and cost more or less the same. Testing them out in this configuration the Forlorn Hope are woefully inadequate compared to the Zweihanders and their cost (as well as their techlevel, which is quite considerable).

Having tested them out with 2 HP they have shown themselves to be slightly weaker than the Zweihanders in combat against strong enemies, but slightly superior in combat against weak opponents. That fits their cost and techlevel perfectly and thus I can only say that it is meant that they should have 2 HP, but it was left out for some reason.

Easy to fix.
Change the '1' in the health line of the descr_export_unit file to a '2'.

[SMZ]

^ Agreed with Kraxis, based on the description and the unit size, Forlorn Hope is definantly supposed to be a 2HP unit. I made a few changes myself but for the most part I try to only make changes where the description or just plain common sense dictates the developers intended something... or just balance issues in rare cases. I was huge fan of Player1's work... so with that in mind:

1. Sherwood Archers are described as being able to hide in any terrain, but lack the equivalent trait...

add "hide_anywhere" to the list of traits

2. Demi Lancers are equipped with lighter armor for the purpose, as described, of being more mobile. However they have neither greater stamina or speed than the majority of cavalry units...

switch mount type to "fast_pony"

3. Russian, Polish and Hungarian Spear Militia are described as having long spears, but lack the anti-cavalry trait...

add "spear bonus 8" to primary attributes for weapon

4. Heavy Billmen have the same stats as Heavy Bill Militia...

increase attack, charge and defense skill values by 1 (16,6,11)

5. Venice and Sicily both get Mortars in their unit rosters, Milan for reasons unknown does not...

add "Milan" to Mortar unit description and to equivalent building description

6. Geneose Militia have the same missile attack as regular Pavise Crossbow Militia...

add 1 point to missile attack (13)

7. Retinue Longbowmen have the same missile attack as previous Yeoman Archers, while Yeoman Archers were 2 points higher than the previous Longbowmen...

add 2 points to missile attack (10)

8. Egypt despite being an advanced and powerful state has no high level artillery...

add "Egypt" to Monster Bombard unit description and to equivalent building description

9. Janissary Heavy Infantry while expensive, late developments and descriped as being trained to be the greatest infantry in the world - have pitiful stats. English Billmen, essentially farmers turned soldiers, consistently beat Janissary Heavy Infantry; crack regiment soldiers who have gone thru grueling training since they were small boys...

increase attack, charge and defense skill values by 3 (15,7,13)

10. The various Knightly Orders are basically identical. Knights of Saint John buildings give a health bonus, and Teutonic Knights have maces as secondary weapons. However the Templar, Teutonic and Knights of Santiago Chapter Houses have no equivalent bonuses. Templars are descriped as running a commercial empire in addition to their warlike abilities. Teutons are described as being noted for their success in forcefully converting the populace. Knights of Santiago are described as having a strong maritime tradition. The Templars, despite being described as fierce fighters, and having a charge without orders trait - do not have their ferocity reflected in their stats. For a little flavor, it would be nice to differentiate the Knights of Santiago a bit as well...

add "trade_base_income_bonus bonus" trait to Templar buildings; add "religion_level bonus" to Teuton buildings; add "navy_bonus" to Knights of Santiago buildings; add 1 point to Templar Knight unit attack, charge and secondary attack (14,9,15); replace "hardy" trait with "very_hardy" for Knights of Santiago unit

11. Kataphractoi are severly handicapped for supposedly being elite cavalry, and described as being very heavily armored. Tsars Guards in Russia, while equipped basically the same have significantly higher attack and defense...

increase attack and defense values by 1 (11,17)

12.High level Castles and Cities have the same number of recruitment slots, despite one of the main supposed benefits of Castles being rapid troop production...

increase Fortress recruitment slots to 4; increase Citadel recruitment slots to 5

13. Northern Peasants are too good, Southern Peasants are too bad...

decrease Northern Peasant attack and defense by 1 (3,2); increase Southern Peasant attack and defense by 1 (2,2)

[Gray_Lensman]

Re 1 above, It might possibly be an incorrect description rather than the ommision of the hide_anywhere trait.

Scanning thru the EDU file only has 2 other units with the hide_improved_forest trait, and both of them already have the hide_anywhere trait. They are Battlefield_Assassins and Hashishim. both of which are referred to as assassins. This tends to make me think they are the only ones that should be capable of the hide_anywhere trait. (being assassins)

However, I am curious what the difference is between hide_forest, and hide_improved_forest traits

[sapi]

Hmmm - just guessing, but maybe it's the range that they can be spotted at?

Either that or improved_forest is long grass?

[SMZ]

you're right, editing the descr for Sherwood Archers would probably fit the unit idea better... anybody know how to do that?



