[maemo-developers] SQLite3

[From: Kamen Bundev]

Nick,

Yes, I know the changes and I'm not against using 3.6.10 on Maemo. I even
compiled 3.6.8 the other day to test it. I had the wrong impression that i
read there are file format changes, but maybe it was for an even older
version. However that's not the point. The point is that 3.6.10 is not
tested on Maemo and MAY break somebody else's application. Even changes like
this can be tricky for someone sometimes:

"The error code is changed to SQLITE_BUSY (instead of SQLITE_ERROR) when an
> attempt is made to ROLLBACK while one or more queries are still pending."
>

Just like you I'm upgrading whenever a new library comes available for my
distribution. the problem is that the distribution lifecycle is half an year
lately, so from 3.4.1 till now at least 3 Ubuntus were released. Are you
sure that applications from the Fawn are working with 3.6.10 flawlessly? I
wouldn't try to upgrade 3.4.1 to 3.6.10 just for the sake of it. I probably
would on new release, but I'm not to decide.

This is the full list of the changes:
http://www.sqlite.org/changes.html

Regards:
Bundyo


On Sun, Jan 18, 2009 at 6:08 PM, Dr. Nicholas Shaw <Doc at docharley.com>wrote:

>  Kamen,
>
>
>
> Even if 3.4.1 worked in command line, I can't see why you would want to use
> it when 3.6.10 is the current release.  Here are some of the changes from
> 3.4.2 to 3.6.0:
>
>
>
> http://www.sqlite.org/34to35.html (from 3.4.2 to 3.5.0)
>
>
>
> ·         The OS interface layer was completely reworked
>
> ·         The optional shared cache and memory management features that
> were introduced in v.3.3.0 can now be used across multiple threads within
> the same process.  Formerly, these extensions only applied to database
> connections operating within a single thread
>
> ·         Restrictions on the use of the same database connection by
> multiple threads have been dropped
>
> ·         There is now a compile-time option that allows an application to
> define alternative malloc()/free() implementations without having to modify
> any core SQLite code
>
> ·         There is now a compile-time option that allows an application to
> define alternative mutex implementations without having to modify any core
> SQLite code
>
> http://www.sqlite.org/35to36.html (from 3.5.9 to 3.6.0)
>
>
>
> ·         Changes to the sqlite3 vfs object
>
> ·         The signature of the xCheckReservedLock method on sqlite3 io
> methods has been modified so that it returns an error code and stores its
> Boolean result into an integer pointed to by a parameter.  A new extended
> error code, SQLITE IOERR CHECKRESERVEDLOCK has been added
>
> ·         When SQLite is ported to new OS, two new functions,
> sqlite3_os_init() and sqlite3_os_end() must be provided as part of the port
>
> ·         The way in which the IN and NOT IN operators handle NULL values
> in their right-hand expressions has been brought into compliance with the
> SQL standard and with other SQL database engines
>
> ·         The column names for the result sets of SELECT statements have
> been tweaked in some cases to work more like other SQL database engines
>
> ·         Changes to compile-time options
>
>
>
> Since the current version is 3.6.10, there have been numerous changes since
> 3.6.0.  So, IMHO, it makes sense to upgrade if you regularly use SQLite.
> Since the files/locations I provided do not match any other files/locations,
> there should be no conflicts.  Over the past 24-hours, I've run several of
> my regular programs and have identified no issues.  Should I do so, I'll
> post those.  I regularly upgrade SQLite on my regular Linux servers (now
> running at 3.6.10 and will so upgrade the 810).
>
> Regards,
>
>
>
> Nick.
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.maemo.org/pipermail/maemo-developers/attachments/20090118/e9913a99/attachment.htm