Collectible Card Game Headquarters

Pux wrote:Somebody already reported it in this thread several pages before, but the game still crashes when the AI plays Squandered Resources.
This bug is pretty serious since Anansi in Jamuraa plays it. If the bug is hard to fix I'd suggest to take the card out of the deck.

R20305: Fixed AI crashing on Squandered Resources (but it still can't use it). Thanks Pux.

Diogenes wrote:More complex bug: I double-spliced Eye of Nowhere with Reweave and Psychic Puppetry, each with its own target, and got the following NPE:

...

On a related note, I've noticed that when a stack action resolves for a spliced spell, the user receives a whole bunch of target selection prompts that don't indicate which effect is getting the target. You can tell in advance by looking at the order of the spliced cards in the list that pops up when you cast the original spell, but even that lists each spell's effect after its name, and the line can easily run off-screen after the first addition. It would be nice if the target-selection prompt reminded you which effect you were targeting, and it would also help if the list simply showed the spell's names rather than appending their entire text.

Both fixed! Thanks Diogenes.

r20302
I cast Crippling Fatigue on my opponents Spellgorger Barbarian. It went to the graveyard and triggered my Cemetery Puca. I paid one, figuring I can get something better later. After the trigger finished resolving I found myself missing a Puca. It seems that it was Cloning the actual P/T and not the clonable characteristics of the Barbarian.

It seems that AB Destroy/DestroyAll effect is not operational at all, or at least largely - yesterday I reported Wall of Corpses not destroying the blocking creature properly, today I've encountered the same bug with Abu Ja'far - when it died it did not destroy the blocked creature. Both cards seem to use the same basic effect (destruction), so maybe all the cards or most of the cards with this effect are affected.

- Agetian

Agetian wrote:It seems that AB Destroy/DestroyAll effect is not operational at all, or at least largely - yesterday I reported Wall of Corpses not destroying the blocking creature properly, today I've encountered the same bug with Abu Ja'far - when it died it did not destroy the blocked creature. Both cards seem to use the same basic effect (destruction), so maybe all the cards or most of the cards with this effect are affected.

- Agetian

I think the common issue here is actually that they check the blocker/attacker after being removed from the battlefield. Probably need to use LKI.

Here are some card-specific problems that I've encountered recently.

Man-o'-War: Infinite loop -> crash when AI cast this. Apparently, Sloth has just fixed this. AEther Adept seemed to have a similar problem, though, and I don't know if the fix helped that one, too (haven't tested yet).

Necromancy: Crashes when the AI casts this. I replaced this with the 5th Edition Animate Dead in the Jamuraa decks because of this.

Smoke / Winter Orb: This is an old bug that still persists. More often than not, you are never prompted to untap that one allowed land or creature, which puts you at quite a disadvantage against the AI. The only time Winter Orb seemed to work properly for me (i.e., I was prompted to untap one land) was in the infamous threaded input revision, which had some major issues of its own (general slowness, misbehaving prompts) - this leads me to wonder if this problem is somehow input/prompt-related?

Contagion: Crashes when the AI casts this. (Not always?) Ok, so the spell is AI-tagged and not supposed to be included in the AI decks so I replaced it with something else (functionally or thematically related, from the roughly same era) in the Jamuraa decks.

Agetian wrote:It seems that AB Destroy/DestroyAll effect is not operational at all, or at least largely - yesterday I reported Wall of Corpses not destroying the blocking creature properly, today I've encountered the same bug with Abu Ja'far - when it died it did not destroy the blocked creature. Both cards seem to use the same basic effect (destruction), so maybe all the cards or most of the cards with this effect are affected.

These are two effects with very similar abilities, and it is definitely not "largely." Regular destroy/destroy all effects work just fine. As marc said, these are caring about battlefield specific properties, that no longer apply when the base card is no longer on the battlefield.

friarsol wrote:

Agetian wrote:It seems that AB Destroy/DestroyAll effect is not operational at all, or at least largely - yesterday I reported Wall of Corpses not destroying the blocking creature properly, today I've encountered the same bug with Abu Ja'far - when it died it did not destroy the blocked creature. Both cards seem to use the same basic effect (destruction), so maybe all the cards or most of the cards with this effect are affected.

