Bugs 0.10.1

Do not post feature requests here, please.
Before posting bugs report check the Image Bug reporting help & guidelines
Forum rules
Do not post Features/Improvments request here (i.e : AI being stupid is not a bug, not being able to do a mulligan is not a bug, etc..etc..)
Before posting, please read the Bugs reporting guideline
Psyringe
Posts: 1163
Joined: Mon Aug 31, 2009 10:53 am

Re: Bugs 0.10.1

Post by Psyringe »

mickey_brown wrote:Not sure what has been mentioned and what hasn't...
The first two or known interface issues, we have tickets for them (not sure if somebody is actively working on a solution at the moment)

The third one is fixed in SVN now, thanks. :)
Psyringe
Posts: 1163
Joined: Mon Aug 31, 2009 10:53 am

Re: Bugs 0.10.1

Post by Psyringe »

Takieda wrote:Hope charm seems to be having difficulty. It works to give me life, or destroy a local enchantment, but it has yet to give a creature first strike
Confirmed. "may (ability) (target)" doesn't seem to work. I've tested a number of abilities, and also tested removing all other auto lines (in case they interfered in any way), but it just doesn't work. The odd thing is that the interface clearly tells you that you *do* cast the ability on the target - it just doesn't actually receive it.

We already have a bugticket for "may", I'll add this issue to it.
Rob
Posts: 33
Joined: Sun Aug 16, 2009 7:34 pm

Re: Bugs 0.10.1

Post by Rob »

wololo wrote:See my reply above yours which explains why the issue occurs, before you start testing ;)
That explains a lot, thank you Wololo.

I have been watching how the AI plays to find cards it is more likely to use to its advantage. I noticed that it does not always attack, even when it is beneficial to do so, so I started to test some "must attack" cards.

I had not seen your reply as I had left my post open, ate diner and came back and hit submit. Oops!
Image Image
· Poof!! · Backlash ·
Psyringe
Posts: 1163
Joined: Mon Aug 31, 2009 10:53 am

Re: Bugs 0.10.1

Post by Psyringe »

Yep - the sliver deck problem that I described must have been something else.

Regarding "mustattack", we probably should give the AI a more fine-grained evaluation of effects eventually ... currently it just rates effects as "good", "bad", or "unknown". Therefore it thinks that having "mustattack" is just as bad for a card as being destroyed, which doesn't really lead to good decisions. ;)
baldersmashed
Posts: 156
Joined: Sat Apr 25, 2009 12:00 am

Re: Bugs 0.10.1

Post by baldersmashed »

Upon looking at the coding for spell snare in DIS, it seems this can counter spells with CMC 1 as well as 2. It should only be able to counter CMC of 2.
Dr.Solomat wrote:"Vegeta! What does the scouter say about Wagics power level?"

"It's over ten thooooouuuuuusaaaaaaaaand!"

"What 10.000? There's no way that can be right!"
Psyringe
Posts: 1163
Joined: Mon Aug 31, 2009 10:53 am

Re: Bugs 0.10.1

Post by Psyringe »

baldersmashed wrote:Upon looking at the coding for spell snare in DIS, it seems this can counter spells with CMC 1 as well as 2. It should only be able to counter CMC of 2.
Yep. The code should read manacost=2 instead of manacost<=2. Fixed, thanks. :)
Psyringe
Posts: 1163
Joined: Mon Aug 31, 2009 10:53 am

Re: Bugs 0.10.1

Post by Psyringe »

Raphael_Carlo wrote:Rain of Filth's (USG)effect didn't last until end of turn but lasted in the whole game period.
Confirmed. I couldn't fix this in the card code, there may be an engine bug involved (or the card can't be coded yet with the instruments available). I opened a ticket for it.
Raphael_Carlo wrote:Mirri the Cursed's (PLC)combat damage to a creature effect didn't trigger.
Card was coded wrongly, the code is now fixed. However, there's an engine bug that still lets the trigger fail when the damaged creature gets destroyed feom the damage. Opened a ticket for that.
Raphael_Carlo wrote:Tainted Field (TOR)doesn't seem to work right. I tapped it for white or black mana and it didn't give me anything even when I have a swamp in the field.
Confirmed and fixed, thanks.
Raphael_Carlo wrote:Reiver Demon's (MRD)effect when put into play from the hand didn't trigger.
Confirmed. I fixed it temporarily, but the fix will lead to other problems once we allow cards to be played from the graveyard. We probably need some added functionality to distinguish between played / played from hand / moved to the battlefield / moved to the battlefield from the hand.

Those were good reports on tricky issues, keep them coming :)
Rob
Posts: 33
Joined: Sun Aug 16, 2009 7:34 pm

Re: Bugs 0.10.1

Post by Rob »

Flame Jet isn't working correctly.

The player can cast it, choose to put in play, select the target but then you are only presented with a cancel option. However, it does cycle correctly.

When the AI played it in demo mode, it taped mana then sat there for a while and eventually burned the mana, did not cast and continued. (This was when the AI played the bottom hand.)

Oddly enough when I played the AI myself, the AI played it correctly and targeted me.

text=Flame Jet deals 3 damage to target player. Cycling {2} ({2}, Discard this card: Draw a card.)
id=15201
target=player
auto=damage:3
autohand={2}:cycling
name=Flame Jet
rarity=C
mana={1}{R}
type=Sorcery

Image
Image Image
· Poof!! · Backlash ·
Psyringe
Posts: 1163
Joined: Mon Aug 31, 2009 10:53 am

Re: Bugs 0.10.1

Post by Psyringe »

This is a known bug that affects all cards with cycling. If you could pay the casting cost as well as the cycling cost, then casting will not work. For some cards, the cycling cost is a subset of the casting cost, so these cards can never be cast successfully at the moment.

There may be AI issues involved as well, but we need to fix the issue described above first, only then can we properly determine whether or not the AI can handle these cards.
Rob
Posts: 33
Joined: Sun Aug 16, 2009 7:34 pm

Re: Bugs 0.10.1

Post by Rob »

Rancid Earth's threshold effect did not trigger correctly when I had 7 cards in my graveyard.
I checked the known bugs and I don't see any comments regarding Threshold.

Is there another place to post AI type observations or are these pointless as the team that is working on the AI may go another way?
AI cast a lightning bolt on a 2/4 creature effectively wasting it.
AI had a lightning bolt in hand, player had 2 life and creatures in play, the AI cast a spell that targeted a creature instead of casting the bolt to win.
When the AI can choose to destroy one of several creatures he does not destroy the highest power/toughness.
(I realize this creates another issue as the most effective creature to destroy will not always be the toughest.)
AI does not choose the most effect land to destroy when casting an effect such as stone rain. (3 plains, 1 forest - may target the plain)
AI does not attempt to retrieve the Hammer of Bogardan when it is in his graveyard and he does not use his mana for anything else.
I suspect this may be true for other cards that behave like the hammer. The player is able to retrieve the Hammer successfully.

ImageImageImageImage
Image Image
· Poof!! · Backlash ·
Post Reply