|From:||Stephen Frost <sfrost(at)snowman(dot)net>|
|To:||Jeevan Chalke <jeevan(dot)chalke(at)gmail(dot)com>|
|Subject:||Re: Review of GetUserId() Usage|
|Views:||Raw Message | Whole Thread | Download mbox | Resend email|
* Jeevan Chalke (jeevan(dot)chalke(at)gmail(dot)com) wrote:
> The following review has been posted through the commitfest application:
> make installcheck-world: tested, passed
> Implements feature: tested, passed
> Spec compliant: tested, passed
> Documentation: tested, passed
> I have reviewed the patch.
> Patch is excellent in shape and does what is expected and discussed.
> Also changes are straight forward too.
> So looks good to go in.
> However I have one question:
> What is the motive for splitting the function return value from
> SIGNAL_BACKEND_NOPERMISSION into
> SIGNAL_BACKEND_NOSUPERUSER and SIGNAL_BACKEND_NOPERMISSION?
> Is that required for some other upcoming patches OR just for simplicity?
That was done to provide a more useful error-message to the user. It's
not strictly required, I'll grant, but I don't see a reason to avoid
doing it either.
> Currently, we have combined error for both which is simply split into two.
> No issue as such, just curious as it does not go well with the subject.
It seemed reasonable to me to improve the clarity of the error messages.
> You can mark this for ready for committer.
I've also claimed it as a committer and, barring objections, will go
ahead and push it soonish.
|Next Message||Pavel Stehule||2015-02-28 05:25:17||Re: Providing catalog view to pg_hba.conf file - Patch submission|
|Previous Message||Stephen Frost||2015-02-28 04:25:31||Re: Improving RLS qual pushdown|