On Wed, Jul 19, 2017 at 12:39:07PM -0400, Tom Lane wrote:
> Justin Pryzby <pryzby@telsasoft.com> writes:
> > FYI, I happened across this commit comment:
> > 3f902354b08ac788600f0ae54fcbfc1d4e3ea765
> > | So, let's accept the removal of the guarantee about
> > | the output tuple's rowtype marking, recognizing that this is a API change
> > | that could conceivably break third-party callers of tupconvert.c. (So,
> > | let's remember to mention it in the v10 release notes.)
>
> > ..but couldn't see that the commit or change is so referenced.
>
> Yeah, I see nothing about 3f902354b in release-10.sgml either.
> We've had varying policies over the years about whether to mention
> internal API changes in the release notes or not, but this one
> I think does belong there, since it's a silent API break rather
> than one that would easily be caught due to compiler errors.
> Bruce, did you have any specific reasoning for leaving it out?
I doubt I saw that sentence in the paragraph. For long text like that,
I am usually looking for "BACKWARDS INCOMPATIBLE CHANGE" or something
like that. Sorry I missed it.
-- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB
http://enterprisedb.com
+ As you are, so once was I. As I am, so you will be. +
+ Ancient Roman grave inscription +