As a man with a marginally popular blog and what I'm told is an extremely masculine writing style, I can assure you that virtually no one applies the principle of charity.
Then go read the top comments either on the blog or HN. The top python geeks (read: core numpy contributors) in the thread had no dispute with me, yet there was a lot of clueless disagreement suggesting I never heard of multiprocessing/zeromq/twisted/etc. The entire point of the post is that multiprocessing+whatever is fundamentally slower/uses more memory/has worse cache locality than shared memory+threads.
People don't read. They don't apply the principle of charity. They don't follow your links to verify things. This is not a gendered problem.
(If you look for popular posts on my blog you can find lots of other examples, I'm just sticking to a purely technical topic.)
It's not a binary; more people apply the principle of charity more often in certain situations than others. It is not mutually exclusive that a certain amount of people don't apply the principle of charity in situations where they feel things about programming and Python specifically, and that a certain amount of men feel things about gender and don't apply it. I feel that "things women write" in my above comment probably should have been "things people write" but otherwise I still see the problem as reflexive rejection by men of tech-and-sex/gender pieces as "oh boy another feminist telling me I'm evil" instead of more rational "I'm not a woman so I should listen to what women have to say about sex/gender in tech even if it turns out I disagree."
I think men in the industry would be more willing to default to your more rational, and also more charitable, approach, had they not so much experience of feminists telling them they are evil. After a while, anyone will get gun-shy.
Exaggeratio ad absurdum is the besetting vice of identity politics. Nuance ceases to matter, and then to exist, because everyone involved comes to feel that his existence as a human being is on the line. For those who make such belief their bread and butter, that's great! For everyone else, it gets hard to take after a while.
Hm. When I made up a quote for the hypothetical man I was not actually trying to posit that this man's encounters (as representative of the tech/programming everyman's) with feminists consisted of him being told that he is evil. I actually value what feminists write, something that you didn't seem to pick up (not that I explicitly put it down).
Also you characterize my second "approach" as "charitable". This is interesting; spending time listening to women is apparently charity, with "oh boy another feminist move along" as the "[less][!] rational" baseline.
I used "charitable" as shorthand for "in accord with the principle of charity", and did not consider the possibilities for misinterpretation thus created. I think it reasonable to hope that the principle applies nonetheless.
I similarly used "evil" in somewhat less than the strictly literal sense; while I understand that someone very familiar with feminist theory may not apply a label like "sexist" with pejorative intent, both trivial surmise and empirical evidence indicate that someone without such familiarity is likely to react defensively when the label is applied to him, because whether it's meant pejoratively or not, it comes across as though it were. Simple pragmatism suggests that the use of the label, as a means of convincing people they should modify their behavior, is counterproductive -- but this, of course, is a tone argument, and therefore verboten.
As you, I also find much value in feminist discourse, which offers suggestions whose adoption would, I think, considerably improve not only our industry, but our society at large. Unfortunately, feminist activists generally have the least facile grasp of tactics for social change of any such movement I've ever seen, studied, read about, or heard of -- if your audience think you're browbeating them, you're browbeating them, and browbeating people is generally not an effective method of convincing them of the rightness of your cause. (I know, I know -- another prohibited tone argument. In my defense I can offer only the axiom that whether or not something rightly should be true has never been demonstrated to have a discernible effect on whether or not it is.)
Of course, when persuasion fails, there's always Alinsky's method: organize to secure power, and use that power to impose punitive sanctions on those who fail to conform. Modern feminist activism seems to me to include an increasing number of people who would happily follow precisely that program -- but who, thus far, have been hamstrung both by their general tactical clumsiness and by their earnestness, an insurmountable obstacle to power in a political system so thoroughly corrupt as that of the United States. Perhaps that'll change, and perhaps it won't; either way, I will not, indeed in good conscience cannot, ally myself with a movement willing to countenance the use of such methods, however fervently I agree with its goals.
Update: You seem to take issue with the continuum of rationality I define in my prior comment; am I right in surmising that, where I use "more rational" and "less rational", you would instead use "rational" and "irrational"? If so, I respond thus: I'm all in favor of extending someone the benefit of the doubt, but when sooner or later it becomes apparent that only one side of the conversation is actually trying to have a conversation, doubt can no longer fairly be said to exist.
For example, consider this post I wrote criticizing my third favorite language: http://www.chrisstucchio.com/blog/2013/why_not_python.html https://news.ycombinator.com/item?id=5986158
Then go read the top comments either on the blog or HN. The top python geeks (read: core numpy contributors) in the thread had no dispute with me, yet there was a lot of clueless disagreement suggesting I never heard of multiprocessing/zeromq/twisted/etc. The entire point of the post is that multiprocessing+whatever is fundamentally slower/uses more memory/has worse cache locality than shared memory+threads.
People don't read. They don't apply the principle of charity. They don't follow your links to verify things. This is not a gendered problem.
(If you look for popular posts on my blog you can find lots of other examples, I'm just sticking to a purely technical topic.)