1. The higher level agent Guild buildings do not appear to have their triggers functioning correctly. An Assassins HQ for instance will give an assassin trained in any city +2 to skill, however it will not also give the local +1 bonus that a 1st level guild building does. Thus, counter-intuitively, better spies, etc are produced in the lower level city. It appears the 'greater than or equal to' sign is not working properly...

Duplicate the 1st level trigger, and tie the duplicates to the 2nd and 3rd level agent guild buildings

2. A Bishop ranking is given to priests produced from a cathedral, however no trigger exists to cause the trait for huge cathedrals...

add a ">" to the priestinit4 trigger or duplicate the trigger and switch out "cathedral" for "huge_cathedral"

3. Pagan Magician ancillary spawns entirely too often, due to the trigger simply calling for a province with less than 90% your religion - which means basically every province, especially in the early game...

change trigger to fire when "PopulationPagan > 10"

[sapi]

Oh, and to clarify #2 - a bishop is created in any city with a cathedral; but it's broken as only the cathedral gives it, not the grand cathedral as well

[Kraxis]

Yes, just add a '>' to the line and it should work.

[SMZ]

sorry, I should've tested those last few before submitting, but the last #3 I posted doesn't work... any clue as to why that command line doesn't function? I saw a "PopulationHeretic" command line somewhere, so I figured there should be equivalents for each religion... connecting the appearance of a Pagan Magician to any other criteria seems a bit non-sensical...

1. unlike all other merchant guild HQ's - the English version does not give a +1 to experience for Merchant Cavalry trained there. I am unsure if this is supposed to reflect englands weaker cavalry, or if it is an oversight, thoughts? in any case, if it is an oversight...

replace the "0" with a "1" in the equivalent merchant cavalry listing within the building list

[kleemann]

Well, the AI has several bugs in it.

BUG: AI stands still on battlefield when your city/castle is under siege and you make counter attack from city with one missile unit or several ones. It just stays thare and lets you to flank and bombard them until all missiles are gone.

So you just go shoot and kill many of AI-s units and AI don't do anything.

Cause: I guess the reason is lack of trigers for this kind of playes side action (this information is to pass on to CA)

FIX:
Add trigers to AI:
Simpler fix: if the situation is same as described above.

1. AI should send his own missile units to counter fire - actually it works.

2. If AI-s missiles are low attack with cavarly. Better if AI attacks enemys missile units with cavarly in the beginning.

3. If AI don't have missile units or cavarly attack with infantry. It is better than letting his units to shoot death.

4. If AI has siege equipment. It should attack walls to get breach and attack city. Use the same action as AI attacks city.

5. If player shoots AI's army and the player has bowmen and archers on wall. AI must move away from missiles range, so that the missile unit has to pursue them to shoot and moves away from walls defending fire. Then AI should counter attack, so that it can kill the missile unit. If it has cavarly it uses cavarly if not then infantry.

Barbarian Invasion has this kind of AI, and it works fine thare.

The idea is that AI should not let his units to bombard.
A) it should move away
B) it should attack

No documents modified yet..

Author .....
Download location: ......

[CeltiberoMordred]

Bug: Cavalry, when charging, kill whole infantry units in two seconds.

Causes: CA "feature".

Fix: Raise the mass of every unit by 10 points in your export_decr_unit.txt file ("soldier" line).

[Whacker]

Seeing how he hasn't got his lazy rear in here to post this, I'll do it for him... Geez some people.

Kobal2fr's outstanding fixed VnV's: https://forums.totalwar.org/vb/showthread.php?t=75505

He's listed it as a mod, but I prefer to think of it as more along the lines of bugfixer for RTW/BI. There's some subjective work in it, for the most part it's just real bugs being squished and tweaked, but that's really no different than bugfixer either.


I should also mention the.seneshcal's Medifix mod, which does have a big more subjective work done to rebalance issues. Either works well depending on what you prefer: https://forums.totalwar.org/vb/showthread.php?t=75326 Also gets .

[Fisherking]

There is still a bug with missions from Noble counsel. When you assault a city or castle they have order taken without sieging it for one turn, you still don't complete the mission. Also noticed joining a crusade on the last passable turn didn't accomplish the mission. They are small but a little bothersome.

[Quillan]

Many, many of these are subjective judgements rather than true bugs. That noble council mission specifically tells you to send a diplomat to secure ownership of the particular province from the other faction, not an army. Janissary Heavy Infantry perform much better on the battlefield than their stats indicate, much as Varangian Guard perform much worse than they should according to their stats. Retinue Longbowmen aren't better archers than Yeomen, but they are armored. That's the difference.

The trait/ancillary triggers are fairly obvious mistakes. Some troop stats are wrong.

