Re: Some belated patch review for "Buffers" explain analyze patch

On Tue, Feb 9, 2010 at 8:18 PM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>> Perhaps instead of looking to change the "actual" times we should look
>> at a way to include total time spent in each node.
>
> You can already get that by multiplying the displayed total time by the
> number of loops.  Robert does have a point that this is subject to a lot
> of roundoff error, though, when the per-loop time is much less than 1
> msec.  I wouldn't object to adding a "total time" field to the
> machine-readable formats.

One possibility we discussed previously is to add some decimal places
to the relevant values when nloops > 1.

If we're going to add a total time field, I think we should add it to
both the machine-readable and human-readable formats. I know it's a
little long-winded, but one of the things that I find really
unfortunate about the current format is that it's sometimes hard to
look at a plan tree and figure out where "the slow part" is, because
some things have been divided through by the number of loops. Reading
through the JSON or YAML format to find the data is, I guess, better
than nothing, but only slightly: I frequently deal with plans that are
25-30 lines long: in XML format, those will be 250-300 lines long. I
wouldn't mind having to do EXPLAIN (ANALYZE, VERBOSE) or EXPLAIN
(ANALYZE, some-other-option) to get the details, but EXPLAIN (ANALYZE,
FORMAT XML) ... is not really a direction I want to go.

...Robert