[sqlite] NULL check in shell.c

2016-09-02 Thread Max Radermacher
With respect to shell.c:shell_exec, there is an inconsistency in how NULL checks are applied. Despite the fact that it seems like `shell_exec` is never called with pArg set to NULL, the method is written such that it could be. In order to be consistent, it seems like pArg should be checked for

Re: [sqlite] BUG: Documentation of API have incorrect anchors...

2016-09-02 Thread Richard Hipp
On 9/1/16, Rob Wehrli wrote: > From: http://www.sqlite.org/c3ref/funclist.html > > All of these functions have the same anchor: > > http://www.sqlite.org/c3ref/backup_finish.html#sqlite3backupfinish > > sqlite3_backup_finish > sqlite3_backup_init >

Re: [sqlite] VARCHAR or TEXT to save sapce

2016-09-02 Thread Richard Hipp
On 9/2/16, Dominique Devienne wrote: > > Does ANALYZE gather actual statistics on lengths? The sqlite_stat1 table has a place to put that information and SQLite will use the information if it is present, but the ANALYZE command does not currently generate it. -- D.

Re: [sqlite] VARCHAR or TEXT to save sapce

2016-09-02 Thread Dominique Devienne
On Fri, Sep 2, 2016 at 5:51 PM, Richard Hipp wrote: > On 9/2/16, Dominique Devienne wrote: > > > > For example, we use Guid extensively, which are typed blob, but often > > add that CHECK(length(guid) = 16) constraint. I know 20 is not far off > > from 16

Re: [sqlite] VARCHAR or TEXT to save sapce

2016-09-02 Thread Richard Hipp
On 9/2/16, Dominique Devienne wrote: > > For example, we use Guid extensively, which are typed blob, but often > add that CHECK(length(guid) = 16) constraint. I know 20 is not far off > from 16 in this particular case, but more accurate is always better, no? The difference

Re: [sqlite] VARCHAR or TEXT to save sapce

2016-09-02 Thread Dominique Devienne
On Fri, Sep 2, 2016 at 5:04 PM, Richard Hipp wrote: > On Fri, 02 Sep 2016 15:10 +0100, Tim Streater > wrote: > > On 02 Sep 2016 at 15:05, Dave Blake wrote: > > > >> Can I confirm that defining feilds as VARCHAR(20) rather than TEXT, if

Re: [sqlite] VARCHAR or TEXT to save sapce

2016-09-02 Thread Richard Hipp
On Fri, 02 Sep 2016 15:10 +0100, Tim Streater wrote: > On 02 Sep 2016 at 15:05, Dave Blake wrote: > >> Can I confirm that defining feilds as VARCHAR(20) rather than TEXT, if I >> know I only want 20 chars or less, will result in a smaller database? > >

Re: [sqlite] VARCHAR or TEXT to save sapce

2016-09-02 Thread R Smith
On 2016/09/02 4:05 PM, Dave Blake wrote: Can I confirm that defining feilds as VARCHAR(20) rather than TEXT, if I know I only want 20 chars or less, will result in a smaller database? In addition to the other useful replies, you might actually be interested in ensuring the text field in

Re: [sqlite] VARCHAR or TEXT to save sapce

2016-09-02 Thread Olivier Mascia
> Le 2 sept. 2016 à 16:05, Dave Blake a écrit : > > Can I confirm that defining feilds as VARCHAR(20) rather than TEXT, if I > know I only want 20 chars or less, will result in a smaller database? That won't change anything. The (20) in VARCHAR(20) is purely 'documentation'

Re: [sqlite] VARCHAR or TEXT to save sapce

2016-09-02 Thread Tim Streater
On 02 Sep 2016 at 15:05, Dave Blake wrote: > Can I confirm that defining feilds as VARCHAR(20) rather than TEXT, if I > know I only want 20 chars or less, will result in a smaller database? Makes no difference whatever. See: --

[sqlite] VARCHAR or TEXT to save sapce

2016-09-02 Thread Dave Blake
Can I confirm that defining feilds as VARCHAR(20) rather than TEXT, if I know I only want 20 chars or less, will result in a smaller database? ___ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org

Re: [sqlite] client app crashes frequently and points sqlite DLL

2016-09-02 Thread Roger Binns
On 02/09/16 03:11, Frantz FISCHER wrote: > I'm almost out of tracks to follow. Any idea on what I could check next? Are you using threads? Note that in this case whatever is happening is making SQLite be the victim. The cause is almost certainly some other C code in your process. Roger

Re: [sqlite] client app crashes frequently and points sqlite DLL

2016-09-02 Thread Frantz FISCHER
Hi, I upgraded the SQLite DLL to 3.14.1. The job finished the first time but crashed the second time mentioning sqlite3.dll. I'm almost out of tracks to follow. Any idea on what I could check next? Regards and thanks, Frantz. On 02/09/2016 11:14, Frantz FISCHER wrote: Hi, Tried with

Re: [sqlite] client app crashes frequently and points sqlite DLL

2016-09-02 Thread Frantz FISCHER
Hi, Tried with the 2.7.12 version but it still fails. I double-checked I didn't take the wrong version and compared the md5sum of python27.dll with the one from python 2.7.12, they are identical -> 797f4566d81c04ed5f21637d2d64197f This time the crash didn't generate a dump and points only

Re: [sqlite] Insert silently failing.

2016-09-02 Thread Simon Slavin
On 2 Sep 2016, at 8:52am, Simon Lepasteur wrote: > I have multiple processes accessing the same database file (sqlite 3.8.7.4 > on linux). The DB contains only one table and every process execute > prepared statements (sqlite3_prepare_v2) like the following one: > >

[sqlite] Insert silently failing.

2016-09-02 Thread Simon Lepasteur
Hello, I have multiple processes accessing the same database file (sqlite 3.8.7.4 on linux). The DB contains only one table and every process execute prepared statements (sqlite3_prepare_v2) like the following one: INSERT INTO table (process,type,status,occurence,data)