VCMI 0.72c - development version

#31 H/next hero button – should switch between heroes with path left not without

#32 if load saved game with garrisoned hero in castle and you enter any of other hero to the castle there always show the meeting screen with the hero even if this hero is already out of the castle (added saved game)

#33 if you change the resolution in console (not setting file) during game play the screen somehow changes: the battle screen, the frames of main window, and there are problems and crashes if you swith between full screen and window mode. I think it should change after new game restart/load and the new resolution should be default and saved in the setting file

I think it is ok. Spacebar switches between underground and surface.
bug_32.zip (70.8 KB)

Hi Zamolxis. Actually, the Badge of Courage description does correspond to the H3 functionality. Whether or not it was intentional on the part of the original H3 coders (it seems to be a fairly powerful ability for that class/value of artifact), mind spell immunity has always been an undocumented feature of the Badge of Courage, which is why we eventually added it to the WoG description (so everyone would know).

In any event, I think the descriptions for artifacts are all stored in editable text resource files.

However, the WoG description says hostile mind spells. I believe it’s all mind spells that are blocked, including Mirth and Frenzy (the two beneficial mind spells), unless it was changed to hostile-only in WoG (unlikely but never know).

  • Fnord

Hi Fnord. It’s good to see you. :slight_smile:

Thanks for the information. I didn’t know that. Actually, while reading your post, it started to ring a bell… I may have probably read that long ago on some WoG forum, but forgotten about it since.

Well, I guess that makes #1 a non-issue. :wink:

Thanks. :slight_smile: Yes, except that the word hostile should probably be removed. :slight_smile: - Fnord

34 Game crashes if hero is in the “school of magic” and another hero comes to transfer creatures to him.After closing meeting screen apears screen from school of magic and game crashes with comment: “R6025 - pure virtual function call” (see attached save). The problem is caused probably because for meeting screen one hero is on the other but should stand next to him. Similar situation is with attacking hero/creature.
crash_34.zip (132 KB)

Correct… I’ve double checked in H3C just to be 100% sure I don’t give misleading reports (again:p), and “hostile” should be removed from the artifact description. I’m gonna update my original report.

//
#35 Heroes in town do not always learn the spells at Mage Guild (even though they have the Skill and the Spell Book).

WORKS:

  • If we build Mage Guild level while hero is in town
  • If hero enters town at the end of its movement
  • If hero enters town with Spacebar

DOES NOT WORK:

  • At game start (if Mage Guild is built & hero has spell book, unless we use Spacebar or movement to enter town).
  • If the hero gains a higher Wisdom skill while at town gate (i.e.: siege), while in town there is a high level Mage Guild built (again, unless we enter town with Spacebar, or later on at movement end)

To easily reproduce it, simply load the attached map and do the following:

  1. Open Neela’s hero screen and check the spell book: only 1 starting spell
  2. Enter town with double click on town or town icon. Open Mage Guild to see available spells.
  3. Double click on Neela’s portrait in town and check the spell book: she still hasn’t learned the spells
  4. Exit town screen and re-enter with Spacebar. Check Neela’s spell book: now she knows the L1/L2 spells
  5. Build Mage Guild L3.
  6. Exit town screen and visit the Scholar to get Basic Wisdom
  7. Enter town with double click on town or town icon.
  8. Check Neela’s spell book: she hasn’t learned the new L3 spells yet

Note: To fix this so that it works like in H3, you only need to make heroes learn new spells whenever they visit the town, regardless of the way you open the town interface.

Perhaps ideally would be to implement it as such that heroes can learn spells when at town gate, without the need of opening the town screen:

  • at map start or beginning of new turn after a siege, check the spells in Mage Guild
  • then check hero’s Wisdom skill to see if he could learn any new spell from the Mage Guild
  • and only after that the new turn autosave to take place, and hero is free to move on
    Though I can see how that would require pretty ugly code, plus potential risk of new bugs. Probably not worth the effort. :stuck_out_tongue:
    Spell Learning.h3m (8.56 KB)

#36 Killing a stack of our own during its turn with a spell, leaves a ghost image of the stack on the battlefield:

  • the stack does not have a number box anymore (see screenshot)
  • the creature can still move (as it was its turn before the spell), but just this once
  • afterwards other creatures can simply walk over it as if it wasn’t there anymore
  • if it’s the last stack of your own that you kill during the battle, after you move the ghost creature the game crashes

It’s easily reproducible if you load the “Spell Learning.h3m” map attached in the post above, then attack with Sir Mullich and use area spells on your own creatures during their turn. Otherwise please see the client log attached here below.

