Thanks, Edzard. For the record, we also found that there is a good explanation of this in the book "Using SQLite" from O'Reilly on pages 153-154. We opened an issue [1] on the Rails project.
best, Peter [1] https://rails.lighthouseapp.com/projects/8994-ruby-on-rails/tickets/5941 On 11/10/10 2:10 PM, Edzard Pasma wrote: > The immediate busy-error is indeed intended behaviour, or rather a > consequence of the technical design. > > It took me a while to find where it is documented. That is not in > "Locking and Concurrency in SQLite version 3" but in the "C/C++ API > Reference". And there it is not in the description of > sqlite3_busy_timeout but under sqlite3_busy_handler. There it is > explained clear enough. > > What has surprised me that this deadlock is unchanged when accessing > a database in WAL mode. Also then it is intended behaviour, this was > explained in an earlier discussion. But I don't know if it will > remain like this in future versions. > > Best Regards, Edzard. > -- Peter Pawlowski | Senior Software Engineer Office: +1.412.422.2499 x116 pawlow...@vivisimo.com | Connect: www.vivisimo.com Vivisimo - Information Optimized _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
