Re: [Fwd: Re: proposal: new long psql parameter --on-error-stop] - Mailing list pgsql-hackers

On Mon, Jun 23, 2014 at 6:06 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:
>
>
>
> 2014-06-23 10:57 GMT+02:00 Fujii Masao <masao.fujii@gmail.com>:
>
>> On Mon, Jun 23, 2014 at 5:10 PM, Pavel Stehule <pavel.stehule@gmail.com>
>> wrote:
>> > Hello
>> >
>> >
>> > 2014-06-23 10:02 GMT+02:00 Fujii Masao <masao.fujii@gmail.com>:
>> >
>> >> On Mon, Jun 23, 2014 at 12:04 AM, Pavel Stehule
>> >> <pavel.stehule@gmail.com>
>> >> wrote:
>> >> > Hello
>> >> >
>> >> > third version with Erik's update
>> >>
>> >> Here are some my comments:
>> >>
>> >> The document of psql needs to be updated. At least the description of
>> >> new
>> >> option
>> >> this patch adds needs to be added into the document.
>> >>
>> >> +    printf(_("  --help-variables         list of available
>> >> configuration variables (options), then exit\n"));
>> >>
>> >> We should get rid of "of" from the message, or add "show" in front of
>> >> "list of"?
>> >>
>> >> +    printf(_("  ECHO               write all input lines to standard
>> >> output\n"));
>> >>
>> >> This message seems not to be correct when ECHO=queries is set.
>> >>
>> >> +    printf(_("  COMP_KEYWORD_CASE  determines which letter case to
>> >> use when completing an SQL key word\n"));
>> >> +    printf(_("  DBNAME             name of currently connected
>> >> database\n"));
>> >> +    printf(_("  ECHO               write all input lines to standard
>> >> output\n"));
>> >>
>> >> I found that some help message line uses a normal form of a verb, but
>> >> other does not.
>> >> We should standardize them?
>> >>
>> >> +    printf(_("  PROMPT1, PROMPT2, PROMPT3  specify the psql
>> >> prompt\n"));
>> >>
>> >> When the option name field is long, we should add a new line just
>> >> after the name field
>> >> and align the starting position of the option explanation field. That
>> >> is,
>> >> for example, the above should be
>> >>
>> >> printf(_("  PROMPT1, PROMPT2, PROMPT3\n"
>> >>              "                     specify the psql prompt\n"));
>> >>
>> >> +    printf(_("  ON_ERROR_ROLLBACK  when on, ROLLBACK on error\n"));
>> >>
>> >> This message seems incorrect to me. When this option is on and an error
>> >> occurs
>> >> in transaction, transaction continues rather than ROLLBACK occurs,
>> >> IIUC.
>> >> I did not check whole help messages yet, but ISTM some messages are not
>> >> correct.
>> >> It's better to check them again.
>> >>
>> >> +    printf(_("  PSQL_RC            alternative location of the user's
>> >> .psqlrc file\n"));
>> >>
>> >> Typo: PSQL_RC should be PSQLRC
>> >>
>> >> +    printf(_("  PGDATABASE         same as the dbname connection
>> >> parameter\n"));
>> >> +    printf(_("  PGHOST             same as the host connection
>> >> parameter\n"));
>> >> +        printf(_("  PGPORT             same as the port connection
>> >> parameter\n"));
>> >> +    printf(_("  PGUSER             same as the user connection
>> >> parameter\n"));
>> >> +    printf(_("  PGPASSWORD         possibility to set password (not
>> >> recommended)\n"));
>> >> +    printf(_("  PGPASSFILE         password file (default
>> >> ~/.pgpass)\n"));
>> >>
>> >> I don't think that psql needs to display the help messages of even
>> >> environment
>> >> variables supported by libpq.
>> >>
>> >
>> > Main reason is a PGPASSWORD -- it is probably most used env variable
>> > with
>> > psql
>> >
>> > PGPASSWORD=****** psql is very often used pattern
>>
>> But it's not recommended as the help message which the patch added says ;)
>
>
> why?
>
> who can see this value?

I'm sure that the document explains this.

http://www.postgresql.org/docs/devel/static/libpq-envars.html
---------------------------------------
PGPASSWORD behaves the same as the password connection parameter.
Use of this environment variable is not recommended for security reasons,
as some operating systems allow non-root users to see process environment
variables via ps; instead consider using the ~/.pgpass file
---------------------------------------

Regards,

-- 
Fujii Masao