Re: [sqlite] possible bug 3.8.1 /3.8.3

2014-02-03 Thread Adam Devita
Good day all, Thank you for your replies. Yes, I can provide the data if required, although I don't think it is needed, as the bug is in the user's code. The point about what happens if several timestamps have the same value is valid, and in this case, I think is the explanation. sqlite>

Re: [sqlite] possible bug 3.8.1 /3.8.3

2014-02-03 Thread Richard Hipp
Can you provide data? Without some sample data, we cannot tell if the answer SQLite is providing is right or wrong. On Mon, Feb 3, 2014 at 12:25 PM, Adam Devita wrote: > Good day, > > I'm debugging some code that uses 3.8.1, and I've tried just upgrading to > 3.8.3,

Re: [sqlite] possible bug 3.8.1 /3.8.3

2014-02-03 Thread Igor Tandetnik
On 2/3/2014 12:25 PM, Adam Devita wrote: This query: SELECT id, data_blob FROM data WHERE timestamp = (SELECT MIN(timestamp) FROM data WHERE stream_num = ?) LIMIT 1 seems to occasionally produce a wrong result (the content of data_blob is incorrect given the values of stream_num) yet this

Re: [sqlite] possible bug 3.8.1 /3.8.3

2014-02-03 Thread Dan Kennedy
On 02/04/2014 12:25 AM, Adam Devita wrote: Good day, I'm debugging some code that uses 3.8.1, and I've tried just upgrading to 3.8.3, which didn't work. The observation is that This query: SELECT id, data_blob FROM data WHERE timestamp = (SELECT MIN(timestamp) FROM data WHERE stream_num = ?)

[sqlite] possible bug 3.8.1 /3.8.3

2014-02-03 Thread Adam Devita
Good day, I'm debugging some code that uses 3.8.1, and I've tried just upgrading to 3.8.3, which didn't work. The observation is that This query: SELECT id, data_blob FROM data WHERE timestamp = (SELECT MIN(timestamp) FROM data WHERE stream_num = ?) LIMIT 1 seems to occasionally produce a