Re: BUG #12670: Can't create postgis extension in single-user mode
От | Andrew Gierth |
---|---|
Тема | Re: BUG #12670: Can't create postgis extension in single-user mode |
Дата | |
Msg-id | 87zj93agba.fsf@news-spur.riddles.org.uk обсуждение исходный текст |
Ответ на | Re: BUG #12670: Can't create postgis extension in single-user mode (Tom Lane <tgl@sss.pgh.pa.us>) |
Ответы |
Re: BUG #12670: Can't create postgis extension in single-user mode
Re: BUG #12670: Can't create postgis extension in single-user mode |
Список | pgsql-bugs |
>>>>> "Tom" == Tom Lane <tgl@sss.pgh.pa.us> writes: >> [ Can't create postgis extension in single-user mode ] Tom> Why in the world would you think that's a good thing to do? The context, for what it's worth, is setting up docker containers, so there's a reasonable desire to be able to do some basic database setup before starting any listener. (Not my project; my only involvement in this is in answering questions on irc, as usual) Tom> Single-user mode is a barely-documented disaster recovery aid. Tom> It's not meant for routine activity. There are a *whole lot* Tom> of behaviors you want that are turned off in single-user mode. Ok, I'll bite: what are they? (I did a quick review of uses of IsUnderPostmaster. Essentially none of them are visible at SQL level except this one (the only other one that comes close is the disabling of event triggers). Almost everything else is about process management; the only exceptions I found are one for accepting tsearch dictionary options and another for running as superuser in spite of missing role entries.) Tom> The specific reason why this doesn't work is this bit in Tom> heap_create_with_catalog: Which personally I think is misguided. There never was any good reason other than rowcount of pg_type not to make array types for catalog tables, and it's inconsistent not to be able to use arrays of system views or i_s views and so on. -- Andrew (irc:RhodiumToad)
В списке pgsql-bugs по дате отправления: