[jhg]

Cultural thing - sure, but keep in mind that professional programmer with 10+ years of experience typically can do more in 9-to-5 than your average college graduate on a 24/7 schedule. He needs to be paid appropriately though, which is something that a "startup" full of college kids may not be able to afford.

[hga]

I tend to ask, can you afford not to have at least one seriously experienced programmer? It's not just a question of units of code per time, but the right code in the right (enough) architecture. As PG has noted, a whole bunch of the dot.bomb crowd were technical failures.

Another I noticed after programming C/C++ for more than decade was an ability to hone in on many bugs in seconds/minutes where the less experienced could take hours/days. How do you value that type of "productivity"?

[steveklabnik]

> As PG has noted, a whole bunch of the dot.bomb crowd were technical failures.

And those are now the guys with 10+ years of experience.

> How do you value that type of "productivity"?

There's also the productivity of not using C/C++.

[hga]

Indeed, I myself learned Lisp before I learned C. Unfortunately it was hard not to use or be forced to use C/C++ in the '80s and early-mid '90s, although what I could do with Perl 4 and early Perl 5 in the latter part of that period changed a few minds.

[steveklabnik]

Absolutely. But now's different... and someone who's programmed in C since the 90s is going to be really upset when they sign up for a startup and they're coding in Ruby.

I just generally find this discussion interesting, since I'm a 'recent graduate' with 10+ years of experience...

[dagw]

I know programmers who programmed Fortran in the 70's, Lisp in the 80's, C++ in the 90's, then switched to Java, now programs python, and is learning Scala. Just because you're over 30 doesn't mean you cannot learn new things.

However I'm guessing the assumption that they can't might be another reason why 23 year old startup founders are loath to hire them.

[steveklabnik]

I myself went from C to C++ to Perl to Java to PHP and Lisp and Python and Ruby and Haskell and tons of other stuff.

But see my comment here: http://news.ycombinator.com/item?id=1439535

[mindcrime]

  Absolutely. But now's different... and someone who's 
  programmed in C since the 90s is going to be really upset 
  when they sign up for a startup and they're coding in Ruby.

For the sake of argument, let's assume that "Ruby" in that sentence is a proxy for any of the "newish" langauges that are out there. If we can go with that assumption...

Well, let' see; as somebody who started programming in the 90's with C and C++, I would say that your assertion is simply not true. Maybe some people are dinosaurs and don't change with the times, but I made the switch from (primarily) C/C++ to (primarily) Java about 2002, and started dabbling in Erlang and Ruby about 2008, and am doing a lot of work in Groovy now, while spending some time learning Scala and Clojure.

(Speaking of which, there's a TriJVM Hack Night tonight, for anybody in the RTP area! http://www.meetup.com/TriJVM/calendar/13771811/ )

And I've written some Python and Ruby as well. So no - based on my experience (and observations of colleagues about my age) I don't think it holds that somebody who was doing C/C++ in the 90's is necessarily a fossil who won't be interested in doing Ruby (or whatever).

To be fair though, I was just starting in programming in the 90's, so I may be younger than whatever stereotypical character you had in mind by saying that. <shrug /> I still think it's dangerous to generalize in this fashion though.

[steveklabnik]

I would also argue that founders are generally not 'average college graduates.'