4 ms·

I can't imagine why Joel felt the question merited that kind of response. This isn't exactly the first time the question has been asked over there. Once you've

I can't imagine why Joel felt the question merited that kind of response. This isn't exactly the first time the question has been asked over there. Once you've read it for the 100,000th time I'm sure your patience tends to wear thin. Meanwhile, though it's conventional wisdom that "programming is a questionable long-term career choice": compared to what? Banking? http://seattletimes.nwsource.com/html/businesstechnology/2008556531_webwamu24.html http://seattletimes.nwsource.com/html/businesstechnology/200... A nice construction job? http://paper-money.blogspot.com/2008/12/new-residential-construction-report.html http://paper-money.blogspot.com/2008/12/new-residential-cons... Middle management at Circuit City? http://money.cnn.com/2008/11/10/news/companies/circuit_city/index.htm?postversion=2008111008 http://money.cnn.com/2008/11/10/news/companies/circuit_city/...

aswanson 18y ago
I think he meant it along the lines of how many successful (however you define that) 50 year old programmers do you run into? Most people who start out their careers doing code eventually migrate into management or something more lucrative by that point.
- potatolicious 18y ago
  That's true for all industries. That 50 year-old banker is not the guy who's running around on the trading floor all day - he's probably a manager of some sort as well. Similarly, if you go into other engineering fields (electrical, mechanical, civil, etc) nearly all the 50 year-olds are managers also.
  - gaius 18y ago
    But how does that scale? What's the ratio of managers to workers? For every 10 20-something new CS graduates, there's maybe 1 50-something manager. What happened to the other 9? This isn't a new problem, after all the Army needs only 1 Colonel for every 12-16 Lieutenants... The option to stay on as a junior officer until retirement doesn't exist, it's "up or out".
- mechanical_fish 18y ago
  One problem with discussions like this is that "programming" has an absurdly narrow definition. If you aren't actually spending the majority of your day typing code into emacs and fixing Javascript bugs, you are not a "programmer": you have "migrated into management". A senior law partner who never personally argues a case or drafts a contract, but does nothing but attend meetings all day and supervise younger partners and paralegals, is still "practicing law". The chiefs of large hospital departments are still "doctors" even when they don't pick up a stethoscope for days at a time. Late in Beethoven's career, when he was almost completely deaf, no longer able to perform or even to critique other performers, he was still a "musician". Whereas (e.g.) Joel Spolsky is now a "manager". Except when he's working on Arc, PG is more likely to be called a "manager" or an "angel investor" or a "venture capitalist" or a "technology pundit" or a "marketer" or a "writer" or a "teacher" or a "designer" or a "strategic consultant" or an "executive chef" (I think he does all of these things!) than a mere programmer. But just because these guys somehow don't fit our hyper-narrow definition of "programmer" doesn't necessarily mean that they had a problem with their software careers. It might just mean that there are many opportunities in software -- some quite rewarding or lucrative; many of which are best performed by someone who also knows how to program -- that aren't captured in the scope of the term programmer. Part of the reason why there seem to be so few "programmers" in their forties and fifties is that by that time many of them have dreamed up a better job title for themselves, like "senior member of the technical staff" or "chief of QA" or "Java architecture consultant" or even "co-founder" or "CEO". Spolsky loves to tell the story about his Bill Gates review: http://www.joelonsoftware.com/items/2006/06/16.html http://www.joelonsoftware.com/items/2006/06/16.html At the time of that review, Bill Gates had not been called a programmer for years and years. He was a CEO and a co-founder, and was paid accordingly (and then some). But what was he doing when he reviewed a spec in gory, technical detail? He was programming. At a higher level than the average college intern, yes, but just because you're not touching a keyboard or dealing with individual calls to malloc doesn't mean you're not programming. Unless we decide that it does. But then we shouldn't be surprised to find that programming as a career is something that many people eventually grow out of. We've defined it that way -- as a term to describe an activity that doesn't scale well and that usually represents only a portion of a well-balanced intellectual or economic diet.
  - 18y ago
- nostrademons 18y ago
  I know a couple - just had lunch with one of them a couple days ago. One does contracting + part time teaching at a college. Another just cashed out of a startup with enough money to never work again (after having previously gone public with a startup with enough money to never work again, but kept it in the company's stock, which tanked along with the company). Both of them have held managerial positions up to VP-Eng, yet decided it was not what they wanted to do and downgraded themselves to rank & file software engineers, albeit very good ones.
- wheels 18y ago
  You find them at big companies. I knew some great coders in their 40s and 50s at SAP. They mostly weren't the "I could rewrite this whole subsystem this weekend" sorts, more "I've solved variations on this problem since before you could write your name. If you'll shut up for a second I'll show you how it's done." I think there are plenty of places where coding wisdom is valued. They're just not the same places, generally, that thrive on geek bravado.
- silentbicycle 18y ago
  How many people born 50 years ago realistically had access to computers, though? It's considerably easier to learn programming when you can test things, and PCs weren't readily available until the late 70s. (Anyone in their 50s now would have already been in their 20s in 1978...) I was lucky enough to get a Commodore 64 when I was six or so, and taught myself simple programming stuff with library books. That's probably more or less true of several other people here. Imagine not even using a computer until after college, or learning by submitting code on punch cards, one batch per day.

