Re: alter_table regression test problem

On 2013-11-07 09:55:55 -0500, Tom Lane wrote:
> Kevin Grittner <kgrittn(at)ymail(dot)com> writes:
> > I think it is pretty much SOP for committers to prefer a patch that
> > adds a new pair of braces around 50 lines of code and only changes
> > non-whitespace of a few lines within that block to leave the block
> > at its old indentation.
>
> Yes. It's much easier to review a patch done that way than to wonder if
> the apparently-just-whitespace changes are masking something substantive.
> In fact, if I'm reviewing a patch that reindents a big chunk of existing
> code, I'll frequently not use the patch but reconstruct it that way,
> just to be sure.

But why not just use git diff/show/whatever -w or, as suggested by
Kevin, --color-words?

ISTM the patch author is much more likely to mistake when indenting code
strangely.

Greetings,

Andres Freund

--
Andres Freund http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services