> pgsql(at)mohawksoft(dot)com writes:
>> Having internal PostgreSQL variables that are not present on disk, or
>> maybe, variables that are mirrored on disk may be good.
> I don't think there's anything wrong with your idea, and there are
> good solutions that implement it already. But what makes you think this
> belongs in Postgres?
> There are plenty of memory and disk based shared databases that are
> non-transactional and non-relational and meant for storing just this kind
> non-relational data. Some are much faster than postgres for simple
> non-concurrent one-record lookups and updates like this.
> Use the right tool for the job. Don't try to make one tool do everything,
> especially something that's anathema to its basic design.
I agree completely with one caveat, when the best tool for the job lacks a
feature what do you do?
> ---------------------------(end of broadcast)---------------------------
> TIP 2: you can get off all lists at once with the unregister command
> (send "unregister YourEmailAddressHere" to majordomo(at)postgresql(dot)org)
In response to
pgsql-hackers by date
|Next:||From: Christopher Browne||Date: 2004-05-29 18:59:22|
|Subject: Re: Win32, PITR, nested transactions, tablespaces|
|Previous:||From: Magnus Hagander||Date: 2004-05-29 18:45:28|
|Subject: Re: dynamic_library_path on Win32 |