From: | Márcio A(dot) Sepp <marcio(at)zyontecnologia(dot)com(dot)br> |
---|---|
To: | "'Tom Lane'" <tgl(at)sss(dot)pgh(dot)pa(dot)us>, "'Rakesh Kumar'" <rakeshkumar464(at)outlook(dot)com> |
Cc: | "'pgsql-general'" <pgsql-general(at)postgresql(dot)org> |
Subject: | RES: Chante domain type - Postgres 9.2 |
Date: | 2016-09-26 16:59:07 |
Message-ID: | 052601d21817$4b8f6990$e2ae3cb0$@com.br |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
> > Can you elaborate? Why would anyone create a text column to store
> customer name or product name which can very well be in varchar(50)
> type of cols.
>
> You sound like you think that varchar(50) is somehow cheaper than text.
> That's backwards (at least in PG, other DBMSes may be different).
> There's no advantage storage-wise, and there is a cost, namely the cost
> of applying the length check on every update.
>
> If you feel that you must have a check for application-specific
> reasons, then sure, use varchar(n). But the number had better be one
> that you can trace to crystal-clear application requirements.
> varchar(n) where n has been plucked from the air is a good sign of bad
> database design.
What a about using text x varchar(n) in primary key?
From | Date | Subject | |
---|---|---|---|
Next Message | Adam Brusselback | 2016-09-26 17:56:18 | Incrementally refreshed materialized view |
Previous Message | David G. Johnston | 2016-09-26 16:23:52 | Re: [HACKERS] temporary table vs array performance |