[Tarantool-patches] [PATCH v3 3/5] sql: introduce SET statement

[Mergen Imeev]

On Thu, Nov 28, 2019 at 12:03:52AM +0100, Vladislav Shpilevoy wrote:
> On 27/11/2019 11:33, Mergen Imeev wrote:
> > Hi! Thanks for the suggestion! I think your points are correct,
> > but I think we should create a new issue after pushing SET.
> > Issue #4621 about using VDBE in SET/control pragmas. All other
> > pragmas already use VDBE. Since SET will use VDBE, I think we
> > should close #4621 with #4511.
> 
> I see that we have different understanding of what it means - use
> VDBE. I mean, that option value should be get/set at runtime.
> Currently the only thing runtime does is return of metadata. All
> the actual work happens at compile time.
> 
> To make binding works the only solution is to make
> PRAGMA/SET/whatever work at runtime totally. When you will
> properly move PRAGMA/SET to VDBE, you will get bindings for
> free.
> 
I see. Now I understand what you meant earlier. I added a
comment to #4621 and replaced “Closes #4621” with
“Part of #4621” in the commit message.

Link to comment:
https://github.com/tarantool/tarantool/issues/4621#issuecomment-559399461

> The patchset LGTM. Perhaps. I don't really know what is
> our plan on session and SQL settings, so I can only approve the
> technical part.
> 
> Seems like in our team design decisions about everything work
> like 'lets see what PostgreSQL/MySQL/Ya.DB do, and we will do
> literally the same'.