Thursday, August 8, 2013

Intent and Impact

Sticks and stones may break my bones, but words will never hurt me.

I call Baloney on that childhood rhyme.

I want to use a simplified model of communication in this post: two people talking to each other. When Alice speaks, she has intent. When Bob listens to her, her words have an impact on him. Good communication happens when [good] intent has a matching impact. When the two don't align (as can be a problem for relationships, religious beliefs, and more), those words can get offensive. (Seriously, read It's not about The Onion.)

If the intent is good, but the impact is bad (sorry, I'm still simplifying), I need more information. Does Alice know her impact is bad? Has Alice actually been told this? Has Alice apologized, sympathized, or empathized? Has Alice thought about how to do better in the future? If the pattern of behavior is all bad impact with no visible effort to improve, it's just toxic waste [see #6].

Intent and impact are two sides of the same coin of communication.

I think intent does matter, but it's not the only thing. In a conversation with two equal participants, Alice's intent should have an equal weight compared to the impact on Bob. If Alice exhibits a pattern of undesirable impact that she hasn't improved (even if bungled), I personally care a whole lot less about her intent.

Impact matters.

UPDATE: Scientific American has a thoughtful blog post on intent versus impact titled "But I didn’t mean it!" Why it’s so hard to prioritize impacts over intents.

Cognitive Bias, continued

Another interesting case of cognitive bias (I encounter bias blind spot and introspection illusion often; this thought continued from earlier post) is mind projection fallacy. Oh, to see the world as it really is (link just for the synopsis of Heart of Darkness) ... or not!

Not Racist

I had a short but uncomfortable conversation with a former neighbor recently about the George Zimmerman verdict. His position was that Trayvon Martin would have been an immediate danger to me, so we are better off without him. The very next words out of his mouth were, "But I'm not a racist because I have black friends."

I need to find a word or phrase to describe what he said there. Logical fallacy? Syllogist fallacy? Ah, fallacy of the undistributed middle and fallacy of composition (with its compatriots hasty generalization and false analogy) are getting closer. Or some other fallacy? Inductive reasoning gives evidence, not proof, of the conclusion, and gets weaker when you add cognitive bias (like bias blind spot and introspection illusion). Just because you like specific example of category doesn't mean that you don't discriminate against general category, knowingly or unknowingly. Given our lame national culture, you really should assume otherwise.

I have many more examples of this behavior, so it's prevalent. If we're discussing broad topic (like racism), how does something else (like the race of a few of your friends) prove that you're free of the typical cultural bias of normative behavior (like discrimination)?

It's very frustrating. But I'd feel a little less frustrated if I could just shoot back, "Bah! Logical fallacy! False analogy!" or similar. Don't try that nonsense on me because I see the large holes making your argument look weak and tattered. And I don't think that's just naïve cynicism on my part.

UPDATE: LinkedIn apologizes for assuming beautiful women can't also be engineers. Does this mean LinkedIn would not have to apologize if the woman in the picture weren't an engineer? If so, they missed the point. How she looks in a photo does not influence how she does her job or what her job is!

Network Traffic Generation

I was recently asked about options for network packet generation (low volume and highly distributed for measurement, so cheaper slower hardware is OK), and I may as well share.

  • Ostinato aims to be wireshark in reverse.
  • tcpreplay is what it says.
  • For more control over emitted packets, look at hping.
  • Or, if you're programming in Python, look at scapy.
  • The perfSonar toolkit would handle the distributed tests.
  • SmokePing might be simpler for the small-but-distributed case.

That should get you started.

Sunday, August 4, 2013

Completion, or Continuation?

I realized that I have my own standard for doughnuts. I used to live in an apartment one block behind a Mr. Donut. Walking past the smell of freshly baked doughnuts often required a stop. My favorite was the "Chocolate Angel" - a doughnut with a light, fluffy chocolate filling that was dusted with powdered sugar. Om nom nom!

So clearly we had to go back so I could judge Daily Donuts on that scale. My report is that the chocolate filling is delicious! It's slightly denser than the Chocolate Angel, but I think the chocolate flavor is better. The powdered sugar is spread on heavier, but tasty. Happily, I don't get that greasy aftertaste that so many doughnuts have. On taste alone, I want to keep Daily Donuts in rotation. There's the personal side too. I am a fan of good customer service, and we were recognized as repeat customers on our second visit. Wow! This appears to be a shop run by a couple; she recognizes us and is very kind, while he bakes these delights. We stocked up.

So, was Doughnut Quest really completed, or does it continue? Or has it changed from a roving Doughnut Quest to an established Doughnut Habit? Hmm, I think it sounds like a Doughnut Habit. And such a tasty habit it is!