aswanson 18y ago

I think he meant it along the lines of how many successful (however you define that) 50 year old programmers do you run into? Most people who start out their careers doing code eventually migrate into management or something more lucrative by that point.

potatolicious 18y ago
That's true for all industries. That 50 year-old banker is not the guy who's running around on the trading floor all day - he's probably a manager of some sort as well. Similarly, if you go into other engineering fields (electrical, mechanical, civil, etc) nearly all the 50 year-olds are managers also.
- gaius 18y ago
  But how does that scale? What's the ratio of managers to workers? For every 10 20-something new CS graduates, there's maybe 1 50-something manager. What happened to the other 9? This isn't a new problem, after all the Army needs only 1 Colonel for every 12-16 Lieutenants... The option to stay on as a junior officer until retirement doesn't exist, it's "up or out".
mechanical_fish 18y ago
One problem with discussions like this is that "programming" has an absurdly narrow definition. If you aren't actually spending the majority of your day typing code into emacs and fixing Javascript bugs, you are not a "programmer": you have "migrated into management". A senior law partner who never personally argues a case or drafts a contract, but does nothing but attend meetings all day and supervise younger partners and paralegals, is still "practicing law". The chiefs of large hospital departments are still "doctors" even when they don't pick up a stethoscope for days at a time. Late in Beethoven's career, when he was almost completely deaf, no longer able to perform or even to critique other performers, he was still a "musician". Whereas (e.g.) Joel Spolsky is now a "manager". Except when he's working on Arc, PG is more likely to be called a "manager" or an "angel investor" or a "venture capitalist" or a "technology pundit" or a "marketer" or a "writer" or a "teacher" or a "designer" or a "strategic consultant" or an "executive chef" (I think he does all of these things!) than a mere programmer. But just because these guys somehow don't fit our hyper-narrow definition of "programmer" doesn't necessarily mean that they had a problem with their software careers. It might just mean that there are many opportunities in software -- some quite rewarding or lucrative; many of which are best performed by someone who also knows how to program -- that aren't captured in the scope of the term programmer. Part of the reason why there seem to be so few "programmers" in their forties and fifties is that by that time many of them have dreamed up a better job title for themselves, like "senior member of the technical staff" or "chief of QA" or "Java architecture consultant" or even "co-founder" or "CEO". Spolsky loves to tell the story about his Bill Gates review: http://www.joelonsoftware.com/items/2006/06/16.html http://www.joelonsoftware.com/items/2006/06/16.html At the time of that review, Bill Gates had not been called a programmer for years and years. He was a CEO and a co-founder, and was paid accordingly (and then some). But what was he doing when he reviewed a spec in gory, technical detail? He was programming. At a higher level than the average college intern, yes, but just because you're not touching a keyboard or dealing with individual calls to malloc doesn't mean you're not programming. Unless we decide that it does. But then we shouldn't be surprised to find that programming as a career is something that many people eventually grow out of. We've defined it that way -- as a term to describe an activity that doesn't scale well and that usually represents only a portion of a well-balanced intellectual or economic diet.
- 18y ago
nostrademons 18y ago
I know a couple - just had lunch with one of them a couple days ago. One does contracting + part time teaching at a college. Another just cashed out of a startup with enough money to never work again (after having previously gone public with a startup with enough money to never work again, but kept it in the company's stock, which tanked along with the company). Both of them have held managerial positions up to VP-Eng, yet decided it was not what they wanted to do and downgraded themselves to rank & file software engineers, albeit very good ones.
wheels 18y ago
You find them at big companies. I knew some great coders in their 40s and 50s at SAP. They mostly weren't the "I could rewrite this whole subsystem this weekend" sorts, more "I've solved variations on this problem since before you could write your name. If you'll shut up for a second I'll show you how it's done." I think there are plenty of places where coding wisdom is valued. They're just not the same places, generally, that thrive on geek bravado.
silentbicycle 18y ago
How many people born 50 years ago realistically had access to computers, though? It's considerably easier to learn programming when you can test things, and PCs weren't readily available until the late 70s. (Anyone in their 50s now would have already been in their 20s in 1978...) I was lucky enough to get a Commodore 64 when I was six or so, and taught myself simple programming stuff with library books. That's probably more or less true of several other people here. Imagine not even using a computer until after college, or learning by submitting code on punch cards, one batch per day.