//
#37 Haste gives speed bonus to War Machines. When their turn comes, we have an area of shaded hexes around them, as if they could move. If we try to move them, the game crashes (easily reproducible - but I’ve attached the logs as well just in case)
090725 - Own stack killed by spell during its turn - log.zip (2.23 KB)
090725 - Crash when trying to move ballista in battle.zip (2.59 KB)

#38 Wrong growth of creatures. See attached save, there is beginning of new week all creatures was bought out at the end of the week.

#39 Moving hero through the mountains. (the same save - hero Axsis)
img182.imageshack.us/img182/3525/clipboard01hjz.th.jpg

#40 Wrong stack queue in the battle if there are creatures with the same speed first attack all creature the attacking hero then opponent, ie: if you attack creatures with speed 6, 5, 5 and 5 against creature speed 5, after creature with speed 6 should attack opponent and then rest of your creatures, but now is that attack all yours creatures and then opponent.
bug_38.zip (128 KB)

I think I managed to figure this one out. And it has nothing to do with the size of the target creature. :stuck_out_tongue:

Steps to reproduce:

  • Load the “Spell Learning” map attached to my post above (3 posts back)
  • Attack Galthran with Sir Mullich
  • Cast Haste in the beginning of the battle, then use Wait on all your creatures
  • Enemy will start moving all the Skeletons as if their only target is the Pikemen (and let’s give this issue log #41, because AI should try to go after the shooter or the flyer, unless the pikemen would be reachable in less number of turns)
  • Move the Archers as in below screenshot, so that they are 4 hexes away (Skeleton Speed) from the nearest Skeleton which was heading for the Pikemen
  • Then move the Pikemen & Griffins somewhere behind the Archers
  • Next turn start by making all your creatures Wait
  • When enemy’s turn comes, he will move the skeletons as in the below screenshot, including the one that went just in front of the Archers, without attacking them

Note that the Skeleton didn’t stop in front of the Archers w/o attacking to block their ranged ability. I’ve reproduced the same scenario switching Archers with Griffins or Pikemen - the Skeleton also stopped just in front of the Griffins/Pikemen w/o attacking. I’m not yet sure, but it seems they only can’t recognize the creature in the next hex at the end of their movement on the horizontal. That because I tried to reproduce it from other angles, and then they did attack me if I was placed on the diagonal. I hope this information helps you find the exact reason behind the bug.

i4.photobucket.com/albums/y104/Zamolxis/VCMI/VCMI%2007/th_090725-Reproduce07273.jpg [size=75](See also the movement area of the Hasted Ballista reported as #37) :wink: [/size]

#42 Monolith bug - you can block the monoliths.
If hero is in monolith and you want to transfer creatures with other hero the meeting screen appears but you can not transfer anything because hero is already teleported to other location.
img36.imageshack.us/img36/3364/clipboard01ecl.th.jpg
Second thing if one hero is in monolith and you want to enter monolith in different location you are unable to do it, there should appear the meeting screen
img41.imageshack.us/img41/1385/clipboard02qic.th.jpg

Attached save for this bug
bug_42.zip (157 KB)

We don’t want do alter original WoG descriptions. Instead of that I’ve modified Badge of Courage bonuses to follow WoG description (blocks only hostile mind spells). It’s a little change to game mechanics IMHO making sense.

Fixed.
[However hero list order is not stored when saving game, it’s known issue.]

Current formula gives 1/24 chance per +1 luck (up to 1/8 with +3 modifier). Which AFAIK follows H3.
random.org/analysis/dilbert.jpg

Fixed.

Fixed.

Fixed.

I’ll leave it without changes for now.

I’m unable to find graphics of grayed out slider arrows.
Has anyone seen them?

