Re: Make mesage at end-of-recovery less scary.

At Tue, 1 Feb 2022 12:38:01 +0400, Pavel Borisov <pashkin(dot)elfe(at)gmail(dot)com> wrote in
> Maybe it can be written little bit shorter:
> pe = (char *) record + XLOG_BLCKSZ - (RecPtr & (XLOG_BLCKSZ - 1));
> as
> pe = p + XLOG_BLCKSZ - (RecPtr & (XLOG_BLCKSZ - 1));
> ?

That difference would be a matter of taste, but I found it looks
cleaner that definition and assignment is separated for both p and pe.
Now it is like the following.

> char *p;
> char *pe;
>
> /* scan from the beginning of the record to the end of block */
> p = (char *) record;
> pe = p + XLOG_BLCKSZ - (RecPtr & (XLOG_BLCKSZ - 1));

> The problem that pgindent sometimes reflow formatting of unrelated blocks
> is indeed existing. But I think it's right to manually leave pgindent-ed
> code only on what is related to the patch. The leftover is pgindent-ed in a
> scheduled manner sometimes, so don't need to bother.

Yeah, I meant that it is a bit annoying to unpginden-ting unrelated
edits:p

> I'd like to set v10 as RfC.

Thanks! The suggested change is done in the attached v11.

regards.

--
Kyotaro Horiguchi
NTT Open Source Software Center