What is the format of the file? I cannot find it documented anywhere.
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
On 04/08/2010, at 11:12 PM, Igor Tandetnik wrote:
> I believe the two are equivalent. SQLite essentially rewrites IN to EXISTS -
> it certainly doesn't generate the complete resultset from the nested select
> and then go searching inside that.
>
> I noticed anecdotally that SQLite may optimize
On 8/10/10, Paweł Hajdan, Jr. wrote:
> So this is another chromium patch I'd like to submit:
> http://src.chromium.org/viewvc/chrome/trunk/src/third_party/sqlite/preload-cache.patch?revision=26596=markup
>
> I'm not the author of that one, but the main idea seems to be
So this is another chromium patch I'd like to submit:
http://src.chromium.org/viewvc/chrome/trunk/src/third_party/sqlite/preload-cache.patch?revision=26596=markup
I'm not the author of that one, but the main idea seems to be that with
preloading we get better performance with many scenarios
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 08/11/2010 12:12 AM, Rosbicn wrote:
> ... sqlite3_initialize() ... sqlite3_shutdown() ...
Why are you calling these functions - ie what do you hope to achieve?
Unless you are in a very small device, or doing behind the scenes
infrastructure, or
On 10 Aug 2010, at 7:45pm, David Bicking wrote:
> Actually, the amount in items is a running balance of the changes. Mind you,
> there are other fields in items (and in changes) that I did not show as they
> didn't seem to impact this decision.
>
> In that 8% case where an item is to be
Thank you, I have backported it to chromium as
http://src.chromium.org/viewvc/chrome?view=rev=55504
Now, how about fts1 and fts2? The original chromium patch is at
http://codereview.chromium.org/174387 . Could you take a look and suggest a
way to upstream those fixes to SQLite?
On Fri, Aug 6,
Hello -- we recently noticed that some SQL statements containing
subqueries (see below) are very slow in 3.6.23 as compared to the
version we were using before, 3.3.17. I tested the newest release of
3.7.0.1 and the queries are fast again.
The problem is that this is a showstopper for us, but we
On 08/08/2010 10:09 PM, Dan Kennedy wrote:
>>> 2) However, one process cannot read from the database while another
>>> is writing -- WAL is irrelevant here.
>>
>> Unless shared-cache mode is turned on, multiple threads each using
>> their own sqlite3* connection should behave in the same way as
>>
Actually, the amount in items is a running balance of the changes. Mind you,
there are other fields in items (and in changes) that I did not show as they
didn't seem to impact this decision.
In that 8% case where an item is to be created, the "change" is in fact the
creation of the item.
On Aug 11, 2010, at 1:25 AM, David Barrett wrote:
> On 08/08/2010 10:00 PM, Dan Kennedy wrote:
>>
>> On Aug 9, 2010, at 11:17 AM, David Barrett wrote:
>>> 3) When an application performs read/write queries on the database
>>> in
>>> parallel to the .backup command being run on the database,
On 08/08/2010 10:00 PM, Dan Kennedy wrote:
>
> On Aug 9, 2010, at 11:17 AM, David Barrett wrote:
>> 3) When an application performs read/write queries on the database in
>> parallel to the .backup command being run on the database, will the
>> application occasionally get the SQLITE_LOCKED return
David,
Your approach contravenes "best practice" by violating the core referential
integrity paradigm: your CHANGES table refers to an item not yet in the
ITEMS table and actually governs whether an ITEM item can be created. The
child is giving birth to the parent. This is unnecessarily
David Bicking wrote:
> I am building an application with these two tables:
>
> CREATE TABLE changes(ChangeDate, Code, AdjAmount, Descr, Creatable);
> CREATE TABLE items(Code, Amount)
>
> Now, what I would like to happen, I insert in to changes, and it updates the
> Amount in
I am building an application with these two tables:
CREATE TABLE changes(ChangeDate, Code, AdjAmount, Descr, Creatable);
CREATE TABLE items(Code, Amount)
Now, what I would like to happen, I insert in to changes, and it updates the
Amount in items.
I can get that with
CREATE TRIGGER
I use sqlite in a .Net dll, which is used in a aspx web page under
IIS. In the code, sqlite3_initialize() is called in the beginning of all
codes, and sqlite3_shutdown() is called in the end of all codes. Now I
find the database file is "hanged" by IIS when many browsers fetch the
web
Hi,
Yes they are.
With "PRAGMA wal_checkpoint" it works as expected.
Yoni.
On Tue, Aug 10, 2010 at 2:50 PM, Dan Kennedy wrote:
>
> On Aug 10, 2010, at 6:03 PM, Yoni Londner wrote:
>
> > Hi,
> > Yes, it explains allot.
> > But, according to you, the following program
I've done this before (especially with db's that don't have rollback).
Add an update flag. Set it to 2 for the records you update. Then set it to 1
when done. Then set it to 0.
2 -- transaction in progress
1 -- transaction being completed
0 -- transaction completed
When you start up the
On Aug 10, 2010, at 6:03 PM, Yoni Londner wrote:
> Hi,
> Yes, it explains allot.
> But, according to you, the following program should work.
> I don't use transactions, shared cache or threads.
> Just run the checkpoint from another connection.
> And still - WAL file is getting bigger and bigger
Hi,
Yes, it explains allot.
But, according to you, the following program should work.
I don't use transactions, shared cache or threads.
Just run the checkpoint from another connection.
And still - WAL file is getting bigger and bigger without limit.
If I do the checkpoint with the same
> Hi,
> I just wanted to add that I changed the program a little bit to not
> use
> transactions and threads, and I still get the same problem (huge WAL
> file).
> All I do is endless loop of insert, and every X insert, I perform a
> checkpoint on another sqlite connection (but in the same
On Aug 10, 2010, at 11:55 AM, Victor Morales-Duarte wrote:
> As it turns out, I can reproduce the failure using a single huge
> insert.
> The code that I'm including below compiles under bcc32 from
> Embarcadero's C++ Builder 2007 and cl from vs2005. Since it's more
> likely that people have
Hi,
I just wanted to add that I changed the program a little bit to not use
transactions and threads, and I still get the same problem (huge WAL file).
All I do is endless loop of insert, and every X insert, I perform a
checkpoint on another sqlite connection (but in the same thread).
It only
Hi Richard,
Thanks for the quick response.
I wrote a little program that demonstrate the error (below).
It opens a new DB, create a table, and start an endless loop of INSERT's.
In a background thread it performs a wal checkpoint.
There is no difference in WAL file size - with or without
On Mon, 9 Aug 2010 17:53:02 -0500, Mike Henshaw
wrote:
>>> Is there a way to create an incremental row or show a row number that is
>>> linked to the current select query that can be used in
>>> calculations?
>>
>>> Basically a row counter for the current query
On Tue, Aug 10, 2010 at 2:24 AM, Yoni Londner wrote:
> Hello,
> I have a questions about the correct use of transactions and WAL.
>
> I am writing an application that:
> 1. should very fast
> 2. should be very responsive
> 3. don't care if the last N minutes of data will
Hello,
I have a questions about the correct use of transactions and WAL.
I am writing an application that:
1. should very fast
2. should be very responsive
3. don't care if the last N minutes of data will be lost (but DB should
never be corrupted)
What I tried to do:
1. open a
27 matches
Mail list logo