Notes on new team difficulty settings

I liked BAB's explanation better. I can't make anything funny from yours.

pohsyb,

When you say "locally" is this client-side or servers side?

Originally Posted by Nethergoat it's NEVER too late to pad your /ignore list!

I was thinking "locally" meant "in a function that's off on its own instead of where the most of the other structs were declared/initialized" as opposed to server vs client. I could be wrong, though.

Originally Posted by PleaseRecycle it has gone from unconscionable to downright appalling that we have no way of measuring our characters' wetness.

Originally Posted by Brillig It's hard to beat the entertainment value of Whackjob Wednesdays.

I would have preferred he got a whipping . . .

. . .

What?

O.O


I really hope you're right.

Somehow I'm reading into this more than I should

That would be my assumption too, but I wanted to bue sure I was thinking through things properly.

Edit: Big kudos again to the Dev team, especially Synapse for keeping up with his analysis, and Pohsyb and company for figuring out where this came from. To all the members of the foil-hat brigade, please note the nature of the issue here as well as the Dev team response to it....

makes sense to me!

good stuff!

Come visit the Burrow!

Originally Posted by Black_Barrier Guess it's hard to click while actively trying to keep the drool away from the keyboard...

Local Variables. Nothing to do with Client/Server architecuture :P

If I had to venture a guess, I would think he meant a local scope variable. Either way, good to see this found out. Explains a few anomalies I've noticed here and there with the drop rates for a while now and just never bothered to look into.

While the bug has existed, I suspect the behavior we saw it cause in I16 has not actually been widespread. There are many ways this can happen. Particularly if your uninitialized structure is created on the stack and not the heap (which a local variable would be), there may have been close to deterministic values in memory where the uninitialized variables were mapped.

While it's impossible to prove now, the investigation was driven by the sense of change, not the other way around.

One thing I have learned playing this game is that people are actually quite sensitive to even small changes in existing patterns when they actually experience the patterns frequently. I once had a long debate here with Arcanaville because I did not believe this to be true, but experience here has taught me otherwise.

As someone who finds humorous the nearly ever present claims of accuracy nerfs every issue, I once found myself with a nagging sense that something was wrong with my own after a new patch. I refused to post on it for some time, convinced that it was a streak, my imagination, or both. However, after some time, I broke down and started making more careful measurements with Herostats. (This was before the hit messages in the combat logs). I found that my measured per-attack hit probabilities were consistently lower than my slotting predicted over a very large number of attacks. There was another patch only a few days later, and the "problem" went away.

So while the potential for this has been around a long time, my strong suspicion is that not-really-random factors masked it until now. In I16, with some significant code reshuffling, something else ended up in those uninitialized memory positions. Something that was a lot more inconvenient, and a lot more noticeable to people with the right experience.

i9 hey? wonder why so many people just now realized a difference because iirc i9 was when inventions came out, so if it was active then and is the same now there would be no difference right?

Since i9, but not plaguing the system till now?

To be honest, it sounds like it existed and affected things to some extent all along, but was exacerbated with some of the other items that changed in i16.

"Locally" has an entirely different meaning when referring to variables. It refers to the scope of the variable--how long its contents are kept around. A local variable usually refers to a variable stored only while the structure it is in is being executed.

Originally Posted by Frosty_Femme I said "ur" which is not a word. It's a sound dumb people make when you ask them to spell out "you are".

As I mentioned earlier, I'm fairly certain this is indeed exactly what he's referring to. I asked merely to have it authoritatively confirmed.

While it existed from i9, previous releases weren't necessarily affected in the same way as current release

</3

I work in software development (nothing remotely as interesting as games sadly ) and we've recently dug a bug out of our system that had no effect for the first 9 months it spent in our live environment. It always ends up being a combination of factors with things like this and it took some other underlying data changes for the bug to start causing things to misbehave.

It made it a complete sod to track down!

I did that back when it first became apparent through statistical analysis that recipe drop rates were lower than they should be.

I've been noticing a slow rise in recipe prices since I16, likely as the supplies dry up. Once it was confirmed by Synapse I put all my market activity on hold, since recipe prices will likely fall once the patch that fixes this goes in.