built-in sqlite support in 29 - still time for changes?

built-in sqlite support in 29 - still time for changes?

[Lynn Winebarger]

I'm curious if there is time or appetite for the following changes to
built-in sqlite support for emacs 29, before any code becomes
dependent on the current implementation.

*  Can rows be returned as arrays rather than lists for consistency
with emacsql programming conventions?

*  Can (or should) sqlite-mode be derived from tabulated list mode
instead of directly from special?

*  Can we add at least one additional primitive type for tables
(corresponding to a prepared sqlite statement) so a database
connection can have multiple outstanding queries?  Or have one query
use the result of an existing query result?

*  Is the current code mature enough to document properly, or is that
being delayed until the implementation is more mature?

Lynn

Re: built-in sqlite support in 29 - still time for changes?

[Eli Zaretskii]

> From: Lynn Winebarger <owinebar@gmail.com>
> Date: Tue, 21 Mar 2023 09:18:16 -0400
> 
> *  Is the current code mature enough to document properly, or is that
> being delayed until the implementation is more mature?

I'm not sure I understand what do you mean by "document properly",
since the SQLite support is already documented in the ELisp reference
manual.

(I have no opinion on the other issues you raise.)

Re: built-in sqlite support in 29 - still time for changes?

[Lynn Winebarger]

On Tue, Mar 21, 2023 at 9:35 AM Eli Zaretskii <eliz@gnu.org> wrote:
> > From: Lynn Winebarger <owinebar@gmail.com>
> > Date: Tue, 21 Mar 2023 09:18:16 -0400
> >
> > *  Is the current code mature enough to document properly, or is that
> > being delayed until the implementation is more mature?
>
> I'm not sure I understand what do you mean by "document properly",
> since the SQLite support is already documented in the ELisp reference
> manual.

That's user error on my part.  Even though I was running my last built
version of master, the last version of emacs I built/installed was
28.3, so I was seeing the old manual.
Thanks

> (I have no opinion on the other issues you raise.)

Would patches still be accepted for emacs-29?  I don't know what the
cutoff is.  For the first point in particular, I don't know if the
emacsql convention is the best, but it seems gratuitous to change the
idiom unless there's a good reason.  It seems like emacsql query
compiler could be adapted to the builtin support with the advantage of
not being restricted to string values, so existing code could be
conveniently ported.


Lynn

Re: built-in sqlite support in 29 - still time for changes?

[Eli Zaretskii]

> From: Lynn Winebarger <owinebar@gmail.com>
> Date: Tue, 21 Mar 2023 10:02:34 -0400
> Cc: emacs-devel@gnu.org
> 
> On Tue, Mar 21, 2023 at 9:35 AM Eli Zaretskii <eliz@gnu.org> wrote:
> 
> > (I have no opinion on the other issues you raise.)
> 
> Would patches still be accepted for emacs-29?

No, emacs-29 is closed for new features.  Only bug fixes are
acceptable.