Thread: Re: Catastrophic changes to PostgreSQL 8.4
[moving this point to -hackers] On Wed, Dec 2, 2009 at 10:49 PM, Adrian Klaver <aklaver@comcast.net> wrote: > When I tried the above Postgres did not ignore the command, instead it gave me > the following error and did not create the database: > CREATE DATABASE bacula ENCODING 'SQL_ASCII'; > ERROR: new encoding (SQL_ASCII) is incompatible with the encoding of the > template database (UTF8) > HINT: Use the same encoding as in the template database, or use template0 as > template. Actually I'm kind of surprised at this. I don't see a reason not to allow converting a template to SQL_ASCII from any encoding given that we're going to allow them to put random bytes into the database afterwards. Why not let them start with random bytes? -- greg
Greg Stark <gsstark@mit.edu> writes: >> CREATE DATABASE bacula ENCODING 'SQL_ASCII'; >> ERROR: �new encoding (SQL_ASCII) is incompatible with the encoding of the >> template database (UTF8) >> HINT: �Use the same encoding as in the template database, or use template0 as >> template. > Actually I'm kind of surprised at this. I don't see a reason not to > allow converting a template to SQL_ASCII from any encoding given that > we're going to allow them to put random bytes into the database > afterwards. Why not let them start with random bytes? Hm, that's a good point, although we should still enforce that the new database's LC_xxx settings be C/POSIX if it's SQL_ASCII. My guess is that the initdb environment had some non-C locale, so this example would have failed the next error check anyway. regards, tom lane
> On Wed, Dec 2, 2009 at 10:49 PM, Adrian Klaver <aklaver@comcast.net> wrote: >> When I tried the above Postgres did not ignore the command, instead it gave me >> the following error and did not create the database: >> CREATE DATABASE bacula ENCODING 'SQL_ASCII'; >> ERROR: new encoding (SQL_ASCII) is incompatible with the encoding of the >> template database (UTF8) >> HINT: Use the same encoding as in the template database, or use template0 as Boy, this looks like a reason why app developers should do some beta testing of new Postgres versions ... --Josh Berkus