These are two effects with very similar abilities, and it is definitely not "largely." Regular destroy/destroy all effects work just fine. As marc said, these are caring about battlefield specific properties, that no longer apply when the base card is no longer on the battlefield.

Even if not largely, it affects a certain type of cards and it's definitely more than one card that is affected. I don't think it's a reason to just forget about the bug on the grounds that it "has been there forever". Hopefully someone will eventually fix this... Things like that really make me wish we had an actual issue tracker so that problems like that could at least be put on a list for somebody to pick up eventually instead of just forgetting about them.

- Agetian

Agetian wrote:Even if not largely, it affects a certain type of cards and it's definitely more than one card that is affected. I don't think it's a reason to just forget about the bug on the grounds that it "has been there forever". Hopefully someone will eventually fix this... Things like that really make me wish we had an actual issue tracker so that problems like that could at least be put on a list for somebody to pick up eventually instead of just forgetting about them.

We did have an actual issue tracker and noone used it. I was the one of the few people even monitoring it. I wasn't trying to say forget it, I was trying to say "don't blow it out of proportion" 5 cards not working in Destroy/DestroyAll is much different than the whole API not working.

I'd say we might actually be getting close to the stage where using a bug tracker would almost be essential - I mean, with so many supported cards, with so many different rules and their interactions, and with so many game modes and subsystems we definitely do get a lot of bugs posted here, and not all of them are easy enough to fix to be fixed immediately. Still, unless issues like that are fixed within foreseeable future, without an issue tracker they would just be forgotten, then re-reported by somebody eventually, then forgotten again, then re-reported ad infinitum. I think it's actually worth considering getting an issue tracker in the future, I don't know what the others would think but the bigger the project and the more (and bigger) the bugs, the more of a necessity it is, in my opinion.

- Agetian

Agetian wrote:I'd say we might actually be getting close to the stage where using a bug tracker would almost be essential

 
There is/was a Mantis based bug tracker on the cardforge site. It may still be in place and be functional. Forge no longer has a button which will connect you to this bug tracker as most people were confused about how to use it. Most bugs should be sent to our Forge forum.

The Forge wiki should have instructions for using the Mantis based bug tracker on the cardforge site. Getting the vast majority of the user base to use this bug tracker may be a difficult goal to reach.

Hmm, maybe it should then be up to the devs to put the issues which are not immediately or quickly fixed on the bug tracker if it's a major hassle for an average user to actually use the system?.. (or something like that, just tossing ideas, really).

- Agetian

I would ask, why Mantis became became unused.

Was it because of requirement to mantain another account on a different site?
Or due to an insane number of fields to be filled to submit a bug?
Because of its ugly interface?

Max mtg wrote:I would ask, why Mantis became became unused.

Was it because of requirement to mantain another account on a different site?
Or due to an insane number of fields to be filled to submit a bug?
Because of its ugly interface?

I would say "all of the above". it's also somewhat impersonal, and it seems most users are more comfortable with the forums.

I regularly check the forums for bug reports related to the changes I'm making. I could start adding any bug reports I see to Mantis so they don't get lost..

Max mtg wrote:I would ask, why Mantis became became unused.

Was it because of requirement to mantain another account on a different site?
Or due to an insane number of fields to be filled to submit a bug?
Because of its ugly interface?

Yea I think it was all of the above. Back when we were still on google code, there was a fair amount of use with the issue tracker since it a) used the same account and b) was easy/clean enough to add things. Of course, not very many forum users were posting issues, they were posting to the forums (as things are now).

But it was also at a centralized location. I was already going to google code to glance at the commit logs, it was a simple step to move over to the issue tracker to report/update something.

We could post long term issues to some issue tracker (I'm not sure why Rob chose Mantis over some other Tracker, maybe because of hosting restrictions?), so they don't get lost, but it certainly wouldn't prevent someone from reposting them.

I guess the important thing about long-term bugs is that they don't get lost so that there's some central repository each developer can access and scan through the list of previously reported issues in case he may be willing to tackle one of the problems. Yes, it won't prevent people from reposting, but then again, it's still beneficial because it can make it more transparent to the developers if a certain bug is a duplicate of a previously reported and known (and registered) bug or a brand new issue.

- Agetian