Collectible Card Game Headquarters

Braids wrote:

Almost_Clever wrote:Current Beta: . . . I cast Fling to end game one of a match and tried to do the same in game two, but I cannot cast Fling a second time.

what creatures were you trying to sacrifice each time? how did the game respond when you tried to cast Fling the second time?

I cannot select a target, pay mana or select a creature to sacrifice; the game will not let me cast the spell. I later figured out that I have three copies in my deck (M12 draft) and I cannot cast Fling [4] (in any match) but can cast the other two...

Braids wrote:

wakka9ca wrote:. . . Worse, if I try to Traumatize the opponent, its entire graveyard will be put back into its library, having the opposite effect. This happens only when the mill effect hit cards like Emrakul, the Aeons Torn, etc.

i think this is what is supposed to happen. the entire effect of Traumatize resolves, so half the library goes into the graveyard. then Emrakul, the Aeons Torn 's ability triggers, causing the entire graveyard to be shuffled back in. if i'm wrong, can you point me to an official ruling or two?

edit: does this happen with Purity or Guile?

Hmm... that's right. If that happened, it may explain why the graveyard is shuffled back to the library. However, it doesn't explain why Glimpse the Unthinkable sometimes doesn't work - and the graveyard doesn't get shuffled back into the library either. That's the more puzzling point...

Other than this, do you know if there is a known memory leak issue with the current version of Forge? The memory usage of java grows well beyond 1GB as the program runs (and keeps growing). I know Java has memory management but if you are using any external library written in other languages (like C), memory leak can still happen very subtlety. Is this normal? Also, I was wondering if a more efficient algorithm doesn't exist to load the card shop or the deck editor in quest mode when your collection has grown a bit larger. The load time can take up to 10 minutes.

My PC is not the fastest computer out there but I can run MATLAB simulations on it no problem.

i've mentioned problems with Ulamog, the Infinite Gyre before, and here's another: i once mulliganned down to 4. i had an ulamog in my hand of 4, i mulliganed again and got 4 card with 2 ulamogs. after this, whenever i mulliganned i got back 2 ulamogs and 1 more card each time

wakka9ca wrote:. . . However, it doesn't explain why Glimpse the Unthinkable sometimes doesn't work - and the graveyard doesn't get shuffled back into the library either. That's the more puzzling point...

i'm not really qualified to help, sorry. if you have a reproducible {even intermittently} description, i would suggest you file a formal bug report with Mantis.

wakka9ca wrote:Other than this, do you know if there is a known memory leak issue with the current version of Forge? The memory usage of java grows well beyond 1GB as the program runs (and keeps growing). I know Java has memory management but if you are using any external library written in other languages (like C), memory leak can still happen very subtlety. Is this normal? Also, I was wondering if a more efficient algorithm doesn't exist to load the card shop or the deck editor in quest mode when your collection has grown a bit larger. The load time can take up to 10 minutes.

i wouldn't say it's normal, but we do know about it. i put a serious dent in Forge's memory usage in version 1.1.0, released on 15 July this year. we still have more memory leaks, but they are becoming increasingly difficult to fix. i am unaware of any linkages with lower level libraries. we use a lot of temporary card lists. the only explanation i can come up with is that we have circular references.

10 minutes? wow. there are plans to consolidate the deck editors, and we will be using memory efficient techniques called generators.

unofficially, i would recommend closing the program completely after two matches.

wakka9ca wrote:My PC is not the fastest computer out there but I can run MATLAB simulations on it no problem.

i usually have a machine with around 2GB of RAM and less than a 2 GHz processor, and it doesn't seem too bad. the deck editor takes a while, but usually much less than 10 minutes.

sorry i'm not much direct help.

Thanks for the swift reply though. Tracing memory leaks can be a pain in the ass. I did custom memory management before when I was embedding Python into C and it was definitely not trivial. Some referencing issues are hard to find. I guess people have to have free time to fish them out.

IBM provide some basic tips on that.
http://www.ibm.com/developerworks/library/j-leaks/
However, it requires a commercial tool JProbe.
But if you want to get your hands a bit more dirty:
http://blog.emptyway.com/2007/04/02/fin ... java-apps/

If you can find a great tool to help tracing the program, that will be great. Shame I cannot commit myself to this wonderful project though. Only got a few days off and will get busy soon again :/

After all, memory leak issues is not that bad (closing/reopening works). The annoying part is the wait time for deck editor and shop (quest mode) so hopefully it will become a priority to fix in the future.

wakka9ca wrote:Thanks for the swift reply though. Tracing memory leaks can be a pain in the ass. I did custom memory management before when I was embedding Python into C and it was definitely not trivial. Some referencing issues are hard to find. I guess people have to have free time to fish them out.

