Re: alter_table regression test problem

On Thu, Nov 7, 2013 at 12:18 PM, Andres Freund <andres(at)2ndquadrant(dot)com> wrote:
> On 2013-11-07 10:10:34 -0500, Tom Lane wrote:
>> Andres Freund <andres(at)2ndquadrant(dot)com> writes:
>> > On 2013-11-07 06:49:58 -0800, Kevin Grittner wrote:
>> >> It's up to the committer whether to indent
>> >> after review and make both substantive and whitespace changes
>> >> together, but I have definitely seen those done separately, or even
>> >> left for the next global pgindent run to fix.
>>
>> > Hm. I don't remember many such cases and I've just looked across the git
>> > history and i didn't really find anything after
>> > a04161f2eab55f72b3c3dba9baed0ec09e7f633f, but that was back when
>> > individuals couldn't run pgindent because of the typedefs file.
>>
>> FWIW, I tend to pgindent before committing, now that it's easy to do so.
>> But in cases like this, I'd much rather review the patch *before* that
>> happens. Basically the point of the review is to follow and confirm
>> the patch author's thought process, and I'll bet you put the braces in
>> before reindenting too.
>
> Well, I did both at the same time, I have an emacs command for that
> ;). But independent from that technicality, reindenting is part of
> changing the flow of logic for me - I've spent a couple of years
> primarily writing python (where indentation is significant), maybe it's
> that.
>
> So, here's the patch (slightly editorialized) with reverted indenting of
> that block.

Gah. Well, of course, I have the opposite preference from Tom on how
this should be indented. Sigh.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company