Bug Reports (snapshot builds)

[Hanmac]

found the culprit:
it wasn't the Farbricate but much deeper in the code.

the LKI copy i create does not have a new map for the counters.
so when i add a counter to the copy, it does add it to the original too.

Did found it because Eclipse debugger does show the Ids of the objects.

should be fixed now.

[tjtillman]

r32230

I control Padeem and Ballista Charger, CPU also controls a Ballista Charger, with each Ballista Charger being the highest CMC controlled artifact for each player.

First time it comes back to me, Padeem's trigger works properly, but the subsequent turn (with artifact controls being the same), Padeem's trigger doesn't hit at upkeep.

Not sure whats going on there.

Sequence was:
Turn 8 I cast Padeem (my only non-land permanent)
Turn 9 CPU attacks with 2 non-artifact creatures (its only non-land permanents), I block with Padeem (non lethal), CPU casts another non artifact creature
Turn 10 I play Ballista Charger
Turn 11 CPU attacks with its 3 non-artifact creatures, I block with Padeem (non lethal), CPU casts its own Ballista Charger
Turn 12 Padeem's trigger correctly goes on the stack, I draw a card. I play a non-artifact creature
Turn 13 CPU plays a non-artifact creature
Turn 14 Padeem's ability doesn't trigger even though it should

EDIT: May not be as complicated as all this. I had another game where I had Padeem out and a Prophetic Prism controlled by me was the ONLY artifact on the battlefield, and his ability didn't trigger on upkeep.

[akuolum]

There's a problem with multiplayer mode. I'm playing on LAN. The person who is supporting server see double faced cards right while the other player, see the wrong face (second phase) with the wrong power and toughness. Once card is flipped second player see wrong face too. The counters doesn't show right for second player. I attach some screenshoots for let you understand. If I start a new match all single faced cards are flipped.

I hope you understand me. My english is not good at all.

[onderzeeboot]

EDIT: May not be as complicated as all this. I had another game where I had Padeem out and a Prophetic Prism controlled by me was the ONLY artifact on the battlefield, and his ability didn't trigger on upkeep.

Padeem seems to only trigger if you control the permanent with the highest converted mana cost _and_ that permanent is also an artifact.

[onderzeeboot]

1. (r32230) When my opponent plays Gonti, Lord of Luxury, Forge shows me which card my opponent picked. The card always stays in a hidden zone, so it shouldn't show me that.

2. (r32223) My opponent will attack Riparian Tiger into a 5/6, even when he hasn't enough energy to pay for the triggered ability, thus slamming a 4/4 into certain death.

Edit3:
I think I fixed Padeem (for real this time)! This triggered reliably for me when it should. Drew me cards when I had highest CMC on the battlefield, then stopped drawing cards when I no longer had highest CMC.

Still broken Padeem | Open

Code: Select all

Name:Padeem, Consul of Innovation
ManaCost:3 U
Types:Legendary Creature Vedalken Artificer
PT:1/4
S:Mode$ Continuous | Affected$ Artifact.YouCtrl | AddKeyword$ Hexproof | Description$ Artifacts you control have hexproof.
T:Mode$ Phase | Phase$ Upkeep | ValidPlayer$ You | TriggerZones$ Battlefield | CheckSVar$ X | SVarCompare$ GEY | References$ X,Y | Execute$ TrigDraw | TriggerDescription$ At the beginning of your upkeep, if you control the artifact with the highest converted mana cost or tied for the highest converted mana cost, draw a card.
SVar:TrigDraw:AB$ Draw | Cost$ 0 | Defined$ You | NumCards$ 1
SVar:X:Count$HighestCMC_Artifact.YouCtrl+inZoneBattlefield
SVar:Y:Count$HighestCMC_Artifact.inZoneBattlefield
SVar:Picture:http://www.wizards.com/global/images/magic/general/padeem_consul_of_innovation.jpg
Oracle:Artifacts you control have hexproof.\nAt the beginning of your upkeep, if you control the artifact with the highest converted mana cost or tied for the highest converted mana cost, draw a card.


