msmillie

In flame graphs heavy iteration tends to give "wider" structures. Recall the stacks aren't ordered on the X axis by time, but by contents.

Guess who's more important to the newspaper: this one guy who encountered some "you were naiive to not expect this" corruption, or the senior city hall politico? And what did he do, anyway, ask for a seat on the board in return for his help? and that position comes with a salary? We can't run that story, quit wasting our time.

http://dtrace.org/blogs/bmc/2011/02/08/llquantize/ describes a very similar log binning technique added to dtrace a few years back. In practice, it's been one of the most useful things in the toolkit.

The following scenario helps explain: you have two radars. One is a wide-angle, general radar, which sees a missle at some coordinates, travelling at some speed.

The other radar is for target acquisition, has a very narrow spread ("range gate"), and must determine the precise location of the missle given an initial set of coordinates, a velocity, and a timestamp.

In other words, the system is actually aiming at targets differently depending on how long it has been up, tracking so far in front (or behind, I don't recall the details) that it can't acquire the target.

That someone can innocently and unironically describe Trader Joe's as being "known for low prices" is really the crux of the problem here.

Perhaps you're not familiar with minimum wage?

> If the OP of the patch had gone through and consistently changed all He|She -> Them|They

If they missed any gendered pronouns in the now-landed PR, surely they would be discoverable in the current source:

  - https://github.com/joyent/libuv/search?q=him&type=Code
  - https://github.com/joyent/libuv/search?q=he&type=Code
  - https://github.com/joyent/libuv/search?q=his&type=Code
  - https://github.com/joyent/libuv/search?q=she&type=Code
  - https://github.com/joyent/libuv/search?q=her&type=Code
  - https://github.com/joyent/libuv/search?q=hers&type=Code

To a reasonable approximation, it turns out the author of PR did consistently change gendered language in the project, there just wasn't very much of it to begin with. That wasn't my assumption either (I thought there'd be more such language somewhere), and the biases that created that assumption are surely an interesting "this is how these problems happen" sort of side topic.

Anyway, definitely a lesson in where we choose to direct the benefit of the doubt. In this case, the author of the PR was the one who deserved it, but wasn't the one who received it.

(edit - link formatting)

As a Joyent engineer, I feel qualified to field this one.

tl;dr: you couldn't be more wrong.

The long version requires a crucial observation: this wasn't one simple mistake, made in haste, ignorance, or confusion. This was a deliberate action to revert to exclusionary language. Speculating on the precise motivation strikes me as pointless; it's the action itself that is problematic.

At Joyent, I have room to make mistakes, and enough respect among my colleagues that I can take corrections without misgivings. Deliberate and repeated "mistakes" on my part would have consequences (but surely that's the literal definition of a problem employee).

In case you aren't already one step ahead of me, here is the corollary: I am protected from the deliberate and repeatedly hostile actions of others.

So: the ability (even obligation) to speak your mind, listen to others, and be protected from bullying, exclusionary behaviour?

That, my friend, is what you call security.