On 21 Mar 2018, at 6:17pm, Jonathan Moules <jonathan-li...@lightpear.com> wrote:
> Simon - I suspected the ORDER BY thing was wrong but wanted to check first
> rather than simply come out with "SQLite is broken!". This may be related to
> the 3.22 regression I brought up a couple of days ago (and why I'm using
> 3.15) - probably why Dr H is suggesting I try his branch.
> I'm executing the query using SQLiteStudio (Or Python).
It is not supposed to be possible to speed up a query by adding an ORDER BY
clause. The very worst that SQLite is meant to do is pick a fast order for the
results. If the only difference between the two queries really is an ORDER BY
clause, it looks like you've found something unexpected in SQLite.
I don't know what operating system you're using, but I hope Dr H will pilot you
through a way to provide the developers with a way to reproduce this problem.
You can provide some useful information by using the precompiled binary of the
sqlite shell tool to give the output from putting "EXPLAIN QUERY PLAN" before
your SELECT commands:
EXPLAIN QUERY PLAN SELECT
u.url, l.error_code <etc.>
The difference between the two might provide some clues about what's going on.
Simon.
_______________________________________________
sqlite-users mailing list
sqlite-users@mailinglists.sqlite.org
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
