Bug Reports (snapshot builds)

[Max mtg]

the oblivion ring problem is about that proxy ability defined near triggerhandler.java:520 was not updated with new methods I've added

I strongly dislike the way that wrapper is made, I'll find a method to overcome the need to create such a proxy.

[ArsenalNut]

 Detention Sphere doesn't exile all permanents with the same name anymore.

[sollecus11]

Hi there! Congrats for the great work you're doing on Forge! It has evolved immensely since its awesome creation. I've been a lurker in these forums but decided it was time to contribute for the project I've been enjoying for so long.

Having said that, I'd like to file a bug report.

I'm using forge-1.3.1-20121108.122459-11.

The Tectonic Rift card caused a crash report (unfortunately I haven't copied it) when targeting a land. The land destruction effect went on stack but didn't resolve. The 2nd effect "Creatures without flying can't block this turn" works correctly though. Thanks for your attention.

[Sloth]

I strongly dislike the way that wrapper is made, I'll find a method to overcome the need to create such a proxy.

That would be nice. This function always felt wrong.

[Diogenes]

I'm not sure if this affects anything, but in cardsfolder.zip Craterhoof Behemoth's text file is craterfoot_behemoth.txt instead of craterhoof_behemoth.txt.

Also, this might have been mentioned but I haven't noticed it. The cursor doesn't change to indicate when you can resize windows in the deck editor again (in-game it works fine.)

[Max mtg]

I strongly dislike the way that wrapper is made, I'll find a method to overcome the need to create such a proxy.

That would be nice. This function always felt wrong.

Well, the soultion is not as clear as I hoped it would be.
I think we need to declare an interface for an object-on-stack with all methods it needs and make SpellAbility and SA-wrapping-a-trigger implement it. This way a wrapper SA class might have less methods than an ordinary spellability.

Yet, for now I have moved the class formerly inlined in forge.card.trigger.TriggerHandler.runSingleTrigger(Trigger, TriggerType, Map<String, Object>) into a separate file. At least that will stop ruining the readability of TriggerHandler class and this made Oblivion ring working again.

[friarsol]

Very low priority, but wanted to report since I noticed it. The Dev Mode "add to play" button actually adds things to the Stack, so it's impossible to add Land into play using it.

[Max mtg]

Very low priority, but wanted to report since I noticed it. The Dev Mode "add to play" button actually adds things to the Stack, so it's impossible to add Land into play using it.

That's true. I did not know how to do make AI play land. If anyone knows the correct code, add it, please.

[Tenneko]

Okay, I'm new to this bug reporting thing, so if I'm doing this wrong, sorry, but I love this game and all the new stuff you all have done with it so decided to check out the newest snapshots. But something weird keeps happening. All I'm doing is attempting to start a new quest, no saved data, no altered preferences but whenever I try to define which packs I want to be limited to, i get a crash report and it doesn't let me save my choices, even if I only have one pack clicked. It has so far done this with

versions:

forge-1.3.1-20121111.121946-14.tar.bz2
forge-1.3.1-20121110.131433-13.tar.bz2
forge-1.3.1-20121109.120457-12.tar.bz2
forge-1.3.1-20121108.122459-11.tar.bz2
forge-1.3.1-20121107.114557-10.tar.bz2

but version: forge-1.3.1-20121106.124336-9.tar.bz2 and before work fine. If this is because I'm doing something, I'm sorry for wasting your time, but if it is an actual problem, I hope the crash report below helps solve it. It was the same crash report every time for each of the versions I stated above.

[friarsol]

r18026

AI casts Puppeteer Clique on resolution I get this NPE.

NPE | Open

java.lang.NullPointerException
at forge.card.abilityfactory.ai.ChangeZoneAllAi.canPlayAI(ChangeZoneAllAi.java:57)
at forge.card.abilityfactory.SpellAiLogic.canPlayAIWithSubs(SpellAiLogic.java:16)
at forge.card.abilityfactory.CommonAbility.canPlayAI(CommonAbility.java:66)
at forge.card.abilityfactory.ai.DelayedTriggerAi.chkAIDrawback(DelayedTriggerAi.java:19)
at forge.card.abilityfactory.CommonDrawback.chkAIDrawback(CommonDrawback.java:70)
at forge.card.abilityfactory.CommonDrawback.chkAIDrawback(CommonDrawback.java:74)
at forge.card.abilityfactory.SpellAiLogic.doTriggerNoCostWithSubs(SpellAiLogic.java:35)
at forge.card.abilityfactory.SpellAiLogic.doTriggerAI(SpellAiLogic.java:26)
at forge.card.abilityfactory.CommonAbility.doTrigger(CommonAbility.java:71)
at forge.card.trigger.WrappedAbility.doTrigger(WrappedAbility.java:151)

[Sloth]

r18026

AI casts Puppeteer Clique on resolution I get this NPE.

NPE | Open

java.lang.NullPointerException
at forge.card.abilityfactory.ai.ChangeZoneAllAi.canPlayAI(ChangeZoneAllAi.java:57)
at forge.card.abilityfactory.SpellAiLogic.canPlayAIWithSubs(SpellAiLogic.java:16)
at forge.card.abilityfactory.CommonAbility.canPlayAI(CommonAbility.java:66)
at forge.card.abilityfactory.ai.DelayedTriggerAi.chkAIDrawback(DelayedTriggerAi.java:19)
at forge.card.abilityfactory.CommonDrawback.chkAIDrawback(CommonDrawback.java:70)
at forge.card.abilityfactory.CommonDrawback.chkAIDrawback(CommonDrawback.java:74)
at forge.card.abilityfactory.SpellAiLogic.doTriggerNoCostWithSubs(SpellAiLogic.java:35)
at forge.card.abilityfactory.SpellAiLogic.doTriggerAI(SpellAiLogic.java:26)
at forge.card.abilityfactory.CommonAbility.doTrigger(CommonAbility.java:71)
at forge.card.trigger.WrappedAbility.doTrigger(WrappedAbility.java:151)

Fixed! Thanks sol.

[RumbleBBU]

Serious bug in GameFormatQuest.java spanning several snapshots...fixed.

The bug was introduced in r17875, evidently with the best of intentions.

Believe me, when I originally wrote the class, my first instinct was to inherit GameFormat.java. It seemed only logical, and would have saved a lot space.

And there's a damn good reason I didn't. What it boils down is this part of GameFormat.java:

Code: Select all

        this.allowedSetCodes = Collections.unmodifiableList(sets);
        this.bannedCardNames = Collections.unmodifiableList(bannedCards);
    protected final transient Predicate<CardPrinted> filterRules;
    protected final transient Predicate<CardPrinted> filterPrinted;


This is simply incompatible with what GameFormatQuest is used for. The whole point is that you can change the lists! To give you two examples, with the r17875 code, you couldn no longer:
- Define your own custom formats when you start a quest, nor
- Unlock any new sets in the course of your quest
If you tried to, you would get a runtime error! (unsupported method)

The only real way to make GameFormatQuest extend GameFormat would be to make the GameFormat privates non-final/non-immutable (and then make sure that the methods of base GameFormat class only give immutable access, while the inheriting GameFormatQuest could have methods that give non-immutable access). (I preferred not to touch the GameFormat class, so I chose to replicate a lot of its code.)

I thought this should have been pretty obvious if you considered for a moment what the GameFormatQuest class was actually used for.

[Max mtg]

Predicates now hold references to lists you change, so they get auto-updated when you change the allowed lists.

[RumbleBBU]

Yes, but the fundamental problem in this case was that you could not change the unmodifiable lists at all.

[Max mtg]

Ok, will changes in 18051 do the job?