[6nne]

You have this a bit wrong.

If ability is normally distributed for a given sex (and the mean of the ability of the same), then it is true that there would be more above-average male programmers than female programmers - just as there are more male programmers than female programmers at any skill level. If you are provided a skilled programmer, they are most likely male. And if you are provided any programmer, they are most likely male. However, If you pick a random male programmer and a random female programmer, the woman is just as likely to be a good coder as the man.

This means that if you are talking to an anonymous person online and they are a good coder, you should assume they are male (if you're going to assume anything). This does not work backwards - if you are talking to a female programmer and you do not know their coding ability, it would not be particularly likely that they are a poor coder.

[mcv]

You should not assume anything. Given any random coder, yes, the chance that they're male is larger than the chance that they are female, but the chance that they are female is still present, and if you jump to the wrong assumption too early, you end up effectively denying their existence, which is a big part of the problem.

Consider it from the point of view of the female coder: in every discussion, people will always start out assuming either that she is male or that she can't code. And it's up to her to either correct that (with all the crap that often comes with it) or not, and let people believe there are no female coders. Constantly running into people assuming you don't exist, can be very demotivating, and may lead her to look for a more welcoming job. Let's not do that anymore, okay?