Hacker News Reader

Hacker News

Comment by kartan | Hacker News Reader

Parent Full thread

kartan

> sure they can work the young guys harder,

If you are NOT a good manager, that is all that you care about.

Did you fail to plan the rest release? Developers will do extra hours.

Did you miss understood the requirements? Developers will fix it with extra hours.

Did you decide to use the wrong technology for the project? Developers will work extra hours.

A company that hires only people that does what is told and stays longer hours have incompetent management.

> I wonder if it’s a good for efficiency.

Companies that value diversity (age, gender, nationality, etc.) are the ones looking for that high-quality output that makes products work in the long term.

Sweden is as successful as China when it comes to software, even that there are orders of magnitude of the difference of population. The main reason is the respect for family and employees. Even that Sweden also is far from perfect.

View on HN

>Companies that value diversity (age, gender, nationality, etc.) are the ones looking for that high-quality output that makes products work in the long term.

What’s the evidence behind this assertion?

Googling to me suggests its not as well studied as it could be; there are papers out there (eg: https://www.theseus.fi/bitstream/handle/10024/63581/Thesisxx... and https://www.sciencedirect.com/science/article/pii/S221256711...) but I'm not seeing anything with a lot of solid data out there (though this probably is a hard subject to quantify in such terms).

Intuitively though this position makes sense to me that, as long as the various issues surrounding natural human tribalism are managed, diversity is generally helpful -- as each "tribe" may offer a unique perspective towards problems and solutions.

Having said that, a lot of articles on "ageism in tech" more or less seem to focus on cultures of work that promote "long hours" for long hours sake, over something more balanced (and, in the end, productive and efficient). There's plenty of evidence out there that shows how wrong-headed of an approach "long hours" cultures are, this is not hard information to look up.

> may offer a unique perspective towards problems and solutions

This is my impression on why diversity generally works well. Unique perspectives and experiences often generate more original and creative approaches to some problem which in turn is much more valuable. Additionally, and especially for FAANGs, having the organizational awareness of a diverse set of perspectives allows you to create solutions that are broad and can adequately meet the inherent diversity in the world.

People often want to look at software engineering as this super technical line of work, but in practice it really isn't. How, What, and Why to engineer something varies significantly. Everyone will probably have a different approach to solving a problem. More diversity means a greater net to catch which one might be the most optimal.

Plus, you might want to create services or apps that specifically target a need. I would honestly question how well an app for women might be if is made by all men. I would question how well a app for rural living might be if it was made by people who only lived in a city. I would question how well an app designed for use in America might be for use in Africa. There are cultural differences and use differences, needs and wants, infrastructure, etc.

It would be assumed a highly diverse team would be much more suited to trying to solve those problems. And I would also believe that a highly diverse team would be the result of authentic merit based system anyways.

There is no evidence for this (I am not saying whether or not this could be empirically verified, I have no clue), but it is politically incorrect, especially on this website, to claim anything otherwise

I think this is actually stated in the reverse, based on my experience. Companies that value high-quality output generally tend to become more diverse as they grow. There are obvious places this falls on its face. I used to manage plenty diverse global team of engineers (China, India, Brazil, Mexico, US, and a few in the UK) but that doesn't mean diversity was a cornerstone of our culture. Our CIO just wanted to cut expenses and do as much as possible in low cost locations. Software quality was predictably mediocre, not because there weren't some good engineers but because the short view was valued far more than the long view, and every dev cycle was a series of compromises followed by releases that were series of P0 emergencies.

> I used to manage plenty diverse global team of engineers

To have Chinese engineers in China, Brazilian in Brazil and Mexicans in Mexico is NOT diversity. That is the contrary of diversity. That is to have homogeneous teams.

I have worked for companies in Stockholm that have people from India, Canada, Australia, Sweden, Spain, Ukraine, Poland, ... working together in the same office. That is diversity. All of them were had the same range of salaries and expectations. Of course, it was not fairytale perfect. But it was quite good.

