On 03/09/2017 10:46 PM, James K. Lowden wrote:
On Tue, 7 Mar 2017 19:04:47 +0100
Vermes Mátyás <ver...@comfirm.hu> wrote:

It is also unnecessarily complex and slow.
The script demonstrates a regression (a bug). It is written in Ruby
so that everybody can run it, and see its _results_. It is absolutely
not interesting that it is slow or complex.
As has been repeated several times in this thread, the act of updating
a SQLite database while a SELECT is being processed interferes with the
SELECT unless WAL mode is used.

Wal mode prevents read-only clients from blocking writers, and vice-versa. But it doesn't actually change anything to do with the way multiple statements executed by the same connection interact.

Dan.







My observations:

1.  Your row-by-row update is not a good design choice.

2.  And it doesn't work correctly in SQLite, by design.

3.  SQLite fails to keep its ACID promise: SELECT is not atomic, and
not isolated from UPDATE.

The ACID fault is a recurring theme on this list.  You are not
the first person bitten by it, and won't be the last.

--jkl

_______________________________________________
sqlite-users mailing list
sqlite-users@mailinglists.sqlite.org
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users


_______________________________________________
sqlite-users mailing list
sqlite-users@mailinglists.sqlite.org
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users

Reply via email to