Hacker News new | past | comments | ask | show | jobs | submit login
Years of irrelevance (37signals.com)
24 points by pius on Feb 5, 2008 | hide | past | favorite | 9 comments



Is it really 8 years of experience? If you spent the last 6 years doing the same thing basically, it's really just 2 years with lots of sets/reps.. Years of experience are overrated and I think HR folks are just trying to control IT/tech departments by imposing their standard lingo/Y.O.E. requirements

In my first programming job, the team was busy with an fix-the-alpha/get-the-beta-out crunch so they assigned me and the other new member a game project to code. I did networked Othello, using Java Swing over Java RMI. The code design was poor since I didn't know better but it kept us busy and we got to present it to the rest of the team (it was a nice diversion from their crunch-time). The code review was strict but good natured (as they knew we didn't know better). But we learned.


> If you spent the last 6 years doing the same thing basically, it's really just 2 years with lots of sets/reps.

There was a comment of TheDailyWTF ( http://www.thedailywtf.com/ ) of the form "Some people get 10 years experience, some people get 10 * one years experience." There's also a quote "Wisdom comes with age but sometimes age comes alone."


Years of experience gainfully employed in a particular industry can be a useful metric but merely years of experience of a particular technology isn't that illuminating. I have 8 full years of experience playing guitar. Can I really play guitar? Nope.


I would say that framework != language. C++ is an example where I would trust someone who has been using it for more than a few years more than I would trust someone who has used it for 6 month (depending on their other experience).

Using C++ under something like MS Visual Studio is closer to the 6 month test...


"Years of experience" is a lousy metric. It only gets used as much as it does because people (especially recruiters and managers) have no better way to distinguish good programmers from bad. It's not only possible but common for a programmer to do something for years, badly.

Would you rather have a good programmer with little experience on a given platform or a mediocre programmer with lots?


Agreed.

In addition to having demonstrable ability (thru side-projects for example), I'd also want a candidate to explain a technical detail or concept to me. I'd phrase it like ("Explain an interesting technical concept to me - your choice"). This would show me a) what they think is interesting b) how well they understand it c) how good they are at sharing/teaching.


What I've always wanted to do before hiring someone is work with them for a while. It usually doesn't take too long to get a sense of how good somebody is and how easy they will be to work with.

How good a new team member will be on a project is a complex question. The only really good indicator of the outcome is the outcome itself. Giving a good technical interview is not that great a predictor (although better than years-of-experience). What I like about this strategy is that it eschews gimmicks and gets as close to the real situation as possible.


Maybe have a list of technologies you've used for 6+ months as a checklist and then a list of completed projects. Certainly a traditional resume isn't a good fit for programmers.


Then again, it takes some people 10 years: http://norvig.com/21-days.html




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: