Re: [HACKERS] IF NOT EXISTS option for CREATE SERVER and CREATE USERMAPPING statements - Mailing list pgsql-hackers

From Arthur Zakirov
Subject Re: [HACKERS] IF NOT EXISTS option for CREATE SERVER and CREATE USERMAPPING statements
Date
Msg-id CAKNkYnxmuH+kVqBZOaK8VGSE9=tXc6WcSokAmr_eLw-59jGbbw@mail.gmail.com
Whole thread Raw
In response to Re: [HACKERS] IF NOT EXISTS option for CREATE SERVER and CREATE USERMAPPING statements  (Ashutosh Bapat <ashutosh.bapat@enterprisedb.com>)
List pgsql-hackers
2017-03-14 15:55 GMT+03:00 Ashutosh Bapat <ashutosh.bapat@enterprisedb.com>:
>
> I noticed that the earlier error message was using "server" instead of
> "foreign server", while the new message uses the later one. Usually,
> when converting an error to notice, we don't expect such changes. But
> many other error messages are using "foreign server" instead of
> "server", so probably this one needed a change anyway. But then, the
> command to create a foreign server is not "CREATE FOREIGN SERVER",
> it's "CREATE SERVER", so users are already getting confused?

Actually, there are other messages with "server". For example, in the
AlterForeignServerOwner() or in the postgres_fdw code.
Maybe it is better to not change "server" to "foreign server" in
"create_foreign_server_if_not_exists.patch"? I think it will be better
to fix all such messages with a separate patch, If we will decide that
it is necessary to change "server" messages.

>
> I don't see similar change in the error message for the user mapping.
> Do we need to change "server" to "foreign server" in case of user
> mapping?  The doc changes didn't compile with error
> "osx:ref/create_user_mapping.sgml:52:15:E: document type does not
> allow element "VARLISTENTRY" here; assuming missing "VARIABLELIST"

Indeed! Missed that.


-- 
Arthur Zakirov
Postgres Professional: http://www.postgrespro.com
Russian Postgres Company



pgsql-hackers by date:

Previous
From: Robert Haas
Date:
Subject: Re: [HACKERS] GSOC - TOAST'ing in slices
Next
From: Sachin Kotwal
Date:
Subject: Re: [HACKERS] Allow pg_dumpall to work without pg_authid