Bug Reports (snapshot builds)

[Sloth]

I will try to fix max hand size.

EDIT: Done.

[friarsol]

Hey Arsenal,

I saw you just made this change.

http://svn.slightlymagic.net/websvn/rev ... &rev=18469

Are you sure that's right? I think some code depends on the "base SA" being checked before moving up to the Parents.

[Sloth]

Hey Arsenal,

I saw you just made this change.

http://svn.slightlymagic.net/websvn/rev ... &rev=18469

Are you sure that's right? I think some code depends on the "base SA" being checked before moving up to the Parents.

Yes, i'm sure as well that some cards will not work this way.

[ArsenalNut]

Hey Arsenal,

I saw you just made this change.

http://svn.slightlymagic.net/websvn/rev ... &rev=18469

Are you sure that's right? I think some code depends on the "base SA" being checked before moving up to the Parents.

Yes, i'm sure as well that some cards will not work this way.

The Ulvenwald Tracker fight ability was not working because the getParentTargtingCard method was returning the target selected by the subability rather than the parent Pump ability. This card used to work and I assumed somebody changed the getParentTargtingCard method incorrectly. From the method name, I assumed the method should not check the "child" SpellAbility that was making the query. There should probably be two sets of methods:

getSATargeting<target class> :
these methods would check the "child" SA before moving on to the parents

getParentTargeting<target class> :
these methods would check only parent SA's

[Sloth]

The Ulvenwald Tracker fight ability was not working because the getParentTargtingCard method was returning the target selected by the subability rather than the parent Pump ability. This card used to work and I assumed somebody changed the getParentTargtingCard method incorrectly. From the method name, I assumed the method should not check the "child" SpellAbility that was making the query. There should probably be two sets of methods:

getSATargeting<target class> :
these methods would check the "child" SA before moving on to the parents

getParentTargeting<target class> :
these methods would check only parent SA's

Yes that would be a good idea. Ulvenwald Tracker should be changed to something like "Defined$ ParentTarget" which leads to getParentTargeting and the normal "Defined$ Targeted" would lead to getSATargeting.

[friarsol]

getSATargeting<target class> :
these methods would check the "child" SA before moving on to the parents

getParentTargeting<target class> :
these methods would check only parent SA's

Yes that would be a good idea. Ulvenwald Tracker should be changed to something like "Defined$ ParentTarget" which leads to getParentTargeting and the normal "Defined$ Targeted" would lead to getSATargeting.

Can we make sure this gets fixed before Release? I'm pretty sure there's dozens of cards this change affected as opposed to the one that gained something from the change.

[ArsenalNut]

getSATargeting<target class> :
these methods would check the "child" SA before moving on to the parents

getParentTargeting<target class> :
these methods would check only parent SA's

Yes that would be a good idea. Ulvenwald Tracker should be changed to something like "Defined$ ParentTarget" which leads to getParentTargeting and the normal "Defined$ Targeted" would lead to getSATargeting.

Can we make sure this gets fixed before Release? I'm pretty sure there's dozens of cards this change affected as opposed to the one that gained something from the change.

I'll get it squared away tonight.

EDIT : fixing the issue with Defined cards is fairly straightforward but some of the of the other areas that use the getParentTargeting methods, like calculateAmount, are not. I am going to revert my changes that fixed Ulvenwald Tracker for now so they don't break the release. I am going to post some observation on this bit code over in the Developing Bugs topic for further discussion.

[Chris H.]

Can we make sure this gets fixed before Release? I'm pretty sure there's dozens of cards this change affected as opposed to the one that gained something from the change.

 
I'll get it squared away tonight.

 
I can delay the next beta release until this issue is taken care of ... no need to rush.

Just let me know when we are ready.

[Agetian]

I've seen a message somewhere above that the AI couldn't properly use Crypt Rats because it didn't obey the mana color restriction, I tried to take a look at the code to see where the proper restriction could be inserted but couldn't find anything useful, the AI code is, for the most part, a mystery for me at the moment, sadly... Would someone be able to take care of it or should Crypt Rats maybe be flagged "RemAIDeck:True" for the time being if it's too difficult to fix before the release?

- Agetian

[Corwin72]

forge-1.3.2-20121129.125941-13
lands have summoning sickness on the ai's turn the turn after they are played.

[moomarc]

Is anyone else getting random freezes when something is about to resolve. I've only come across it in multiplayer, but aside from that I can't seem to see anything else common between cases. The game will be going smoothly then a player (doesn't matter who) will cast or activate something that goes to the stack, and when you press ok, nothing happens. I don't know if priority isn't being passed properly or something, but the stack is frozen. The game itself doesn't hang though and you can still quit or restart, but it's rather annoying in the middle of a fun match. I've had it while Moonglove Extract was entering play on my side, while an opponent activated Urza avatar's pinging ability (it worked in previous turns) and when an apponent cast a random instant.

[friarsol]

Got this crash.

Attacked with Juggernaut, Talruum Minotaur blocked it. I activated Oracle's Attendant activated to save the Juggernaut, Minotaur was the chosen source. Received the crash, and then my Juggernaut died

Oracle's Attendant crash | Open

java.lang.NullPointerException
at forge.card.abilityfactory.AbilityFactory.getAbility(AbilityFactory.java:130)
at forge.card.replacement.ReplacementHandler.executeReplacement(ReplacementHandler.java:185)
at forge.card.replacement.ReplacementHandler.run(ReplacementHandler.java:144)
at forge.card.replacement.ReplacementHandler.run(ReplacementHandler.java:74)
at forge.Card.replaceDamage(Card.java:8138)
at forge.Card.addCombatDamage(Card.java:7771)

[swordshine]

The spelldescription of Judge Unworthy missed a word "Choose".

[Agetian]

While implementing color restriction for X costs for the AI I noticed the following bug with the AI using Fireball: it'll only spend one mana on Fireball itself and it won't actually spend mana on X to deal damage to me or to any of my creatures. I thought it could have been caused by my changes but even after reverting everything the AI still won't play it correctly for some reason. Here's a test case for dev mode (fireball_ai.test):

Code: Select all

ActivePlayer=Human
ActivePhase=Main1
AICardsInPlay=Swamp; Swamp; Swamp; Swamp; Mountain; Mountain; Plains; Plains
AICardsInHand=Fireball
HumanCardsInPlay=Raging Goblin; Raging Goblin
HumanLife=2


The same happens if there are no creatures in play, so:

Code: Select all

ActivePlayer=Human
ActivePhase=Main1
AICardsInPlay=Swamp; Swamp; Swamp; Swamp; Mountain; Mountain; Plains; Plains
AICardsInHand=Fireball
HumanLife=2


- Agetian

[friarsol]

When I attempt to cast Magma Burst with kicker I have to sacrifice lands before activating mana, which is quite confusing since sacrificing costs always happen after paying mana for normal costed things.