Edit4: I don't actually have/know how to access the code base, so if someone else could review and commit the change to Padeem, that would be great.

[fmartel]

Description: [in commander waiting after AI to play Patchwork Gnomes. Java killed Forge]

OutOfMemoryError | Open

Code: Select all

Forge Version:    1.5.56-SNAPSHOT-r32193
Operating System: Windows 7 6.1 amd64
Java Version:     1.8.0_101 Oracle Corporation

java.lang.OutOfMemoryError: GC overhead limit exceeded


[Marek14]

A minor thing for Diligent Farmhand and Pardic Firecat:

Their ability to be counted for Muscle Burst/Flame Burst is implemented as a part of the later cards. But in case there's Yixlid Jailer active, the cards in graveyard lose their abilities and Diligent Farmhand and Pardic Firecat should not be counted for Bursts in this case.

[Hanmac]

@Marek14: fixed it,
but you don't want you to find such things,
because they are ugly to code ;P

used StaticAbility and Hidden Keywords so they can be negated by Yixlid

[Marek14]

I control Vivid Creek, Island and Reflecting Pool. Reflecting Pool can give colorless mana, even though I don't have anything that produces it.

[Hanmac]

I control Vivid Creek, Island and Reflecting Pool. Reflecting Pool can give colorless mana, even though I don't have anything that produces it.

should be fixed in r32237.
there was a leftover when was turned into {C}.

i checked the function usage and it should not break any other stuff.

[Hanmac]

2. (r32223) My opponent will attack Riparian Tiger into a 5/6, even when he hasn't enough energy to pay for the triggered ability, thus slamming a 4/4 into certain death.

fixed in r32239:
i added a check if cost can be paid for the Trigger in predictPowerBonusOfAttacker and predictToughnessBonusOfAttacker.

that should fix the thing with Riparian Tiger.

it probably does not fix it for Thriving Turtle, but there i need to see where it does go wrong.

@all: also Please checkout if that might break something.

[onderzeeboot]

Dangnabbit, my Padeem still doesn't work. If neither player has an artifact, you still draw an additional card, because the highest converted mana cost of your nonexisting artifacts is equal to the highest cmc of your opponent's nonexisting artifacts. Grmbl...

[friarsol]

Dangnabbit, my Padeem still doesn't work. If neither player has an artifact, you still draw an additional card, because the highest converted mana cost of your nonexisting artifacts is equal to the highest cmc of your opponent's nonexisting artifacts. Grmbl...

I had already fixed Padeem... in r32232 which was before you posted your fix, did you just miss it?

[onderzeeboot]

Dangnabbit, my Padeem still doesn't work. If neither player has an artifact, you still draw an additional card, because the highest converted mana cost of your nonexisting artifacts is equal to the highest cmc of your opponent's nonexisting artifacts. Grmbl...

I had already fixed Padeem... in r32232 which was before you posted your fix, did you just miss it?

Yeah, I was still working on r32230. I installed r32242 last night and couldn't reproduce the bug anymore. That's when I decided I better go to bed Glad to see it was fixed though, now I can stop agonizing over it

[Agetian]

Ok, after some more testing I concluded that actually all cards with either Enchanted Cost or Exiled Cost expression for mana cost payment appear to be broken as of the current (r32246) version. For example:

- Merseine does not work, always asking to pay 1 generic mana instead of the enchanted creature's mana cost (expression Cost$ Mana<1\EnchantedCost> is broken)
- Back from the Brink does not work, always asking to pay 20 generic mana instead of the exiled card's mana cost (expression Cost$ Mana<20\Exiled> is broken)

This used to work relatively recently.
The culprit appears to be the cost-copying process at CostAdjustment.java:48 which does not copy costs like that correctly. What would be a good way to write it though such that it works correctly and does not break anything else? :/

EDIT: in r32247-32249, I tried fixing this by explicitly copying the isEnchantedManaCost and isExiledManaCost parameters, but it's a rather hacky solution and I'm not sure if there is a bigger and deeper issue underneath and/or a potential for breaking something this way, so please take a look and update if possible/necessary.

- Agetian