Re: Run pgindent now? - Mailing list pgsql-hackers

From Tom Lane
Subject Re: Run pgindent now?
Date
Msg-id 26328.1432478650@sss.pgh.pa.us
Whole thread Raw
In response to Re: Run pgindent now?  (Andrew Dunstan <andrew@dunslane.net>)
Responses Re: Run pgindent now?  (Bruce Momjian <bruce@momjian.us>)
List pgsql-hackers
Andrew Dunstan <andrew@dunslane.net> writes:
> On 05/23/2015 11:37 PM, Tom Lane wrote:
>> No, pgindent has *always* been wonky about lines that contain a typedef
>> name but are not variable declarations.

> Well, that sounds like something we should try to patch, doesn't it? 
> (No, I'm not volunteering.)

In the past, the main argument against changing pgindent has been that
it would cause reformatting of settled code.  For example, when Bruce
twiddled its right margin limit for comments around 8.1, that caused
literally years worth of back-patching pain.  If we can get to an
agreement on re-indenting back branches at the same time as master,
then this problem would go away, and I for one would get a lot more
enthusiastic about trying to improve pgindent rather than just working
around its oddities.

Not that I'm volunteering either right now.
        regards, tom lane



pgsql-hackers by date:

Previous
From: Andrew Dunstan
Date:
Subject: Re: Run pgindent now?
Next
From: Tom Lane
Date:
Subject: Re: fsync-pgdata-on-recovery tries to write to more files than previously