[SMZ]

I don't care how good JHI's perform. They still get beat by English Billmen, which are the same basic troop type, just better... which does not make a lick of sense

subjective? yes - but it's a glaring error nonetheless - either the billmen should suck more, or the JHI's should be better - take your pick

secondly, every other line of troops sees the higher level forms increase their attack power - ie: Heavy Billmen aren't just armored versions of Billmen, they also have a higher attack rating to reflect that they are the more advanced form of the unit - Heavy Pike Militia aren't just Scots Pike Militia given armor, their attack increases as well - etc, etc, etc across the board... it makes no sense to thus call Retinues simply Yeoman given armor

as the description says, Yeoman are freeholders expected to train in their downtime and be ready for war... Retinues are essentially mercenary bands who's business is war - their training is better... they should be better archers

[Schugger]

Maybe it is just me, but something seems odd about how the game is calculating turns (one turn equals two years, alternating seasons) and the age of characters (they age just by one year each turn).
Wilhelm "Methusalix" the Conquerer deceased in my campaign in 1130. Considering that he was 50 in 1080, he almost became of biblic age.
Known issue?

And merry Christmas, guys

[Foz]

Maybe it is just me, but something seems odd about how the game is calculating turns (one turn equals two years, alternating seasons) and the age of characters (they age just by one year each turn).
Wilhelm "Methusalix" the Conquerer deceased in my campaign in 1130. Considering that he was 50 in 1080, he almost became of biblic age.
Known issue?

And merry Christmas, guys

I agree, it feels slightly strange, but I think it's done to maintain game balance (especially in traits), since the passage of time actually IS a setting we can control. IIRC the default is the 2 years per turn that you described, but we can also set it for 1 year per turn and some other one I can't remember. Since buildings go up in a given number of turns, and the trait system is based primarily on per-turn calculations, characters always have to age at 1 year per turn, or else different timescales would mean your characters had a disproportionately short or long length of time to develop, which would make those time scales far easier or harder, correspondingly, from a character usability standpoint. For instance if your characters aged as fast as the years in 2-year-per-turn mode, they'd live about 30 turns, see far less buildings built under them, develop less traits and to a lesser degree of development, and have difficulty earning anything near the max number of command (or any other stat) stars. Similarly, Agents in 6-month-per-turn mode would live insanely long 120-turn-ish lives, and would be uber in no time. So I think it is that way to keep the characters developing at the rate CA intended, but still allow us to make time pass more slowly or quickly, which affects when inventions and events come about, as well as how long we have to accomplish the objective.

[DukeKent]

Hope this is the right place to list this?

Description of Bug and how to replicate it .......
English Princess charms a Commanding French General in a stack of units inclusing the presumed future French Faction Heir Henry. The Charmed former French General, and all units in the former French Crusader stack can be combined with other English units as expected. Attempts to combine the future French Faction Heir caused a CTD on that turn. As an unecpected bonus I got about 18 French crusader units. Best Diplomatic/Charm deal I ever sas in any MTW game I do not believe that the French family member (Future Faction Heir) was designated as a "Faction Heir" until a few turns after the General stack he was in were charmed by the English pricess. I could move this now red English single unit stack around. But a mouse over on the unit shows this unit as the "French Faction Heir". Attempts to move the single unit stack containing the French Faction Heir cause a windows popup error, then a CTD. The log shows the following entry in it.

00:30:59.164 [system.rpt] [error] Medieval 2: Total War encountered an unspecified error and will now exit. Should I up the logging level and try again? Had anyone else seen a similiar CTD. My current work around is to just leave the unit alone.


Probable Cause ........
Confustion on the MTW programs part about allowing a French Faction Heir in an English unit stack. Caused by a princess charming.

Suggested Fix .......
Probably a CA code fix. Smells like a code bug to me. Interesting how CA does its mod txt files. Like no programming, or data base I have seen in the business world

[DukeKent]

Update, the charned French general ended up going rebel. Attemps to attack the unit caused a CTD Even from a backup before the attack on the rebel general. That pretty much ended the campaign

[Wizzie]

Maybe it is just me, but something seems odd about how the game is calculating turns (one turn equals two years, alternating seasons) and the age of characters (they age just by one year each turn).
Wilhelm "Methusalix" the Conquerer deceased in my campaign in 1130. Considering that he was 50 in 1080, he almost became of biblic age.
Known issue?

And merry Christmas, guys

This is not a bug. It's a feature, or perhaps better described as simply the way the game is supposed to work. Since there is such a massive time scale covered in this game, for it to progress at a reasonable rate while still activating all events like Gunpowder, the Americas, etc. in a reasonable timestamp, a turn rate of two years per turn was implemented. However, if characters aged at this rate they would die off in 1/4 of the time they did in Rome. Since character stats and progression mechanics etc. are presumably hardcoded in the game engine, it made more sense to maintain the Rome aging rate of 1 year for every 2 turns - in order for good generals to develop before dying of old age. This has resulted in an abstract aging model which CA left easily moddable.

