Re: API change advice: Passing plan invalidation info from the rewriter into the planner?

On Tue, Jun 17, 2014 at 9:45 PM, Brightwell, Adam
<adam(dot)brightwell(at)crunchydatasolutions(dot)com> wrote:
> I absolutely appreciate all of the feedback that has been provided. It has
> been educational. To your point above, I started putting together a wiki
> page, as Stephen has spoken to, that is meant to capture these concerns and
> considerations as well as to capture ideas around solutions.
>
> https://wiki.postgresql.org/wiki/Row_Security_Considerations
>
> This page is obviously not complete, but I think it is a good start.
> Hopefully this document will help to continue the conversation and assist in
> addressing all the concerns that have been brought to the table. As well, I
> hope that this document serves to demonstrate our intent and that we *are*
> taking these concerns seriously. I assure you that as one of the
> individuals who is working towards the acceptance of this feature/patch, I
> am very much concerned about meeting the expected standards of quality and
> security.

Cool, thanks for weighing in. I think that page is a good start. An
item that I think should be added there is the potential overlap
between security_barrier views and row-level security. How can we
reuse code (and SQL syntax?) for existing features like WITH CHECK
OPTION instead of writing new code (and inventing new syntax) for very
similar concepts?

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