I don't see the connection between the title and the graph.
I know the adjective "cowboy" was used for some programmers in the early 1970's. BASIC, a dynamically typed language, was wildly popular for microcomputers in the 1980's. There were schema-free databases for IBM mainframes in the 70's too. For that matter, LISP was a dynamically typed language back in 1958.
Although "scrum", "extreme programming", "ruby" and things like that are new, all of the real fighting points in software engineering practice have been battlegrounds since the 1970's if not sooner.
As someone who did his first contract gig in 1987, I can say that all of those rules are conditional -- I can pound out certain kinds of projects quickly with languages like PHP and there are other ones where only C++ will do. Premature optimization is a mistake, but its also a mistake to build a system on an architecture that is doomed to forever be slow, unscalable and (worst of all) high latency.
"I don't see the connection between the title and the graph."
My thought was that startups target younger devs as they believe that they'll be more 'liberal' engineers. I think the graph shows that to be false.
edit: well I don't think the graph shows anything - but it suggests that age and liberalism are not correlated. Many people have suggested that the survey is flawed - and that Yegge's original post is biased against conservatives.
The whole thing is wonky - there is no software "liberalism" vs "conservatism".
I always assumed startups target younger devs because they generally fit startup culture more: cheap, work horses, full of energy, willing to take on risk.
I think they target younger devs for a few reasons.
When you've got a family you're more likely to want real cash instead of equity. Many startups are cash poor and they wish they could get some geniuses who'd live on ramen noodles for two years in hope of a future payout. A 40 year old programmer who's good can get a job with good salary and benefits and doesn't want to hear about it.
I think many IT employers like younger workers because they're easier to intimidate. If you've lived through a few "charge of the light brigade" projects, you eventually learn how to say "no".
Yeah. Also while I think that premature optimization is the root of all evil, I don't think it is simply true that rigid schemas slow down development. They slow down certain phases and types o development and it is a tradeoff.
Please, now that you have a half-billion records in Mongo, I want a report on statistics of these in ways you hadn't thought of before on my desk tomorrow...
The survey behind this reminds me a lot of those internet "what's your political alignment" polls that go around from time, where you always end up discovering that you're a Libertarian.
Take a quick spin through the questions and look at how they're worded. It's no surprise that the results point to "Software Liberal". I don't doubt that you could spend a few minutes rewriting the quiz in such a way that the overwhelming majority of participants ends up on the other end of the scale.
Point is not the absolute "liberalism" value but that relative values are the same through all ages. And if you'll reword the quiz than picture will be the same - age is irrelevant (for this topic).
But that's not what it shows. It just shows that people of all ages give similar answers to a set of questions designed specifically to have only one answer.
Similarly, if I gave a quiz with the question "I don't think it's right to run over kittens with my car. Strongly Agree ... Strongly Disagree", then gave that quiz to a broad spectrum of people across all age groups, you'd suddenly find that the overwhelming majority of them were all "cat people".
But if you simply asked whether they liked cats better or dogs better, you'd get a more accurate and less skewed view on the situation.
For me, that is behind a login screen. I heard a little while ago about users who have a google account but aren't logged in being unable to access G+ pages; I'll edit after a cookie-clear and refresh (which may or may not help).
EDIT: De-cookieing helped, and I can now read the article without problems. Screw you, Google. Observers be warned.
EDIT 2: Is the downvote for any particular reason? My point was not unhelpful nor especially off topic; help, anyone?
There have been a lot of studies and surveys done regarding the peak age of entrepreneurship. Very few of those studies support the idea that youth is a benefit to starting a business.
I'll post a few quotes and links here:
"In 2008, I led a research team in exploring the backgrounds of 652 U.S.-born chief executive officers and heads of product development in 502 successful engineering and technology companies established from 1995 to 2005. These were companies with real revenue -- not just the start-ups founded by the college dropouts that some venture capitalists like to fund. We learned that the average and median age of successful founders was 39. Twice as many founders were older than 50 as were younger than 25. And there were twice as many over 60 as under 20."
http://www.washingtonpost.com/national/on-innovations/the-ca...
"It does not take but one minute to look around the world and prove any thesis of a peak tech founder age incorrect. There are countless entrepreneurs over the age of 30, including Reid Hoffman (age 35 in 2002), Evan Williams of Twitter (age 35 in 2007), Mark Pincus of Zynga (age 41 in 2007), Arianna Huffington of the Huffington Post (age 54 in 2005), among many others."
http://techcrunch.com/2011/05/28/peak-age-entrepreneurship/
"“It turns out that over the past decade or so, the highest rate of entrepreneurial activity belongs to the 55-64 age group. The 20-34 age bracket, meanwhile, which we usually identify with swashbuckling and risk-taking youth (think Facebook and Google), has the lowest rate.”