[Bobalude]

Description of Bug and how to replicate it: Can't build watchtowers or forts / limited movement. Don't know how to replicate, just happens on some regions.
Probable Cause: Disaster event got stuck or something so the game thinks its still active many turns later.
Suggested Fix: Wait or cause another disaster in the region that clears up correctly.

Working Fix (workaround): Edit the disasters file to force a disaster event to occur, run the saved game for a turn or two so the event happens, save, and revert the disaster file back to the original.

The file is: data\world\maps\base\descr_disasters.txt

Make a copy of the file first, then edit the original and make a copy of the first entry, for earthquake, and delete the "climate" line. Replace that line with "Position" specifying the affected region. You need to lookup the position of the city with the affected region using "show_cursorstat" in the game console with the mouse cursor over the city.

Change min/max scale to 1 to reduce the effects of the disaster, and the year/frequency from 20 to either 1 or 2. I'm playing with 2 turns/year (0.50 time scale), so if you are using the default 2 years per turn, I would try 2 instead of the 1 that I used. These numbers appear to be in years, so with a 0.50 timescale, or 2 turns per year, it takes 2 game turns for it to occur.

Save the file (its read-only so you have to undo that property). Load your game, and run a turn or two. Hopefully an earthquake will properly hit and undo the watchtower/fort bug. Save post-earthquake and then go back and restore the original descr_disasters file.

--------
Lastly, I noticed that this oddly didn't work in the odd game turns using 0.50 timescale. That is if the saved game was during a particular season or 1/2 of a year, the disaster event will never trigger for some reason. Simply go back to a saved game 1 turn earlier seems to fix it. Guessing it only counts properly either every even or odd turn number... or if the game starts at a half year, it never triggers (i.e. year 1110.5).

[sapi]

GrandpaTroll, i've moved your post to the Apothecary - you won't be able to post there (pm me if you need to say anything) but you'll get much more help.

[GrandpaTroll]

GrandpaTroll, i've moved your post to the Apothecary - you won't be able to post there (pm me if you need to say anything) but you'll get much more help.

Thanks

Gramps

[Radioman]

fast_pony is typed as, fast pony in data file...

[Foz]

I've proposed a temporary (i.e. imperfect) fix for the bug where shields incorrectly detract from a unit's defense in combat. I believe it to be the best fix at our disposal, borking the least other things in the game while still fixing this otherwise game-breaking problem. You can find it here: https://forums.totalwar.org/vb/showp...&postcount=110

Now version 1.1 with a slightly different fix. The link above still gets you there.

[Zuraffo]

Hi, I would like to recommend two fixed I found in another forum (are they against the rules?)

zxiang's fix on 2H-Axe:
http://www.twcenter.net/forums/showt...t=74088&page=4

and zxiang's fix on the ballista/cannonball tower:
http://www.twcenter.net/forums/downl...do=file&id=973

No doubt the vets have already known about these fixes, but it's useful to have it somewhere easy to search for newbie like me. TY

[FactionHeir]

High level Castles and Cities have the same number of recruitment slots, despite one of the main supposed benefits of Castles being rapid troop production...

increase Fortress recruitment slots to 4; increase Citadel recruitment slots to 5

The higher level agent Guild buildings do not appear to have their triggers functioning correctly. An Assassins HQ for instance will give an assassin trained in any city +2 to skill, however it will not also give the local +1 bonus that a 1st level guild building does. Thus, counter-intuitively, better spies, etc are produced in the lower level city. It appears the 'greater than or equal to' sign is not working properly...

Duplicate the 1st level trigger, and tie the duplicates to the 2nd and 3rd level agent guild buildings

Those fixes would be nice if you actually detailed which file these are in. Otherwise I take it you only meant them as suggestions but don't actually know how to do it.

[Zuraffo]

I am posting the link to a patch mod that fixed most of the problems

http://www.twcenter.net/forums/showt...=72874&page=16

This mod however, did not fixed the spy guild problem. You can get this mod to fix all the other problems and get Kobal2fr's outstanding fixed VnV's (posted earlier in this thread but I c/p for completion) : https://forums.totalwar.org/vb/showthread.php?t=75505

Just copy the text file to the litemod data folder. If yoou own a good text editor you can also try to compare and merge the two.

[Ring_Master\]

Easy to fix.
Change the '1' in the health line of the descr_export_unit file to a '2'.

Gahhhh!! I can't seem to find the descr_export_unit file no matter how hard I try.. could someone please direct me to it's direct location within the game files? Any response is fine as long as I can understand it. But seriously though, this simple problem of mine is starting to get on my nerves.

[FactionHeir]

Use the unpacker and read its readme.