Comments on: Code Reviews http://tleaves.com/2005/03/16/code-reviews/ Creativity x Technology Sat, 17 Mar 2012 05:09:58 +0000 hourly 1 http://wordpress.org/?v=3.3.1 By: Strider http://tleaves.com/2005/03/16/code-reviews/comment-page-1/#comment-1103 Strider Thu, 17 Mar 2005 16:04:34 +0000 http://tleaves.com/?p=330#comment-1103 This reminded me a bit of an excellent Den Beste article. Thought you might be interested: It's OK to be wrong (http://denbeste.nu/cd_log_entries/2002/10/ItsOKtobewrong.shtml) This reminded me a bit of an excellent Den Beste article. Thought you might be interested:

It’s OK to be wrong (http://denbeste.nu/cd_log_entries/2002/10/ItsOKtobewrong.shtml)

]]> By: peterb http://tleaves.com/2005/03/16/code-reviews/comment-page-1/#comment-1102 peterb Thu, 17 Mar 2005 05:06:20 +0000 http://tleaves.com/?p=330#comment-1102 That's fair. I think the point I was really trying to get at didn't quite come across. I wasn't trying to slam open source as much as I was trying to say that a code review with 15 people in a room is less effective than a code review with 3 people in a room, because with 15 people it's easier for each individual reviewer to simply not contribute (and, if they all have jobs, they've even more incentive to not talk because they want to get out of the room in less than 4 hours, which won't happen if they are all actively participating). Which brings up another point I forgot to mention: if you're doing code reviews in person, set a time limit. Personally, I can't review code in a meeting context for more than about an hour without wanting a nap. (For some reason, if I'm reviewing someone else's code "on my own" for commenting via email, I can concentrate for longer periods of time). That’s fair. I think the point I was really trying to get at didn’t quite come across. I wasn’t trying to slam open source as much as I was trying to say that a code review with 15 people in a room is less effective than a code review with 3 people in a room, because with 15 people it’s easier for each individual reviewer to simply not contribute (and, if they all have jobs, they’ve even more incentive to not talk because they want to get out of the room in less than 4 hours, which won’t happen if they are all actively participating).

Which brings up another point I forgot to mention: if you’re doing code reviews in person, set a time limit. Personally, I can’t review code in a meeting context for more than about an hour without wanting a nap. (For some reason, if I’m reviewing someone else’s code “on my own” for commenting via email, I can concentrate for longer periods of time).

]]> By: Brian Hook http://tleaves.com/2005/03/16/code-reviews/comment-page-1/#comment-1101 Brian Hook Thu, 17 Mar 2005 04:28:49 +0000 http://tleaves.com/?p=330#comment-1101 I believe that's a misapplication of the diffusion of responsibility phenomenon. With a typical DoR you have a large group of witnesses or bystanders with no vested interest in a particular situation who are implicitly expected to act on that situation, often with no personal gain for themselves. Contrast this with a typical open source situation -- contributors have many, many reasons to invest effort and time in finding bugs, providing optimizations, porting to new targets, etc. Sometimes it's economic -- I need MySQL to run on my old VAX/8600!!! -- and many times it's just flat out ego -- I was the first person to port Linux to my toaster!!!! That's very different than driving by a crime in action and not doing anything about it. It comes down to return on investment of effort. I believe that’s a misapplication of the diffusion of responsibility phenomenon. With a typical DoR you have a large group of witnesses or bystanders with no vested interest in a particular situation who are implicitly expected to act on that situation, often with no personal gain for themselves.

Contrast this with a typical open source situation — contributors have many, many reasons to invest effort and time in finding bugs, providing optimizations, porting to new targets, etc. Sometimes it’s economic — I need MySQL to run on my old VAX/8600!!! — and many times it’s just flat out ego — I was the first person to port Linux to my toaster!!!!

That’s very different than driving by a crime in action and not doing anything about it. It comes down to return on investment of effort.

]]>