| From: | Amul Sul <sulamul(at)gmail(dot)com> |
|---|---|
| To: | Alvaro Herrera <alvherre(at)alvh(dot)no-ip(dot)org> |
| Cc: | PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: GetStandbyFlushRecPtr() : OUT param is not optional like elsewhere. |
| Date: | 2022-07-21 04:08:17 |
| Message-ID: | CAAJ_b97+2RYTFJFnaL3tAQFEM8fftfNXpL8fakmiK7CXthgyGQ@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Thanks Aleksander and Álvaro for your inputs.
I understand this change is not making any improvement to the current
code. I was a bit concerned regarding the design and consistency of
the function that exists for the server in recovery and for the server
that is not in recovery. I was trying to write following snippet
where I am interested only for XLogRecPtr:
recPtr = RecoveryInProgress() ? GetStandbyFlushRecPtr(NULL) :
GetFlushRecPtr(NULL);
But I can't write this since I have to pass an argument for
GetStandbyFlushRecPtr() but that is not compulsory for
GetFlushRecPtr().
I agree to reject proposed changes since that is not useful
immediately and have no rule for the optional argument for the
similar-looking function.
Regards,
Amul
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Kyotaro Horiguchi | 2022-07-21 04:20:12 | Re: Use "WAL segment" instead of "log segment" consistently in user-facing messages |
| Previous Message | John Naylor | 2022-07-21 03:44:26 | Re: Custom tuplesorts for extensions |