On Dec 4, 2012, at 7:45 PM, Chris Cunningham <[email protected]> wrote:

> On Tue, Dec 4, 2012 at 6:30 AM, Henrik Sperre Johansen 
> <[email protected]> wrote:
> On 04.12.2012 12:49, Stéphane Ducasse wrote:
> Comments are ***REEEAALLLLY*** important because they help a guy reading the 
> code to say yes you are smart and you got it right
> I will explain you why.
> 
> Even writing comment in test is cool.
> 
> Stef
> I would go so far as to say comments in tests are *especially* important.
> It's sometimes hard to tell whether a failure is due to a faulty test, or an 
> actual bug, when you have no idea (even when looking at the code) what case 
> the test was actually written to cover.
> With a comment of what the test was intended to do, you have a better feel 
> for how to handle it.
> True, to a point.  As long as later coders that alter the 
> method/class/package either adhear to the original comments, or update the 
> comments when they change the behaviour/intent of the code, or at the very 
> least, delete the comment when they make significant changes that would 
> invalidate the comment, this is true.  If, on the other hand, the comments 
> are left as is while potentially drastic changes are made, then the comments 
> become seriously misleading.
> 
> Not to say comments are bad - but we must all work to make sure the comments 
> that exist stay relevant.
> 

And of course it is important to actually change comments when someone finds a 
problem.
"How can you change the world if you can't even change a typo in a comment?"

        Marcus


Reply via email to