Thread

Commits

  1. Tweak wording of documentation for pg_checksums

  2. Add options to enable and disable checksums in pg_checksums

  1. clean up pg_checksums.sgml

    Justin Pryzby <pryzby@telsasoft.com> — 2019-03-29T14:32:10Z

    PFA patch with minor improvements to documentation.
    
    Also, what do you think about changing user-facing language from
    "check checksum" to "verify checksum" ?  I see that commit ed308d78 actually
    moved in the other direction, but I preferred "verify".
    
  2. Re: clean up pg_checksums.sgml

    Michael Paquier <michael@paquier.xyz> — 2019-03-30T01:51:23Z

    On Fri, Mar 29, 2019 at 09:32:10AM -0500, Justin Pryzby wrote:
    > PFA patch with minor improvements to documentation.
    
    Patch does not apply, and I have reworded the last paragraph about
    failures while operating.
    
    > Also, what do you think about changing user-facing language from
    > "check checksum" to "verify checksum" ?  I see that commit ed308d78
    > actually moved in the other direction, but I preferred "verify".
    
    Yes, that's a debate that we had during the discussion for the new
    switches, and we have decided to use --check over --verify for the
    default option.  On the one hand, "Check checksums" is rather
    redundant, but that's more consistent with the option name.  "Verify
    checksums" is perhaps more elegant.  My opinion is that having some
    consistency between the option names and the docs is nicer.
    --
    Michael
    
  3. Re: clean up pg_checksums.sgml

    Justin Pryzby <pryzby@telsasoft.com> — 2019-04-08T00:15:46Z

    On Sat, Mar 30, 2019 at 10:51:23AM +0900, Michael Paquier wrote:
    > On Fri, Mar 29, 2019 at 09:32:10AM -0500, Justin Pryzby wrote:
    > > PFA patch with minor improvements to documentation.
    > 
    > Patch does not apply, and I have reworded the last paragraph about
    > failures while operating.
    
    Sorry, the patch was on top of an brief effort I made to rename "check
    checksums" to "verify checksums", before asking about the idea.
    
    PFA patch to master.
    
    Justin
    
    > > Also, what do you think about changing user-facing language from
    > > "check checksum" to "verify checksum" ?  I see that commit ed308d78
    > > actually moved in the other direction, but I preferred "verify".
    > 
    > Yes, that's a debate that we had during the discussion for the new
    > switches, and we have decided to use --check over --verify for the
    > default option.  On the one hand, "Check checksums" is rather
    > redundant, but that's more consistent with the option name.  "Verify
    > checksums" is perhaps more elegant.  My opinion is that having some
    > consistency between the option names and the docs is nicer.
    
  4. Re: clean up pg_checksums.sgml

    Michael Paquier <michael@paquier.xyz> — 2019-04-08T06:37:03Z

    On Sun, Apr 07, 2019 at 07:15:46PM -0500, Justin Pryzby wrote:
    > Sorry, the patch was on top of an brief effort I made to rename "check
    > checksums" to "verify checksums", before asking about the idea.
    > 
    > PFA patch to master.
    
    Thanks for the patch, Justin.  That looks indeed clearer after
    considering your proposal, so I have applied most of it.  There were
    some terms I found fuzzy though.  For example, I have replaced
    "checksum state" by "data checksum configuration", but kept
    "verifying" because "check checksums" sounds kind of redundant.
    --
    Michael