Fwd: Re: [CORE] temporal tables (SQL2011)

Started by Stefan Scheidabout 9 years ago5 messages
#1Stefan Scheid
ssc@4braincells.de

Hi all,

are there plans to introduce temporal tables?

best,

Stefan

-------- Weitergeleitete Nachricht --------
Betreff: Re: [CORE] temporal tables (SQL2011)
Datum: Fri, 4 Nov 2016 10:27:40 -0400
Von: Peter Eisentraut <peter_e@gmx.net>
An: Stefan Scheid <ssc@4braincells.de>
Kopie (CC): pgsql-core@postgresql.org

On 11/1/16 12:08 PM, Stefan Scheid wrote:

how about implementing this feature?

Want to have a real argument to move 150 customers from mysql to
postgresql ...
cause they are not able or willing to use DB2 or Oracle ...

The core team does not coordinate the development effort. Please write
to pgsql-hackers to discuss development ideas.

#2Craig Ringer
craig@2ndquadrant.com
In reply to: Stefan Scheid (#1)
Re: Fwd: Re: [CORE] temporal tables (SQL2011)

On 7 November 2016 at 05:08, Stefan Scheid <ssc@4braincells.de> wrote:

Hi all,

are there plans to introduce temporal tables?

I don't know of anybody working on them, but someone else may. Try

searching the list archives.

PostgreSQL development happens because people who want features step up and
either implement them or convince someone else to implement what they need.
The roadmap, such as it is, is "what the contributors and their various
customers want".

If this is important to you, look into what you need to do to make it
happen.

--
Craig Ringer http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services

#3Robert Haas
robertmhaas@gmail.com
In reply to: Stefan Scheid (#1)
Re: Fwd: Re: [CORE] temporal tables (SQL2011)

On Sun, Nov 6, 2016 at 4:08 PM, Stefan Scheid <ssc@4braincells.de> wrote:

are there plans to introduce temporal tables?

I don't know of anyone who is actually working on it, but I agree that it
would probably attract some users if we did.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

#4Craig Ringer
craig.ringer@2ndquadrant.com
In reply to: Craig Ringer (#2)
Re: Fwd: Re: [CORE] temporal tables (SQL2011)

On 8 Nov. 2016 15:11, "Craig Ringer" <craig@2ndquadrant.com> wrote:

On 7 November 2016 at 05:08, Stefan Scheid <ssc@4braincells.de> wrote:

Hi all,

are there plans to introduce temporal tables?

I don't know of anybody working on them, but someone else may. Try

searching the list archives.

I should've mentioned that one of the reasons it doesn't seem to be that
high on many people's priority lists is that it's fairly easy to implement
with triggers and updatable views. There's a greater performance cost than
I'd expect to pay for the same thing done as a built-in feature, but it
works well enough.

Many ORMs and application frameworks also offer similar capabilities at the
application level.

So I think temporal tables are one of those nice-to-haves that so far
people just find other ways of doing.

#5Stefan Scheid
stefan.scheid@4braincells.de
In reply to: Craig Ringer (#4)
Re: Fwd: Re: [CORE] temporal tables (SQL2011)

Hi,
thanks for elaborating.

yes, of course, I can implement
it with 3 triggers, adding a couple of columns. It doesn't affect design and testing which stay the same.
As we are developing a product that must support a couple of databases and as I am not really happy with Maria e.a.,
I want to switch our standard DBMS. We need to support ms and ora as well, so there are h2, db2 and pg, or maybe we switch to some nonrel stuff like neo.
A couple of years ago I migrated a cms from db2 to pg, and was quite impressed... thats my current "mind map" :-)

Von meinem iPhone gesendet

Show quoted text

Am 10.11.2016 um 01:26 schrieb Craig Ringer <craig.ringer@2ndquadrant.com>:

On 8 Nov. 2016 15:11, "Craig Ringer" <craig@2ndquadrant.com> wrote:

On 7 November 2016 at 05:08, Stefan Scheid <ssc@4braincells.de> wrote:

Hi all,

are there plans to introduce temporal tables?

I don't know of anybody working on them, but someone else may. Try searching the list archives.

I should've mentioned that one of the reasons it doesn't seem to be that high on many people's priority lists is that it's fairly easy to implement with triggers and updatable views. There's a greater performance cost than I'd expect to pay for the same thing done as a built-in feature, but it works well enough.

Many ORMs and application frameworks also offer similar capabilities at the application level.

So I think temporal tables are one of those nice-to-haves that so far people just find other ways of doing.