Re: Replication vs. float timestamps is a disaster

On 20/02/17 08:03, Andres Freund wrote:
> On 2017-02-19 10:49:29 -0500, Tom Lane wrote:
>> Robert Haas <robertmhaas@gmail.com> writes:
>>> On Sun, Feb 19, 2017 at 3:31 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>>>> Thoughts?  Should we double down on trying to make this work according
>>>> to the "all integer timestamps" protocol specs, or cut our losses and
>>>> change the specs?
>>
>>> I vote for doubling down.  It's bad enough that we have so many
>>> internal details that depend on this setting; letting that cascade
>>> into the wire protocol seems like it's just letting the chaos spread
>>> farther and wider.
>>
>> How do you figure that it's not embedded in the wire protocol already?
>> Not only the replicated data for a timestamp column, but also the
>> client-visible binary I/O format, depend on this.  I think having some
>> parts of the protocol use a different timestamp format than other parts
>> is simply weird, and as this exercise has shown, it's bug-prone as all
>> get out.
> 
> I don't think it's that closely tied together atm. Things like
> pg_basebackup, pg_receivexlog etc should work, without having to match
> timestamp storage.  Logical replication, unless your output plugin dumps
> data in binary / "raw" output, also works just fine across the timestamp
> divide.
> 
> It doesn't sound that hard to add a SystemToIntTimestamp() function,
> given it only needs to do something if float timestamps are enabled?
> 

It's definitely not hard, we already have
IntegerTimestampToTimestampTz() which does the opposite conversion anyway.

That being said, I did wonder myself if we should just deprecate float
timestamps as well.

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