Ecopreneurist (http://s.tt/12HYH)http://ecopreneurist.com/2009/09/21/the-average-age-of-an-en...
Aside from the quotes, I would also appeal to your own experience. If you are over the age of 25, then ask yourself, have you learned anything useful over the last 5 years? Have you learned important things about, say for instance, software, money, managing your time, communicating with people who are different from you, politics, self-discipline, the law, or perhaps a problem domain that you find exciting? If you haven't learned anything useful in the last 5 years, I would say you are doing something very wrong with your life. If you have learned something useful in the last 5 years, doesn't that suggest you have some new skills that will help you when you launch your business?
I think a huge factor is attitude to risk. The younger entrepreneurs are in the no-ties shoot-for-the-sky phase.
A larger proportion of people in the 30-40 range will be becoming more risk averse as they need to have financial stability to pay mortgages/support kids/spouses or whatever. But people tend to become more able to take financial risk as they get past young-kids stage, and develop enough of a savings cushion to go off on their own for a bit.
I think risk might be the opposite. Very young people might have no risk... unless they don't want to move back in with their parents.
But older people might have the least risk if their kids are grown and don't need help, their own parents gone, and they have enough money saved not to worry about failing.
Thus over 50 could have less risk than 25 and under.
Anecdotally, many of the older people I know starting businesses have their kids moved out, and either a spouse whose income is more than enough for both of them or enough savings to retire. That might be why so many people in the 55-64 age group are starting businesses.
"The younger entrepreneurs are in the no-ties shoot-for-the-sky phase."
But of course, we know that these types of endeavors have the highest failure rates, of all the different kinds of new businesses that you can start. The shoot-for-the-sky businesses seem to appeal mostly to certain venture capitalists, who are looking for the occasional big payout. And that kind of venture firm doesn't really care about the fact that shoot-for-the-sky projects have a high failure rate, since a high failure rate is already part of the venture model. But if we are talking about the majority of successful businesses that get launched in the USA, then the average age of entrepreneurs is much older than the youthful image that gets much discussed on certain startup sites.
My thought is that startups tend to target 'liberal' developers so they can iterate fast and innovate rapidly. It's the 'fuck it ship it' kinda attitude.
The thing that keeps a 40-something like me from wanting to be in an SV type of startups relate more to priorities. I've been through the grind from the first bubble, and would rather not go through it again.
I'm talking about:
* Giving up salary for equity/options/signing bonuses. No thanks.
* Working stupid long hours that bring my effective hourly rate down to a fast-food restaurant manager's hourly rate. Because of this, I'd probably be flagged as someone who has "no passion" for what I do. No, I just like a balanced life.
* Working in a "frat house" environment. I've been there and had fun in those environments, but I'm past it, and prefer a more diverse workplace.
* Working in a VC-funded company - at least in my past experience - usually means the focus shifts from ramping up quickly vs. building value and/or a solid business model
If you don't want to be in a startup, fine, that's not ageism. The problem is if you're turned away because of your age if you do. There's a perception that older folks are going to be "stodgy", overly concerned with outdated engineering principles and less with cool new features, and that startups need the latter. This attribute seems to map reasonably well onto Yegge's conservative/liberal framework, and to the degree that it does (and the degree to which the survey is representative) this seems to not be the case empirically.
It doesn't preclude the possibility, of course, that there may be other reasons someone might not be a good fit that age may predict.
Having been part of that hiring process during the first bubble, we'd drop a lot of older, skilled candidates for "cultural fit", which was related to the exact same things as to why I wouldn't really want to be in a SV startup at my current age.
* Candidate asks about overtime? Probably won't work late with us during crunches.
* Candidate won't take equity in lieu of a pay cut? Doesn't get at all what our mission is.
I know the adjective "cowboy" was used for some programmers in the early 1970's. BASIC, a dynamically typed language, was wildly popular for microcomputers in the 1980's. There were schema-free databases for IBM mainframes in the 70's too. For that matter, LISP was a dynamically typed language back in 1958.
Although "scrum", "extreme programming", "ruby" and things like that are new, all of the real fighting points in software engineering practice have been battlegrounds since the 1970's if not sooner.
As someone who did his first contract gig in 1987, I can say that all of those rules are conditional -- I can pound out certain kinds of projects quickly with languages like PHP and there are other ones where only C++ will do. Premature optimization is a mistake, but its also a mistake to build a system on an architecture that is doomed to forever be slow, unscalable and (worst of all) high latency.