On 2026-Sep-11, Chao Li wrote:

> > On Sep 11, 2026, at 08:54, Fujii Masao <[email protected]> wrote:

> > -  long_desc => '-1 disables the timeout and waits for the receiver to
> > catch up; 0 does not wait for the receiver to catch up.',
> > +  long_desc => '-1 disables the timeout. 0 means do not wait for the
> > receiver to catch up.',

> 
> Yeah, here -1 is really more like “wait forever”, so just saying “disables 
> the timeout” feels like it loses that distinction.

Maybe it should say "0 means do not wait, -1 means wait indefinitely."
This complements the short_desc,

  Sets the maximum time the server waits during shutdown for all WAL data
  to be replicated to the receiver.

I think the words "for the receiver to catch up" is somehow at odds with
the short_desc's "for all WAL data to be replicated".  They should be
saying the same thing, but aren't.

-- 
Álvaro Herrera         PostgreSQL Developer  —  https://www.EnterpriseDB.com/
Si no sabes adonde vas, es muy probable que acabes en otra parte.


Reply via email to