Lazy CardFactory loading is the problem . . . or is it?

[Braids]

i've been trying to reproduce and track down the alleged bugs with the lazy CardFactory loading:

• New quest mode hangs
• Only colorless cards placed into 3-color generated decks
• Unacceptable behavior in deck editor (vague)

i wrote some unit tests. i slapped a stderr progress meter into ReadCard so i would know when it was done loading cards.

the unit tests passed with flying colors.

i added a call to getCards within CardFactory's constructor, so that it behaves like it did before r10095. it loads all cards at time of construction. the bugs are still there mostly.

• New quest mode still hangs badly
• I noticed a two-colored card in my hand in my 3-color generated deck, but most of the cards were still colorless.
• The deck editor takes a long time to load, even after all the cards have loaded into memory. It prefers the window to be resized before it updates its contents (i.e., after clicking on a card or scrolling).

i promised myself no more late night commits, so the changes i made are still private.



btw, Goblin Hero has given us strikethrough at my request. use the s button, between Font colour and spoiler= in the BBCode editor. it makes for more dramatic redactions and has immense humor potential. Goblin Hero is awesome.

[Chris H.]

Using rev 10131 starting a new quest also appears to hang my computer. It pegged the processor, I waited about 1 minute and I gave up and did a force quit.

I can apparently continue an existing quest. I can scroll through the quest mode deck editor and the card descriptions in the card detail panel look OK.

When I scroll through the constructed mode deck editor the card descriptions in the card detail panel look OK for some cards. Other cards have either no description or are unparsed.

The filters for the constructed mode deck editor have some problems. This may be related to the updating the card list in the table.

Playing a match by selecting Constructed -> Generate Deck - > 2-Color Deck (new) results in an null error exception:

An error has occurred | Open

An error has occurred. You can copy/paste this message or save it to a file.
Please report this, plus what you tried to do, to:
viewforum.php?f=26
If you don't want to register an account, you can mail it directly to
mtgerror@yahoo.com


n must be positive


Version:
Forge -- official beta: $Date: 2011-01-06 11:34:48 -0500 (Thu, 06 Jan 2011) $, SVN revision: $Revision: 4891 $

OS: Mac OS X Version: 10.6.8 Architecture: x86_64

Java Version: 1.6.0_24 Vendor: Apple Inc.

Detailed error trace:
java.lang.IllegalArgumentException: n must be positive
at java.util.Random.nextInt(Random.java:250)
at forge.deck.generate.Generate2ColorDeck.get2ColorDeck(Generate2ColorDeck.java:180)
at forge.Gui_NewGame.generate2ColorDeck(Gui_NewGame.java:953)
at forge.Gui_NewGame.genDecks(Gui_NewGame.java:833)
at forge.Gui_NewGame.startButton_actionPerformed(Gui_NewGame.java:743)
at forge.Gui_NewGame$13.actionPerformed(Gui_NewGame.java:552)
at javax.swing.AbstractButton.fireActionPerformed(AbstractButton.java:2028)
at javax.swing.AbstractButton$Handler.actionPerformed(AbstractButton.java:2351)
at javax.swing.DefaultButtonModel.fireActionPerformed(DefaultButtonModel.java:387)
at javax.swing.DefaultButtonModel.setPressed(DefaultButtonModel.java:242)
at javax.swing.plaf.basic.BasicButtonListener.mouseReleased(BasicButtonListener.java:236)
at java.awt.Component.processMouseEvent(Component.java:6374)
at javax.swing.JComponent.processMouseEvent(JComponent.java:3267)
at java.awt.Component.processEvent(Component.java:6139)
at java.awt.Container.processEvent(Container.java:2085)
at java.awt.Component.dispatchEventImpl(Component.java:4736)
at java.awt.Container.dispatchEventImpl(Container.java:2143)
at java.awt.Component.dispatchEvent(Component.java:4566)
at java.awt.LightweightDispatcher.retargetMouseEvent(Container.java:4621)
at java.awt.LightweightDispatcher.processMouseEvent(Container.java:4282)
at java.awt.LightweightDispatcher.dispatchEvent(Container.java:4212)
at java.awt.Container.dispatchEventImpl(Container.java:2129)
at java.awt.Window.dispatchEventImpl(Window.java:2478)
at java.awt.Component.dispatchEvent(Component.java:4566)
at java.awt.EventQueue.dispatchEventImpl(EventQueue.java:680)
at java.awt.EventQueue.access$000(EventQueue.java:86)
at java.awt.EventQueue$1.run(EventQueue.java:639)
at java.awt.EventQueue$1.run(EventQueue.java:637)
at java.security.AccessController.doPrivileged(Native Method)
at java.security.AccessControlContext$1.doIntersectionPrivilege(AccessControlContext.java:87)
at java.security.AccessControlContext$1.doIntersectionPrivilege(AccessControlContext.java:98)
at java.awt.EventQueue$2.run(EventQueue.java:653)
at java.awt.EventQueue$2.run(EventQueue.java:651)
at java.security.AccessController.doPrivileged(Native Method)
at java.security.AccessControlContext$1.doIntersectionPrivilege(AccessControlContext.java:87)
at java.awt.EventQueue.dispatchEvent(EventQueue.java:650)
at java.awt.EventDispatchThread.pumpOneEventForFilters(EventDispatchThread.java:296)
at java.awt.EventDispatchThread.pumpEventsForFilter(EventDispatchThread.java:211)
at java.awt.EventDispatchThread.pumpEventsForHierarchy(EventDispatchThread.java:201)
at java.awt.EventDispatchThread.pumpEvents(EventDispatchThread.java:196)
at java.awt.EventDispatchThread.pumpEvents(EventDispatchThread.java:188)
at java.awt.EventDispatchThread.run(EventDispatchThread.java:122)

[slapshot5]

Playing a match by selecting Constructed -> Generate Deck - > 2-Color Deck (new) results in an null error exception:

I did not get this error when I just tried it in r10003 (the last beta revision). Just FYI.

-slapshot5

[Braids]

i have backed out the changes to the LazyCardFactory and ReadCard.

they are still in the svn tree, but they have been reverse merged out as of r10141.

Edit 1: the rollback seems to have had the following positive effects:

• Generating a 3-color deck now adds mostly colored cards.
• Starting a new quest no longer freezes.
• The deck editor seems faster and more responsive.

i am frustrated. the changes to CardFactory and ReadCard were supposed to be transparent. but they introduced subtle systemic bugs that i do not comprehend.

Edit 2: please keep in mind that i even forced the CardFactory to load all the cards in its constructor, so the behavior was allegedly the same as before the changes. that did not even help.

the quotes below are from another thread at http://www.slightlymagic.net/forum/viewtopic.php?f=26&t=4884&p=63547#p63547

Braids, don't give up on it....

don't give up what exactly? the lazy CardFactory? i don't have the patience right now to diagnose the strange systemic behaviors above. i won't give up on speeding up my unit tests though.

After startup the only major delay is loading the main deck editor . . .

you are right. the program seems to run more smoothly with the old CardFactory.

There are ways to reduce the load time without necessarily having a lazy CardFactory. If you use the cardsfolder.zip option, then it should be a bit faster . . .

my main reason for doing this was to eliminate the long startup time for unit tests. thanks for the tip. i will see if that helps.