IBM provide some basic tips on that.
http://www.ibm.com/developerworks/library/j-leaks/
However, it requires a commercial tool JProbe.
But if you want to get your hands a bit more dirty:
http://blog.emptyway.com/2007/04/02/fin ... java-apps/

If you can find a great tool to help tracing the program, that will be great. Shame I cannot commit myself to this wonderful project though. Only got a few days off and will get busy soon again :/

After all, memory leak issues is not that bad (closing/reopening works). The annoying part is the wait time for deck editor and shop (quest mode) so hopefully it will become a priority to fix in the future.

someone is tinkering with the deck editors, but we don't have an official assignee yet. so i won't name names.

your quest shop experience is unusual; i have never had slowness there. how many cards are in your collection in your quest?

i was using Eclipse Memory Analyzer Tool to explore hprof output. it quickly became obvious that the main thing was our card database -- it has OVER NEARLY NINE THOUSAND items with bells and whistles {code abilities, not just text}. we had many method calls that would make copies of the entire database. {well, it wouldn't copy the cards, but we would recreate the list of thousands of items pointing to those cards.} i converted most of them to use generators instead of copying lists. it substantially {but not completely} helped at least one person. timmermac, if you're reading this, thanks for hanging in there.

i had an instant, similar to one that i had before with ulamog, where i could click Ichorclaw Myr, and i could drag it around, but i couldn't cast it. 2 other copies do work, but one doesn't

@Braids: only 9477 cards in total. That's not even all the cards even if they are all one-of's (which they aren't)

What is weird is that deck editor and quest shop loading time are about the same: very slow as the collection increases. I have started many new games and everything works fine at the beginning. As soon as the collection start growing, the load time will take longer and longer.

I believe it is not necessary to load all the info about a card in shop or deck editor. A text render of its card text should be enough so the object instance with all its functional code (only used in game match) should not be loaded. Using a wrapper class may help (or not). The load time should not be much longer than an Excel spreadsheet (or Magic Album, etc...) which is 5-10 seconds.

wakka9ca wrote:@Braids: only 9477 cards in total. That's not even all the cards even if they are all one-of's (which they aren't)

What is weird is that deck editor and quest shop loading time are about the same: very slow as the collection increases. I have started many new games and everything works fine at the beginning. As soon as the collection start growing, the load time will take longer and longer.

I believe it is not necessary to load all the info about a card in shop or deck editor. A text render of its card text should be enough so the object instance with all its functional code (only used in game match) should not be loaded. Using a wrapper class may help (or not). The load time should not be much longer than an Excel spreadsheet (or Magic Album, etc...) which is 5-10 seconds.

i couldn't agree more {except about the wrapper class}. unfortunately, this is an issue of technical debt.

i'm not sure a wrapper class is what we need; i think it would be better to fully distinguish a card outside a game from a card inside a game.

i used Sower of Temptation on my opponent's Stuffy Doll, but its ability still dealt damage to me

s1886x wrote:i used Sower of Temptation on my opponent's Stuffy Doll, but its ability still dealt damage to me

The first time i stole a Stuffy Doll i was smiling, lets use this one against the opponent . Then i was the one who received the damage and was like .

But it is correct, the text on Stuffy Doll says it deals damage to the chosen player. The player to receive the damage is only chosen once, when Stuffy Doll enters play. No matter who takes control of it, the chosen player remains the same.

Braids wrote:

wakka9ca wrote:Thanks for the swift reply though. Tracing memory leaks can be a pain in the ass. I did custom memory management before when I was embedding Python into C and it was definitely not trivial. Some referencing issues are hard to find. I guess people have to have free time to fish them out.

