[functional_test]

Dang, I have the highest respect for what your moderation has done for HN.

That said, unkilling this is just wrong. This sort of drivel is what I'd expect on Tumblr -- nit picking minor things, focusing on people's phrasing rather than meaning, not accepting common terminology as if that advances some sort of social justice, and asking people (inherently Bayesian creatures) to ignore their priors for the sake of "equality".

These articles detract from the real issues, and serve only to further separate women in technology, rather than integrating them. There's a reason this was flag killed (and it was not sexism).

[robbrown451]

I hope it stays. It was very well written and balanced, and did what it set out to do, which is get me to look at a couple things I've done or defended (for instance saying "guys" or calling women "girls" in a social context) and seen them in a different light.

Also it is discouraging people from calling software "sexy", and I support that -- because it is just annoying and sounds like something lame marketing people would say.

[dang]

For the most part we don't turn off flags in this way; at most we unkill flagged stories so ongoing discussion can continue. The flagging system works well overall.

But the community is divided on this issue. Many users flag these stories, but many other users care deeply about them. Those users are also a valuable part of HN. They argue that if we never turn off the flags then we are de facto taking this issue off the table (or allowing a segment of the community to do so). That may be a bit exaggerated—some stories about sexism and/or gender in tech do stay on the front page despite flags—but it's still a fair point. Whatever you think about these questions, they're not off-topic for HN the way most politics are. Moreover, many HN users who make this case are active members who participate in lots of threads. They're not just trying to use HN to further an agenda.

If we had better software, maybe it would all balance out in a way that didn't require any intervention, but that's not where we're at (and in fairness is a pretty hard problem). In cases where the software leads to unrepresentative outcomes, it's our job to adjust. Is this article the best one for that? I don't know, but given that the thread turned out to be a vigorous discussion that managed not to be horrible, I think it was an ok call.

[deciplex]

> people (inherently Bayesian creatures)

If only this were true.

[deciplex]

sigh downvoting my post is not going to turn humans into Bayesian creatures. And, if you think humans are inherently Bayesian, I am very interested to hear your reasoning, since it would be: 1. wonderful news and, 2. fly in the face of thousands of years of human history.