Re: Roadmap for FE/BE protocol redesign - Mailing list pgsql-hackers

From Emmanuel Charpentier
Subject Re: Roadmap for FE/BE protocol redesign
Date
Msg-id b4uo54$1q15$1@news.hub.org
Whole thread Raw
In response to Re: Roadmap for FE/BE protocol redesign  (Justin Clift <justin@postgresql.org>)
List pgsql-hackers
Justin Clift wrote:

[ ... ]

> The problem Dave is suggesting this as a first attempt at a solution for 
> is that with ODBC, a frontend (i.e. OpenOffice) asks the ODBC driver 
> which columns are NULLable, etc.  And the ODBC driver is getting the 
> info wrong, then passing back the incorrect info.

And that's probably why OpenOffice doesn't allow you to update a view, 
event if there are rules allowing this from psql ...

This, in my book is an EPITA for end users ...

> So, when a person goes to insert a row into a table with a 
> SERIAL/SEQUENCE based column, OpenOffice has been told the column isn't 
> NULLable and forces the user to enter a value.  Voila, it doesn't work 
> with sequences.  :(
> 
> It's likely possible to add to the ODBC driver some way of getting the 
> info right, but Dave is also looking for a way of making this easier 
> into the future for similar problems.  i.e. Let the database explicitly 
> have info about what each column can do.

I'd second that, even if it takes time ...
                Emmanuel Charpentier

-- 
Emmanuel Charpentier



pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: [INTERFACES] Upgrading the backend's error-message infrastructure
Next
From: Bruno Wolff III
Date:
Subject: Re: No index maximum? (was Re: No merge sort?)