IBM provide some basic tips on that.
http://www.ibm.com/developerworks/library/j-leaks/
However, it requires a commercial tool JProbe.
But if you want to get your hands a bit more dirty:
http://blog.emptyway.com/2007/04/02/fin ... java-apps/

If you can find a great tool to help tracing the program, that will be great. Shame I cannot commit myself to this wonderful project though. Only got a few days off and will get busy soon again :/

After all, memory leak issues is not that bad (closing/reopening works). The annoying part is the wait time for deck editor and shop (quest mode) so hopefully it will become a priority to fix in the future.

someone is tinkering with the deck editors, but we don't have an official assignee yet. so i won't name names.

your quest shop experience is unusual; i have never had slowness there. how many cards are in your collection in your quest?

i was using Eclipse Memory Analyzer Tool to explore hprof output. it quickly became obvious that the main thing was our card database -- it has OVER NEARLY NINE THOUSAND items with bells and whistles {code abilities, not just text}. we had many method calls that would make copies of the entire database. {well, it wouldn't copy the cards, but we would recreate the list of thousands of items pointing to those cards.} i converted most of them to use generators instead of copying lists. it substantially {but not completely} helped at least one person. timmermac, if you're reading this, thanks for hanging in there.

No problem... I'm finding that I can play at least 5 or 6 matches in a row now without everything bogging down, so progress has definitely been made. Wakka9ca, could you post your computer's specs? Mine's a Pentium-4 3.06 Ghz with 1 GB RAM, Intel 915 graphics and Windows XP SP 3.

I cast Bitter Ordeal after having 40 or so token creatures go to my graveyard and got the following error:

Code: Select all

This is a Crash Report. An error has occurred. Please save this message to a file.
Please follow the instructions at this address to submit this Crash Report, plus what you were doing at the time:
   http://tinyurl.com/3zzrnyb
Reporting bugs in Forge is very important. We thank you for your time.

For input string: "GravestormCount1"


Version:
Forge version 1.1.2, build ID 9793

OS: Windows 7 Version: 6.1 Architecture: amd64

Java Version: 1.6.0_24 Vendor: Sun Microsystems Inc.

