|
|
Post by MintDragon on Jul 2, 2016 16:35:41 GMT -5
On the prospector's path, my team unglommed, but one of them ended up in an adjacent area. Original team, Vraes leading.  iPad Air 2, iOS 9.3.2, HoS 4.2.39 |
|
|
|
Post by MintDragon on Jul 2, 2016 17:23:21 GMT -5
adding one more similar, still in prospector's path. It feels like the proximity logic calculates that certain points on the map are closer than un-glomming them single file. Definitely not a game breaker in the below example, but another unexpected outcome.  |
|
|
|
Post by fallen on Jul 2, 2016 17:54:09 GMT -5
adding one more similar, still in prospector's path. It feels like the proximity logic calculates that certain points on the map are closer than un-glomming them single file. Definitely not a game breaker in the below example, but another unexpected outcome. Thanks for the pics. We've improved this code over the years, but in some cases it still can be messed up. We've added these cases to our list to examine. |
|
|
|
Post by Someone on Jul 2, 2016 18:03:50 GMT -5
Hahaha!
I particularly like when you unglom and pop! you're in a store or other new map... sometimes it's faster than walking, even with fast walk. ;-p
Not fun when it takes you two turns out of combat by popping you into a new tunnel tho... I've had some hurt done when a tank has the long walk of shame imposed on 'em. Another reason for mp buffs.
|
|
|
|
Post by Cory Trese on Jul 3, 2016 12:43:43 GMT -5
Hahahaha! We've reduced the un-glom issue repeating rate to about 1 in 8000.
We will keep trying to reduce it until it is averaging 1 in a billion.
Thanks for all the feedback and posts.
|
|
|
|
Post by anrdaemon on Jul 3, 2016 13:19:36 GMT -5
May be you can check if heroes can reach one another in 3-4 steps (ignoring themselves blocking the way)? Should be quite doable, and would save in all cases.
|
|
|
|
Post by Cory Trese on Jul 3, 2016 13:20:34 GMT -5
May be you can check if heroes can reach one another in 3-4 steps (ignoring themselves blocking the way)? Should be quite doable, and would save in all cases. yeah, definitely doing that. Thanks for the suggestion! |
|
|
|
Post by anrdaemon on Jul 3, 2016 21:56:52 GMT -5
Another idea (though, I don't know, how doable it would be), is to remember the last 4 tiles the group moved through. Kind of a "cursor tail" from the ancient Windows versions.
|
|
|
|
Post by Cory Trese on Jul 4, 2016 11:00:34 GMT -5
Another idea (though, I don't know, how doable it would be), is to remember the last 4 tiles the group moved through. Kind of a "cursor tail" from the ancient Windows versions. Yeah, the code does that too. |
|
|
|
Post by Someone on Jul 4, 2016 13:49:59 GMT -5
I imagine bugs which are rare are harder to find the best fix for.
I must get it more often, cause it's happened to me at least a half dozen times and I don't think I've done more than 8k ungloms.
I'm on hiatus atm due to Amazon issues anyway.
|
|
|
|
Post by Cory Trese on Jul 4, 2016 20:43:47 GMT -5
I imagine bugs which are rare are harder to find the best fix for. I must get it more often, cause it's happened to me at least a half dozen times and I don't think I've done more than 8k ungloms. I'm on hiatus atm due to Amazon issues anyway. I don't expect the distribution to be average across players, just that is the rate that it occurs across all sessions. |
|
