Re: Any exceptions for the 15-line rule?

[Dmitry Gutov]

On 01.05.2013 22:06, Glenn Morris wrote:
> Dmitry Gutov wrote:
>
> > Thank you. I don't think I've seen a good definition of "tiny change"
> > anywhere, and there's no s/foo/bar in either of these patches, but if
> > you suggest to use my own judgment, that's fine by me.
>
> s/foo/bar was just an example.
>
> You need to look at the change and ask, "conceptually, how big is this"?
> Deleting says 15 lines of common code from two places and turning it
> into a separate function that those two places call only "counts" as a
> few lines IMO (the function definition part, the function body is the
> same code as before, just moved around).

I see, that's good.
The two function invocations should probably also count, as +1 or +2.

What about the second patch?

https://github.com/mooz/js2-mode/pull/100.diff

Should it be counted as just +3 (for the added docstring), or does the 
tweaked list of identifiers also contribute? I'm thinking no, because 
the change only removes elements from the list.

What if the diff was adding new elements (like the reverse of this diff 
would do)? Would I be able to discount those lines then, on the basis 
that this is just copying information from the specification?