missing documentation detail?
The following documentation comment has been logged on the website: Page: https://www.postgresql.org/docs/18/app-pgverifybackup.html Description: I was disappointed not to find the return code information on the pg_verifybackup pages - I am using the utility in a script and it seems I cannot distinguish between an invalid or 'bad' backup and a failed run (due to e.g. bad path, missing manifest file) - it *seems* that either will simply return as 1 but I am not certain. I would like to differentiate between these states.
md5 reference
The following documentation comment has been logged on the website: Page: https://www.postgresql.org/docs/18/auth-pg-hba-conf.html Description: Hi, In the last pg_hba example on the page md5 is used as method. Maybe it should be changed to scram-sha-256 as that is the preferred method? KR Mikael Gustavsson
Re: md5 reference
On 17.12.25 08:21, PG Doc comments form wrote: The following documentation comment has been logged on the website: Page: https://www.postgresql.org/docs/18/auth-pg-hba-conf.html Description: Hi, In the last pg_hba example on the page md5 is used as method. Maybe it should be changed to scram-sha-256 as that is the preferred method? Yes, I think that would make sense to modernize that example.
Re: md5 reference
> On 17 Dec 2025, at 13:20, Peter Eisentraut wrote: > > On 17.12.25 08:21, PG Doc comments form wrote: >> The following documentation comment has been logged on the website: >> Page: https://www.postgresql.org/docs/18/auth-pg-hba-conf.html >> Description: >> Hi, >> In the last pg_hba example on the page md5 is used as method. Maybe it >> should be changed to scram-sha-256 as that is the preferred method? > > Yes, I think that would make sense to modernize that example. Agreed, I was just looking at it after reading the report. @Mikael: would you like to try your hand at proposing a patch for this? -- Daniel Gustafsson
SV: md5 reference
Från: Daniel Gustafsson Skickat: den 17 december 2025 13:33 Till: Peter Eisentraut Kopia: Mikael Gustavsson; [email protected] Ämne: Re: md5 reference > On 17 Dec 2025, at 13:20, Peter Eisentraut wrote: > > On 17.12.25 08:21, PG Doc comments form wrote: >> The following documentation comment has been logged on the website: >> Page: https://www.postgresql.org/docs/18/auth-pg-hba-conf.html >> Description: >> Hi, >> In the last pg_hba example on the page md5 is used as method. Maybe it >> should be changed to scram-sha-256 as that is the preferred method? > > Yes, I think that would make sense to modernize that example. Agreed, I was just looking at it after reading the report. @Mikael: would you like to try your hand at proposing a patch for this? -- Daniel Gustafsson Absolutely, i can try, but I think I might need some guidance. /Mikael
Re: missing documentation detail?
On Tuesday, December 16, 2025, PG Doc comments form wrote: > The following documentation comment has been logged on the website: > > Page: https://www.postgresql.org/docs/18/app-pgverifybackup.html > Description: > > I was disappointed not to find the return code information on the > pg_verifybackup pages - I am using the utility in a script and it seems I > cannot distinguish between an invalid or 'bad' backup and a failed run (due > to e.g. bad path, missing manifest file) - it *seems* that either will > simply return as 1 but I am not certain. I would like to differentiate > between these states. > This isn’t a good place to request feature enhancements. There isn’t an issue with the documentation - the program doesn’t provide multiple conditional exit codes. Post a patch to -hackers or a feature request to -general (mailing list). David J.
Re: md5 reference
> On 17 Dec 2025, at 14:16, Mikael Gustavsson wrote: > Absolutely, i can try, but I think I might need some guidance. Great! Feel free to ask anything either here, on the PostgreSQL Hacking Discord or in an off-list email, whichever you prefer. The file in the source tree you want to be looking at is doc/src/sgml/client-auth.sgml. -- Daniel Gustafsson
Re: missing documentation detail?
On 17 Dec 2025, at 14:20, David G. Johnston wrote:This isn’t a good place to request feature enhancements. There isn’t an issue with the documentation - the program doesn’t provide multiple conditional exit codes.Post a patch to -hackers or a feature request to -general (mailing list).Absolutely right, that's the best way to propose new (or changes to) features.The proposed change here seems pretty reasonable and might be as simple as the(untested) attached, modeled around pg_amcheck. I noticed that pg_amcheckdoesn't document its different return values either which seems like anoversight. --Daniel Gustafsson pg_verifybackup_exit.diff Description: Binary data
