Minor chain bug caught on screen capture

[burtosis]

I was screen capturing to see what x chain I could get on the blueprints and got the chain bug on a screen capture.

About 2% of the time, for some reason, a chain does not go off right - it just starts over as a match. On a blueprint, after you win it, if you have chains stacked up they continue to fire.





See the first shot - the wind falls and fails to chain. This only happens rarely (guesstimate at 1 in 50).

[Jason Citron]

Crazy. Thanks for the heads up. Any idea if you are doing something in particular that is causing it to not fire?

[burtosis]

I have seen this about 20x, it happens about once in 50 to once in 100 times you chain - so it is not very noticable in most cases.

I can't find anything in common with the various times that I have seen it - so sorry but I don't have an idea...

If it helps, I have an iphone 3g with the latest update 2.1 (5F136) and the latest update on AF.

Edit: Not the latest AF (just noticed the update). I was using the previous version which, IIRC is 1.0.1.2

I am upgrading to 1.1.0.0 and will post if I see it again.

Edit again: noticed that I said iphone when I meant iphone 3g sry.

[burtosis]

After upgrading, I got the bug again.

Not a big deal anyhow. I really like this game, so keep up the awesome work.

[jemni]

I observe it probably about the same frequency. I haven't narrowed it down to how exactly it occurs. It's not exclusively after a device flip, but maybe more common. Or that timing may be a separate issue. Any chance of getting that timeout increased?

[burtosis]

FWIW I almost never rotate the device - so many of the times I have seen it, the device has been upright and not rotating.

[jemni]

Jason, Danielle, are you any closer to finding the source of this bug and fixing it?

[burtosis]

I have version 1.1.0.1 (newest IIRC) and this bug has continued to hit me.

Makes getting over a x200 very hard. I lost a x220ish chain to this bug on sunday...

[jemni]

Oh, stop it with your x200 chains! You're making the rest of us feel bad.

[burtosis]

An important update to finding the cause of this bug is also in this thread:
aurorafeint.proboards100.com/index.cgi?action=display&board=gamebug&thread=1286&page=1

Edited to add:
This bug does not hit if you have multiple overlapping timing (i.e. the stack of chains/combo popping timing is 2 deep or more). I have never seen the bug when this this is the case - chains appear to work perfectly. But if you are on one set of timing (i.e all popping bricks on the screen pop at the identical time) then this bug will hit occasionally.

FWIW I have seen this bug hit with horizontal (as above) and vertical chains.

[burtosis]

Update 1.2 is here and I downloaded it. Unfortunately, this bug is still present. I haven't played enough to be sure (could be chance) but this bug looks as if it might be more frequent now, I will try and find out by playing more!

Anyhow, keeping up a multichain (multiple timing on when the bricks pop) still keeps this bug at bay.

[Subaru]

Can someone tell me what CPU is in iPhone??? If ARM and coding is similar to what I know I think I know what can be problem
It's just theory
Let's say we have endless loop that makes game working whole time
CPU should refresh whole thing in 60 FPS
so whole endless loop should take one sec
Theory is that sometimes it jumps over and don't catch chain I mean it checked but didn't found chain and jumped over
I know a bit complicated but developers will probably know what I mean ;D

[burtosis]

I think I see what you mean, though I am not sure how the debugging differences work between an actual device and an emulator, much less how different compilers/OS for/on different devices could influence the occurence of the bug. I know that the few times I have written code I sure tried to find if the fault was in something I didn't have control over

I have seen this bug perhaps 200 times now and have not been able to replicate the logic behind how it happens. It appears to effect all kinds of chains. Keeping a multichain up still lets in the bug, but it counts as a match of 3 and keeps the current multiplier. So, it is probably a book-keeping problem as you mention. Either the check for chain function ends prematurely due to a framing problem, never counts some bricks because of a logic exception case problem, or whatever, the one thing I am fairly sure of is it would not be something that a user of the game could easily pick up on (unless they wrote the code).

It would be nice to know what devices this is happening on, perhaps is in the interface between the code and OS version that creates the problem.

Hopefully this will be fixed for the arena version, because there it would impact the gameplay more, imo, and I expect more out of paid software than free.

This is still my most favorite game for the iPhone (though I have a few other favorites as well) and is worth the few bugs because of the overall game play and free price.

[burtosis]

So I have been playing 1.2 for awhile now and have seen this bug about 10x so far. It has been happening closer to 1 in 20 times now or about 5% of the time.

Also, I swear some minor timing/rule changes went into 1.2 but I haven't been able to put my finger on exactly what the changes are. So, perhaps some of these changes were responsible or indirectly responsible for the increase in the chain bug.

Hopefully the devs can fix this as it is now a worse problem then before.

[burtosis]

OK,
Just loaded up 1.2.1 and the chain bug is now at the same frequency as before, but it is still here. I thought the timing/rules were the same as before (1.1.0.1) too - so either this rule/timing change has something to do with it or I am just imagining it...