Externalizing costs to a low-cost country has nothing to do with diversity, as I understand it.

So my global, remote team isn't diverse... because they're remote? The shared office doesn't matter IMO.

I believe it boils down to how often they all interact with each other. If they're all part of one team, and are regularly interacting, that's one thing, but if they're all kind of self-contained units that don't interact with each other normally, that's not that diverse.

> To have Chinese engineers in China, Brazilian in Brazil and Mexicans in Mexico is NOT diversity. That is the contrary of diversity. That is to have homogeneous teams.

I'm curious what definition of Diversity you're using that a team of people from all over the world doesn't qualify.

They're saying that if you have one team that's all Mexican and another team that's all Chinese, neither of these teams are diverse. Your company might have diversity if you look at the employees as a whole, but segregation is not part of what diversity is supposed to mean.

Ethnicity is not the only measure of diversity. To boil diversity down to one metric like that is missing the point.

You're right, but I think it's still important to consider. If I had a team of engineers that were from diverse cultures, but all happened to be male, I would feel that lack of (gender) diversity would still be a fair criticism. It's not boiling it down to one metric, it's pointing out that there's an important metric that might have been missed.

I didn't boil down diversity to one metric. I was responding to this:

> I'm curious what definition of Diversity you're using that a team of people from all over the world doesn't qualify.

What you wrote would work better as a response to that than to what I wrote.

Yea i echo this - genuinely different backgrounds, influences on business processes and styles as it pertains to the company is discountes as not diverse because of where they are pgysically located.

Like an office in sf with all genders and colors represented is diverse...even tho they are all impacted by being im sf and sf culture

I should have been clear. People across these teams were working together on projects, so in that context it meets your bar (as far as I'm concerned). That said, when I started at Google there were 7 countries represented in my initial team of 15 people, which is an entirely different situation.

here is a nice piece of article about that: https://www.scientificamerican.com/article/how-diversity-mak...

Maybe software for international markets. But there's a lot of software produced in China for the huge domestic market. I doubt Sweden has the same software output as China.

Per capita, I'm pretty sure Sweden has a higher output.

Per capita they are also the largest arms dealers in the world. Per capita doesn't say much about a country overall.

It's the only argument that makes sense for disparate population sizes.

The US and UK (and less so Germany, France, Israel and Japan) produce a sizable majority of software products, especially at an enterprise level. At a per capita level, that gap is even larger with China; and countries like Sweden and Norway become top players.

> It's the only argument that makes sense for disparate population sizes.

Or you accept that a comparison simply can't be made and don't make the argument in the first place.

Incorrect.

It’s a meaningless comparison, the median Swedish/Norwegian citizen is far more educated, wealthy, and capable of producing software than the median Chinese citizen.

China’s per capita software production is dragged down by its massive rural population, many of whom have never even owned a computer other than a smartphone. If you don’t account for that, then obviously wealthier/more urbanized countries will have higher per capita production.

China is 55% urban. Its had a massive urban influx, and in fact the largest growth area in population is urban.

However 1 in 5 are 'migrant workers' with no fixed address. I'd guess they are the least educated demographic.

There are dozens of us, dozens!

Judging by any Stockholm coffee shop as well we’re all programmers, so the math checks out.

Many management issues you mention are not cut and dry. The fact is early stages of a product development tends to be experimentation and fast iteration. Experienced, heavy hitters don’t work well in this situation, they are great when the product has matured and you need stability and need to fix the last 10% of the issues. Hiring new grads who can code certainly helps in man many situations, but not all.

It's exactly about the company i currently work. They're just adding more and more developers.

>A company that hires only people that does what is told and stays longer hours have incompetent management.

Only when quality of end product / developer hapiness is concerned.

Which, lip service aside, are not metrics that managers are usually measured buy.

Do they bring in profit?

Sounds like the TSB way to manage things :-P

Reply on news.ycombinator.com