Re: Advocacy of code reviews: how the heck do you do it?

I work in an academic department, too, and have also seen the lack of discipline that you mention.

I have had some luck encouraging code review using two methods. First, I ask some folks to look at my code and comment. Often they will do so, and in turn feel more comfortable in asking me to do so. It helps to first put your ego on the line before asking others to do the same.

The second approach is to note that part of the scientific method is reproducibility, so programs should be made publically available along with the papers. Then code review becomes just another part of the peer review process and prevents public shame in distributing buggy, sloppy code.

This has only worked with three people in my dept., but we have a nucleus of code quality that at times really helps my own work.

-Mark

Comment on Re: Advocacy of code reviews: how the heck do you do it?

Replies are listed 'Best First'.
Re^2: Advocacy of code reviews: how the heck do you do it? by FoxtrotUniform (Prior) on Oct 17, 2004 at 20:43 UTC
I have had some luck encouraging code review using two methods. First, I ask some folks to look at my code and comment. Often they will do so, and in turn feel more comfortable in asking me to do so. It helps to first put your ego on the line before asking others to do the same. I should have mentioned that I've tried this approach; the standard response was "I'd love to, but I don't have time". Since the sticking point seems to be time rather than ego, I thought I'd put my time on the line instead. I like the point about making code available with papers; my supervisor has been talking about this a lot lately (basically, his theory is that more people will cite your paper if they can easily compare their method to yours on the same platform; providing code makes that a lot more likely). `--` Yours in pedantry,`F o x t r o t U n i f o r m` "Lines of code don't matter as long as I'm not writing them." -- merlyn	[reply]

Replies are listed 'Best First'.

Re^2: Advocacy of code reviews: how the heck do you do it?
by FoxtrotUniform (Prior) on Oct 17, 2004 at 20:43 UTC

I have had some luck encouraging code review using two methods. First, I ask some folks to look at my code and comment. Often they will do so, and in turn feel more comfortable in asking me to do so. It helps to first put your ego on the line before asking others to do the same.

I should have mentioned that I've tried this approach; the standard response was "I'd love to, but I don't have time". Since the sticking point seems to be time rather than ego, I thought I'd put my time on the line instead.

I like the point about making code available with papers; my supervisor has been talking about this a lot lately (basically, his theory is that more people will cite your paper if they can easily compare their method to yours on the same platform; providing code makes that a lot more likely).

--
Yours in pedantry,F o x t r o t U n i f o r m
"Lines of code don't matter as long as I'm not writing them." -- merlyn

[reply]