Bug reports

[ShawnieBoy]

creture exiled by Banisher Priest did not return to the battlefield when the priest died

Not able to replicate this - was there anything else going on on the battlefield?

[jerichopumpkin]

Banisher Priest died because of combat damage. I don't remember the exact situation, only that the AI was attacking me, with Banisher Priest and a couple other creatures. I killed it with a creature with first strike. I'm sorry, but I'm unable to give more details, since I've lost the game log. As soon as I can I'll try to replicate it

[PalladiaMors]

I was trying to use Galepowder Mage's groovy to do other cards, when I noticed an issue: it can target itself, contrary to its oracle text.

[ShawnieBoy]

I was trying to use Galepowder Mage's groovy to do other cards, when I noticed an issue: it can target itself, contrary to its oracle text.

Nice find. Submitted a fix.

[Agetian]

I have encountered a rather weird issue with Magarena 1.53 (fresh install into a new folder). For some reason, no matter which AI I choose (I tried all three), the game will "hang" indefinitely in the End Turn step (of the very first human's turn), with the little "thinking" icon rotating in the box saying "Play a card or ability". I tried waiting for about 10 minutes and the game still stayed in this "waiting" state as if the AI was thinking. I tried playing with various decks but also to no avail.

I don't have this issue when running Magarena 1.52 or earlier (1.52 installed to a different folder still works just great). In case it matters, I'm running Magarena on Java 7 (OpenJDK 1.7.0_51) on Linux Mint 15 (64-bit).

Apparently the new version (1.53) works fine on Windows but not on Linux (works ok on my Windows laptop but not on my two desktop PCs running Linux Mint).

Can you please tell me how to get Magarena 1.53 to function properly on Linux?

- Agetian

[Lodici]

Windows 7 is my OS of preference (for now) but I have Linux Mint 16 running in a VirtualBox so I will have a look. When you say "hang", does the UI become completely unresponsive? Did you import settings from a previous version when you first ran 1.53?

[Agetian]

Windows 7 is my OS of preference (for now) but I have Linux Mint 16 running in a VirtualBox so I will have a look. When you say "hang", does the UI become completely unresponsive? Did you import settings from a previous version when you first ran 1.53?

On the positive side, it seems like this might be some kind of a OpenJDK version issue. I'm trying various newer VirtualBoxed Linux systems as well and it seems like v1.53 acts much better in a newer environment. I'll try upgrading my OpenJDK on the main system (and possibly the kernel as well) to a newer version and see if that helps it, I think it might as well.

I try not to use Windows unless I absolutely must (the laptop has it 'cause my family still uses it).
As for the issue itself, the UI does not become completely unresponsive, it's possible to look at cards (they get magnified and stuff), it's possible to exit to the menu, concede the game, restart the duel, etc., it's just that the AI seems to go into "eternal thinking" mode (for some reason in the least expected place - that is, human's End Turn, with not much going on).
I did import the pictures from my previous installation with the "import" feature but an absolutely clean installation acts the same way as well (no pictures, no import procedure at all).



- Agetian

[Lodici]

Ok, thanks. Here is the JDK I have running...

Code: Select all

java version "1.7.0_55"
OpenJDK Runtime Environment (IcedTea 2.4.7) (7u55-2.4.7-1ubuntu1~0.13.10.1)
OpenJDK 64-Bit Server VM (build 24.51-b03, mixed mode)


Seems to work ok, although there are a few dialogs where the text falls off the edge due to the font - not sure if that is a linux issue or VBox though?

[Lodici]

When I run magarena in Linux Mint 16, OpenJDK version "1.7.0_55", VirtualBox the decks list is not sorted correctly. Anyone else experience this? EDIT: Also happens on Zorin, so looks like a linux-specific problem - I have raised (yet another ) new task 683.

[Agetian]

Ok, thanks. Here is the JDK I have running...

Code: Select all

java version "1.7.0_55"
OpenJDK Runtime Environment (IcedTea 2.4.7) (7u55-2.4.7-1ubuntu1~0.13.10.1)
OpenJDK 64-Bit Server VM (build 24.51-b03, mixed mode)


Seems to work ok, although there are a few dialogs where the text falls off the edge due to the font - not sure if that is a linux issue or VBox though?

I can confirm that upgrading to 1.7.0_65 makes the game behave correctly.
It seems that OpenJDK 1.7.0_51 may be the culprit (or some of its interaction with the other system components, I don't know). Anyhow, I can confirm that I can play matches normally under Linux Mint 15 running OpenJDK 1.7.0_65.

- Agetian

[Agetian]

The AI tapped Mutavault to activate its own ability (for a few turns in a row), which resulted in wasting mana and potential opportunity to attack for 2 each turn (the latter made sense in context).

- Agetian

[ShawnieBoy]

The AI tapped Mutavault to activate its own ability (for a few turns in a row), which resulted in wasting mana and potential opportunity to attack for 2 each turn (the latter made sense in context).

- Agetian

That's strange, the AI really shouldn't be doing that (for obvious reasons), I've looked at the card code and all appears ok. It could be that the artificial condition for colorless land-animation isn't quite right. Thanks for the find.

(For self-animating permanents that can produce mana, we fool the AI into thinking it needs more mana before it can activate it, to prevent tapping-out to animate a now-tapped permanent.)

[ShawnieBoy]

Looks like the man-lands have confused the AI (They weren't too bad before) - Can highlight the problem with the following AI-confusing deck in a mirror Match:

4 Blinkmoth Nexus
4 Inkmoth Nexus
4 Mishra's Factory
4 Mutavault
4 Petrified Field
4 Adventuring Gear
4 Infiltration Lens
4 Paradise Mantle
4 Peregrine Mask
4 Reito Lantern

The artifacts are decoys (except possibly, the Adventuring Gear)

In the test games I played the AI would rather cast more permanents than animate and attack. Only managed to animate a land on my turn, while I was tapped out; or when it knew it was losing and self-animated and equipped a tapped creature with as much stuff as it could.

Raising an issue

[jerichopumpkin]

Ember Swallower (under opponents control) triggered ability was activated when I activated monstrosity on Ill-Tempered Cyclops. I don't know if the bug is related to the trigger for mostrosity abilities or only to Ember Swallower, I'll run some tests.

Edit: I can confirm, Arbor Colossus also activates when another permanent became monstrous (Fleecemane Lion in my test)

Edit: another confirm, activating Hydra Broodmaster monstrosity triggered both Arbor Colossus and Wildfire Cerberus (both of them where still not monstrous)

[ShawnieBoy]

Ember Swallower (under opponents control) triggered ability was activated when I activated monstrosity on Ill-Tempered Cyclops. I don't know if the bug is related to the trigger for mostrosity abilities or only to Ember Swallower, I'll run some tests.

Edit: I can confirm, Arbor Colossus also activates when another permanent became monstrous (Fleecemane Lion in my test)

Edit: another confirm, activating Hydra Broodmaster monstrosity triggered both Arbor Colossus and Wildfire Cerberus (both of them where still not monstrous)

Nice find - I've pushed a fix. Let me know if there's still any issues with it