This Reuters article has more detailed info. Among U.S.-based tech employees the stats are 3% Hispanic and 1% black, vs. 4% and 2% respectively among Facebook's global workforce. 17% of its global tech employees are women.
As one data point of comparison, here is some demographic data for AP Computer Science test takers in California for the year 2012. Looking at students who take the AP exam may be a good proxy for identifying students who will one day be applying for top-tier positions. Among this group, 7% were Hispanic, 1% were black and 21% were women. If those stats are representative of the pool of top-tier talent in the workforce, then Facebook isn't far off in terms of its hiring of blacks and women. It seems further off with respect to Hispanics. Though, California has a higher-than-usual Hispanic population, so maybe nationally the % of Hispanic AP Exam takers is less than 7.
This article in USA Today also has some stats. They looked at the demographics of CS and CE graduates from "top" U.S. universities. Not sure what "top" means. They claim that 4.5% of such graduates are black, 6.5% are Hispanic. They didn't report on what % were women.
I recant the part about Apache then, and limit my answer to running Rails in process-per-request mode. Unfortunately the project I work on has a ton of legacy code that's probably not thread-safe, so rather than go through the pain of refactoring it it we've elected to just buy more RAM, i.e. have more instances in the pool.
Was going to be sarcastic and answer "Ruby on Rails", but I'll go with "process per request" instead. That is, the Apache model, or for Rails the model where you keep a pool of instances that only handle one request at a time.
Terrible in terms of scalability, but generally works for small workloads. Plus it largely sidesteps developers having to understand how to write thread-safe code.
No list, obviously. But it's easy to make a reasonable guess. I don't work on crypto. I don't have a lot of friends from "suspect" foreign countries. I'm not a terrorist. I don't associate with terrorists. I don't say incendiary things about elected officials. I'm not an "activist" of any kind. I don't traffic in protected IP, child porn, weapons or drugs. I'm not a black hat.
Basically, the NSA has no reason to care about me. I'm a nobody. And my online profile makes it extremely unlikely that I'd be tagged by an algorithmic solution.
Did you miss the part where I said I don't begrudge them their decision to leave for greener pastures? Employee loyalty to company is almost entirely dead. Company loyalty to employee is also dead. It's not fair to employers to expect them to be loyal to their employees when those same employees aren't loyal to their employer.
Honestly, I'd rather my employer not risk spending money to train him (not to mention having to keep him on for another six months) only to have him perform exactly the same at the end of that period. I'd rather "we" as a development organization start to have actual hiring standards, and apply those standards retroactively to current employees based on the body of work we've already seen from them.
"They" (meaning employees where I work) frequently "discard" my employer whenever a higher paying position comes along; there's no real "company loyalty" any more. (And I don't begrudge them that decision.) So why should my employer feel an obligation to not respond in kind? Besides: I'm not convinced training would help. The guy writing bloated, buggy, poorly performing code isn't going to magically starting writing clean, robust, succinct, performant code if we send him to a 3 week class.
I'm relatively certain I could go get another job inside a week or two, possibly making more than I do now. So I'm winning to risk being the one who gets laid off. It's pretty liberating, actually, being in that position. Or, at least, perceiving myself to be in that position. I can speak a-politically about things at work (hard truths, etc.) and have no anxiety about possible consequences.
Certainly seems as if the market considers it the right move w.r.t. the success of Seagate, Inc. If you disagree, and feel Seagate should have retained those 6500 employees, why should that be the steady state? Couldn't Seagate afford to take on additional employees? What should factor into Seagate management's calculus when they try to determine what the size of their workforce should be? As large as possible with the constraint that the company remain profitable? Whatever size maximizes long-term growth? Something else?
Honestly, I often wish my employer would lay off 15% of the company. Or, better yet, lay off the lowest-performing 30% and back-fill half of those positions with more competent new-hires. Productivity would stay roughly constant but payroll would shrink by ~10%. (This assumes the new hires would need to be paid more in order to guarantee the caliber of employee likely to be more productive than those being replaced.)