We’ve used formula given in Strategija (pp. 22 - 24, link here: heroescommunity.com/viewthread.php3?TID=27539 )

It’s somewhat random, if creature is set to “aggresive” its internal “mood” level set randomly to 1 - 10. It’s a big spread, so there is possibility that creatures will be as aggresive as if they were set to “savage” in editor, but they may also be “friendly”.
Depends on luck.
So, I see no bug here. (unless the H3 and VCMI behaviour is always different?)

Fixed.

Fixed.

I’ll make big changes to pregame (Main menu screens, scenario/save selection, etc) soon after 0.73 is released. Then I’ll fix pending issues and add many of the missing ones.
This code is very old (written for VCMI 0.2) and should receive big changes.

I’m not observing this, maybe was fixed somehow.
Btw, VCMI gives currently 40 - 90 exp points. Have you observed higher/lower values in H3?

Fixed.

The snow version of the Subterranean Gate is causing problem. WIll be fixed by 0.73.

Fixed.

Can’t open, save format has been changed :frowning:

Is this different from H3 behaviour?

Makes sense indeed… I never much liked the artifacts that give “resistance” to benefic spells on friendly troops anyway. :stuck_out_tongue:

If my understanding is correct, the remaining bug is not such a bad thing for the time being: we can check whenever we load if the newly recruited heroes are ordered correctly as per alignment. :stuck_out_tongue:

LOL :smiley: - Dilbert rocks. :stuck_out_tongue:

Though I still have some doubts/questions regarding the way it’s (supposed to be) implemented:

a) Does “up to 1/8 with +3 modifier” mean Expert Luck makes Luck bonuses from artifacts & fountains obsolete? Theoretically it doesn’t make much sense, but if that’s how H3 implemented it, there’s not much I can say (but we still need to make sure that’s how it was in H3)

b) But even with the “up to 1/8” rule I had a hero with Expert Luck (+ another 3 bonus luck from artifacts/fountains - if that matters at all) who got only once luck in 3 rounds for a full army. As Luck should kick in for retaliations as well, on a 7 creatures army, all attacking + some retaliating, it means at least 1 lucky hit per round. It can be that I missed one of the VCMI (very) short luck animations at some point, but I doubt it… I’ll keep a closer look it. I’m suspicious because same as spell damage was implemented using correct values on the paper, but in VCMI it got tripled somehow, it can be that you started from the correct values to implement this, but for some reason it doesn’t work like as desired just yet. I’ll keep a closer look at it.

c) And lastly - just a wild assumption based on the behavior I (subjectively) noticed in H3 : Could it be that NWC implemented a more complex formula, to give a higher chance that Luck kicks in at least once during the first round for a full army? For example - in case of Basic Luck (1/24) - something like double the chance for Luck to occur during the first 12 chanced of the 24? As said - just a wild guess. :stuck_out_tongue:

I’ll see if I have time to prepare/test some special study cases, but I’m curious of your feedback on the above until then…

Any chance a tech-n00b like me would be able to find them, if told at least where to start looking for them? :stuck_out_tongue: :blush:

After trying to reproduce it in a test map (really identical hero/army/stats/terrain/etc) without success, I agree that this is most probably a non-issue. Btw, thanks for the link - lots of useful/interesting H3 details in that .pdf :wink:

Looking forward to the big changes. :slight_smile: - Though probably they’re mainly at code level, so I won’t see much on my side. :stuck_out_tongue:

The VCMI values are indeed correct. No idea why I had the feeling I’ve seen others as well. But I retested now 10 restarts on a 8-hero test map (so 80 checks) and the extreme min/max were 40 & 90.

Regarding the bug, I’ll have to check in 0.72d with the same map (to see if it’s either fixed, or there’s sth about that custom map causing it)

I’ve opened it in my 0.72c. The bug is obvious: +50% Growth bonus for all creatures, as if he would have the Grail (probably he didn’t notice it’s exactly +50% because some of them were also influenced by hoard buildings). The Castle interface was displaying the correct growths according to Castle upgrade + hoard buildings, but at week start he was also getting the Grail bonus - i.e.: 53 Imps (30+8+15) i/o 38, 3 Devils i/o 2, etc. I’m not sure if it’ll help, but I’m attaching here the logs I got after playing his saved game.