Detailed error trace:
java.lang.NumberFormatException: For input string: "GravestormCount1"
   at java.lang.NumberFormatException.forInputString(Unknown Source)
   at java.lang.Integer.parseInt(Unknown Source)
   at java.lang.Integer.parseInt(Unknown Source)
   at forge.card.abilityFactory.AbilityFactory.calculateAmount(AbilityFactory.java:1150)
   at forge.card.abilityFactory.AbilityFactory_Copy.copySpellResolve(AbilityFactory_Copy.java:592)
   at forge.card.abilityFactory.AbilityFactory_Copy.access$600(AbilityFactory_Copy.java:17)
   at forge.card.abilityFactory.AbilityFactory_Copy$6.resolve(AbilityFactory_Copy.java:422)
   at forge.card.abilityFactory.AbilityFactory.resolve(AbilityFactory.java:1764)
   at forge.card.spellability.SpellAbility_Requirements.finishPaying(SpellAbility_Requirements.java:136)
   at forge.card.spellability.Cost_Payment.payCost(Cost_Payment.java:565)
   at forge.card.spellability.SpellAbility_Requirements.startPaying(SpellAbility_Requirements.java:127)
   at forge.card.spellability.SpellAbility_Requirements.needPayment(SpellAbility_Requirements.java:117)
   at forge.card.spellability.SpellAbility_Requirements.fillRequirements(SpellAbility_Requirements.java:90)
   at forge.GameAction.playSpellAbility_NoStack(GameAction.java:2103)
   at forge.card.trigger.TriggerHandler$2.resolve(TriggerHandler.java:901)
   at forge.card.abilityFactory.AbilityFactory.resolve(AbilityFactory.java:1764)
   at forge.MagicStack.resolveStack(MagicStack.java:782)
   at forge.Phase.passPriority(Phase.java:697)
   at forge.ComputerAI_General.stackResponse(ComputerAI_General.java:442)
   at forge.ComputerAI_General.stack_not_empty(ComputerAI_General.java:387)
   at forge.gui.input.InputControl.updateInput(InputControl.java:183)
   at forge.GuiInput.update(GuiInput.java:30)
   at java.util.Observable.notifyObservers(Unknown Source)
   at java.util.Observable.notifyObservers(Unknown Source)
   at forge.MyObservable.updateObservers(MyObservable.java:17)
   at forge.gui.input.InputControl.resetInput(InputControl.java:89)
   at forge.Phase.passPriority(Phase.java:688)
   at forge.gui.input.Input_PassPriority.selectButtonOK(Input_PassPriority.java:46)
   at forge.GuiInput.selectButtonOK(GuiInput.java:57)
   at forge.GuiDisplay4.okButtonActionPerformed(GuiDisplay4.java:1117)
   at forge.GuiDisplay4.access$300(GuiDisplay4.java:52)
   at forge.GuiDisplay4$32.actionPerformed(GuiDisplay4.java:870)
   at javax.swing.AbstractButton.fireActionPerformed(Unknown Source)
   at javax.swing.AbstractButton$Handler.actionPerformed(Unknown Source)
   at javax.swing.DefaultButtonModel.fireActionPerformed(Unknown Source)
   at javax.swing.DefaultButtonModel.setPressed(Unknown Source)
   at javax.swing.AbstractButton.doClick(Unknown Source)
   at javax.swing.plaf.basic.BasicRootPaneUI$Actions.actionPerformed(Unknown Source)
   at javax.swing.SwingUtilities.notifyAction(Unknown Source)
   at javax.swing.JComponent.processKeyBinding(Unknown Source)
   at javax.swing.KeyboardManager.fireBinding(Unknown Source)
   at javax.swing.KeyboardManager.fireKeyboardAction(Unknown Source)
   at javax.swing.JComponent.processKeyBindingsForAllComponents(Unknown Source)
   at javax.swing.JComponent.processKeyBindings(Unknown Source)
   at javax.swing.JComponent.processKeyEvent(Unknown Source)
   at java.awt.Component.processEvent(Unknown Source)
   at java.awt.Container.processEvent(Unknown Source)
   at java.awt.Component.dispatchEventImpl(Unknown Source)
   at java.awt.Container.dispatchEventImpl(Unknown Source)
   at java.awt.Component.dispatchEvent(Unknown Source)
   at java.awt.KeyboardFocusManager.redispatchEvent(Unknown Source)
   at java.awt.DefaultKeyboardFocusManager.dispatchKeyEvent(Unknown Source)
   at java.awt.DefaultKeyboardFocusManager.preDispatchKeyEvent(Unknown Source)
   at java.awt.DefaultKeyboardFocusManager.typeAheadAssertions(Unknown Source)
   at java.awt.DefaultKeyboardFocusManager.dispatchEvent(Unknown Source)
   at java.awt.Component.dispatchEventImpl(Unknown Source)
   at java.awt.Container.dispatchEventImpl(Unknown Source)
   at java.awt.Window.dispatchEventImpl(Unknown Source)
   at java.awt.Component.dispatchEvent(Unknown Source)
   at java.awt.EventQueue.dispatchEventImpl(Unknown Source)
   at java.awt.EventQueue.access$000(Unknown Source)
   at java.awt.EventQueue$1.run(Unknown Source)
   at java.awt.EventQueue$1.run(Unknown Source)
   at java.security.AccessController.doPrivileged(Native Method)
   at java.security.AccessControlContext$1.doIntersectionPrivilege(Unknown Source)
   at java.security.AccessControlContext$1.doIntersectionPrivilege(Unknown Source)
   at java.awt.EventQueue$2.run(Unknown Source)
   at java.awt.EventQueue$2.run(Unknown Source)
   at java.security.AccessController.doPrivileged(Native Method)
   at java.security.AccessControlContext$1.doIntersectionPrivilege(Unknown Source)
   at java.awt.EventQueue.dispatchEvent(Unknown Source)
   at java.awt.EventDispatchThread.pumpOneEventForFilters(Unknown Source)
   at java.awt.EventDispatchThread.pumpEventsForFilter(Unknown Source)
   at java.awt.EventDispatchThread.pumpEventsForHierarchy(Unknown Source)
   at java.awt.EventDispatchThread.pumpEvents(Unknown Source)
   at java.awt.EventDispatchThread.pumpEvents(Unknown Source)
   at java.awt.EventDispatchThread.run(Unknown Source)


I used Liliana Vess('s) ultimate ability with 2 copies of AEther Flash in play. Neither of them triggered.

@timmermac: Intel Pentium Dual CPU E2200 2.2GHz, 3GB RAM, GeForce 9600

I can run Supreme Commander with no issue here so I don't think my PC is that bad.

The thing is, the game doesn't slow down after 5 or 6 matches. Only the memory grow. Also, the quest shop and deck editor are slowed down based on the number of cards in the collection, but on how long the game has been running.