Re: [sqlite] https://www.sqlite.org/draft/gencol.html Typo

On 2019-10-29 6:20 a.m., Simon Slavin wrote: • Every table must have at least one non-generated column. I greatly admire this restriction. From a relational purist perspective that restriction is counter-productive. But then, disallowing a table or a key/unique constraint from having

Re: [sqlite] https://www.sqlite.org/draft/gencol.html Typo

On 29 Oct 2019, at 10:16pm, Warren Young wrote: > One question I had after reading the draft doc is whether an > application-defined SQLITE_DETERMINISTIC function can be used to compute a > generated column. Hmm. Well, I can see that any non-deterministic function would be a problem here.

Re: [sqlite] https://www.sqlite.org/draft/gencol.html Typo

On Tue, Oct 29, 2019 at 5:08 PM Keith Medcalf wrote: > > On Tuesday, 29 October, 2019 16:17, Warren Young > wrote: > > >On Oct 29, 2019, at 7:20 AM, Simon Slavin wrote: > > >One question I had after reading the draft doc is whether an application- > >defined SQLITE_DETERMINISTIC function can

Re: [sqlite] https://www.sqlite.org/draft/gencol.html Typo

On Tuesday, 29 October, 2019 16:17, Warren Young wrote: >On Oct 29, 2019, at 7:20 AM, Simon Slavin wrote: >One question I had after reading the draft doc is whether an application- >defined SQLITE_DETERMINISTIC function can be used to compute a generated >column. My immediate use case for

Re: [sqlite] https://www.sqlite.org/draft/gencol.html Typo

On Oct 29, 2019, at 7:20 AM, Simon Slavin wrote: > > > > OMG. Much welcomed feature. Yes, I can see immediate use for this. One question I had after reading the draft doc is whether an application-defined SQLITE_DETERMINISTIC function can be used

Re: [sqlite] https://www.sqlite.org/draft/gencol.html Typo

te mailing list > Subject: Re: [sqlite] https://www.sqlite.org/draft/gencol.html > Typo > > > On Tuesday, 29 October, 2019 12:25, Doug > wondered: > > >The draft says "Nor may a generated column depend on the ROWID." > > >If my table uses ROWID by default:

Re: [sqlite] https://www.sqlite.org/draft/gencol.html Typo

On Tuesday, 29 October, 2019 12:25, Doug wondered: >The draft says "Nor may a generated column depend on the ROWID." >If my table uses ROWID by default: >CREATE TABLE foo (id INTEGER PRIMARY KEY, a INTEGER, b AS (id+a)); >where id is ROWID by default, is the generated column disallowed

Re: [sqlite] https://www.sqlite.org/draft/gencol.html Typo

Keith Medcalf wrote: > At the end of the second paragraph of section 2.1: > > Only VIRTUAL tables can be added using ALTER TABLE. > > should be > > Only VIRTUAL columns can be added using ALTER TABLE. Above typos is already corrected, but here are 2 other typos in the same page:

Re: [sqlite] https://www.sqlite.org/draft/gencol.html Typo

WID? Doug > -Original Message- > From: sqlite-users > On Behalf Of Simon Slavin > Sent: Tuesday, October 29, 2019 6:21 AM > To: SQLite mailing list > Subject: Re: [sqlite] https://www.sqlite.org/draft/gencol.html > Typo > > <https://www.sqlite.org/draft/genco

Re: [sqlite] https://www.sqlite.org/draft/gencol.html Typo

OMG. Much welcomed feature. > • Every table must have at least one non-generated column. I greatly admire this restriction. Can I suggest an addition to gencol.html ? An explicit statement saying whether VIRTUAL and/or STORED columns can be used in

[sqlite] https://www.sqlite.org/draft/gencol.html Typo

At the end of the second paragraph of section 2.1: Only VIRTUAL tables can be added using ALTER TABLE. should be Only VIRTUAL columns can be added using ALTER TABLE. -- The fact that there's a Highway to Hell but only a Stairway to Heaven says a lot about anticipated traffic volume.