This needs to be monitored in 0.72d/0.73 I guess, as we don’t know if the bug was general, or triggered by one particular action taken during the game.

That could be indeed a small mistake from the map editor. The only way to check it is by loading it in H3 as well.

@ Boulie : Could you please say which map were you testing? And if it’s a custom one, could you either test it in H3 or attach it to a post?

@ Tow : (Hope I’m not getting annoying with this) :stuck_out_tongue: This reminds me again of my suggestion to have the Scenario Information implemented sooner rather than later (it’s somewhere in the Missing features thread). You never know when a tester might come back to give further information like on which map a bug was found, but with the saved game anyone of us could have identified the map and check if it’s a map issue or a bug. :wink:
0.72c - Grail growth without Grail.zip (2.77 KB)

This problem is general. I’ve attached save with other map (ver 0.72c if someone can still open it) and also I’ve noticed it before.
@Zamolxis - i’ve noticed it’s as there was grail built but I did not want to suggest what can cause the problem because I wasn’t sure.

I’ll check it later. The map name is “Island King” (attached)
*
EDIT:
Already checked. It’s map bug.Sorry,my mistake.*
bug_38ver2.zip (155 KB)
Island King.h3m (35.1 KB)

Page 44 of the HoMM3 manual lists the effects of luck. A luck of 1 should give a 4.2% for maximum damage, a luck of 2 should give 8.3%, and a luck of 3 should give 12.5%. That’s 1/24, 1/12, and 1/8 respectively.

I’m pretty sure that that’s how luck was implemented in HoMM3, there are no additional benefits if luck>3. Higher luck will probably work as a buffer against the Misfortune spell (i.e., having a luck of 5 means that an enemy could cast expert misfortune on your troops, which reduces their luck by 2, and you’ll still have the full effect of maximum luck). I haven’t actually checked that though.

That makes luck a “weak” skill, which is the reason why it has been buffed by WoG’s script48 (“Enhanced secondary skills”), which adds a minor increase to attack and defense of 1-3 stacks based on your luck skill.

It’s not impossible, but I doubt it, it seems unnecessarily complicated and even if implemented that way, there’s no reason not to inform the player about it. The theory also doesn’t conform to my own observations when playing the game, although I have to say that I never monitored the luck effect closely. That’s not to say that there aren’t any hidden effects on luck - I wouldn’t rule out a hidden game difficulty modifier that boosts AI luck chances, for example -, but it wouldn’t be easy to track down. We’d need a testbed and keep tabs - so far we’re both only presenting anecdotal evidence, and personal impressions can easily be mislead by our minds’ tendency to focus on unexpected or special results.

Fixed,

Fixed.
It was case only on RoE/AB maps, that’s why I thought it had been already fixed.

It works on SoD/WoG maps. RoE/AB map formats doesn’t make this possible, they use only one value for hero experience and it’s set to 0 if exp is not initialized. Then we cannot say if experience is 0 because it was set to 0 or because it was not set at all.

Fixed.

I think it’ll be possible after new pregame is done. In the current save format the map name and description is at the beginning (after ~24 initial bytes) of vlgm1 file, it can be read quite easily.

There is also another factor making luck a “weak” skill, when compared to Leadership.
Negative morale can cause stack to loose it’s turn. Moreover, the chance for it is 1/12 per -1 point (max -3 penalty applies), so it’s twice “stronger” than positive morale/luck effect.
The negative luck effect was originally planned by NWC but for some reason this idea has been abandoned. There are several texts in game that mentions that negative luck can cause stack to deal only the half of damage.
Maybe this effect would make Misfotune spell a reasonable choice in battle. Now I never use it and I don’t think it has any sense. There are practically always better alternatives.

Fixed.

Fixed.

Fixed.

Fixed.

Fixed.

Fixed.

Fixed by Onion Knight.

Fixed more or less. Applied original fonts to Marketplace window.

Fixed.

Fixed.

Fixed.

Fixed (but please test in the next release if in all cases).