Doing SQLDIFF against a non-existent database (e.g., due to a typo) produces a
diff which is a series of DROP TABLE ... and also creates an empty file by that
name.
This is misleading as it looks like a possibly valid diff.
Shouldn?t it instead give an error that there is no such database?
>From: Domingo Alvarez Duarte
>I have a question, is it enough to vacuum a database to update to the new
>page size ?
Apparently all you need to is "pragma page_size=4096; vacuum;" using the
appropriate page size.
This makes very easy to convert any(all) database(s) with a single command
from
>From: Richard Hipp
>Can you run sqlite3_analyzer on some of your databases and send me the
>output?
I will as soon as I can manage to build it under Windows. Although I
regularly build sqlite3, lemon, and sqldiff, with sqlite3_analyzer I get a
lot of errors like these:
...
I ran some tests and almost all of my databases (about 100 of them with
different content mix and with the biggest one being around 500MB) inflated
by a lot while only a couple or so shrunk in size by not much. (It could be
that I'm just unlucky with my database contents and this change simply
Compiling under Linux with SQLITE_OMIT_WAL I get this error:
./sqlite3.c: In function ?sqlite3PagerJrnlFile?:
./sqlite3.c:50209:16: error: ?Pager? has no member named ?pWal?
return pPager->pWal ? sqlite3WalFile(pPager->pWal) : pPager->jfd;
-Original Message-
From: Richard Hipp
Sent:
[52c16603] now works OK for me. Thanks.
-Original Message-
From: to...@acm.org
Sent: Friday, January 08, 2016 11:55 PM
To: SQLite mailing list
Subject: Re: [sqlite] SQLite3 v3.10.0 compilation error
Got the latest [3392f8fa] fix from the trunk but now get these
errors/warnings
2: warning: ?unixGetpagesize? used but never defined
[enabled by default]
static int unixGetpagesize(void);
^
-Original Message-
From: to...@acm.org
Sent: Friday, January 08, 2016 8:52 PM
To: General Discussion of SQLite Database
Subject: [sqlite] SQLite3 v3.10.0 compilation error
(Under
(Under Linux)
I?m using the SQLITE_OMIT_WAL option and because of this the #define osReadlink
is not defined (amalgamation line # 27508) which is later required by
unixFullPathname function, and compilation fails.
Thanks.
*** SORRY IF YOU HAVE RECEIVED THIS TWICE ? NOT SURE IT MADE IT THE
1. I?m trying to create a new repo by taking a branch from an existing one.
So, I thought the easier way would be the new ?bundle? feature.
I tried this batch file (Win7):
-
f new xxx.fossil --date-override 2014-01-01
f o xxx.fossil
f bun export xxx --standalone --branch %1 ?R
Here?s a report about an SQLDIFF utility problem.
sqldiff: SQL statement error: near "IS": syntax error
To reproduce, put the schema below (produced automatically by .IMPORT from a
CSV file so don?t bother with its appearance ? but it works in sqlite3) inside
some file like xxx.sql, and then do
I?ve been compiling new versions for a long time now, and I just noticed this
(so I don?t know how long it?s been there).
Here?s the relevant output from the compilation:
--
tclsqlite.c
lib.exe /NOLOGO /LIBPATH:c:\tcl\lib /OUT:libtclsqlite3.lib tclsqlite.lo
That's very likely what happened the first, must have put the -lm option
before the .c files :(
(Not a regular gcc user.)
Thanks.
-Original Message-
From: Rowan Worth
Sent: Monday, October 19, 2015 6:43 AM
To: General Discussion of SQLite Database
Subject: Re: [sqlite] Compilation
) when FTS5 is
enabled
On 10/18/2015 02:21 AM, tonyp at acm.org wrote:
> I'm trying to compile the shell using the amalgamation (three files
> involved: shell.c sqlite3.c sqlite.h).
> This is a process I've done a zillion times before without problems.
> The only new thing I
shell.c sqlite3.c
tonyp at xxx:~/temp$
-Original Message-
From: Dan Kennedy
Sent: Saturday, October 17, 2015 10:36 PM
To: sqlite-users at mailinglists.sqlite.org
Subject: Re: [sqlite] Compilation fails under Linux (Ubuntu) when FTS5 is
enabled
On 10/18/2015 02:21 AM, tonyp at acm.org
Yes, it does!
-Original Message-
From: Simon Slavin
Sent: Saturday, October 17, 2015 10:38 PM
To: General Discussion of SQLite Database
Subject: Re: [sqlite] Compilation fails under Linux (Ubuntu) when FTS5
isenabled
On 17 Oct 2015, at 8:21pm, tonyp at acm.org wrote:
> I'm try
.)
-Original Message-
From: Dan Kennedy
Sent: Saturday, October 17, 2015 4:35 PM
To: sqlite-users at mailinglists.sqlite.org
Subject: Re: [sqlite] Compilation fails under Linux (Ubuntu) when FTS5 is
enabled
On 10/17/2015 06:47 PM, tonyp at acm.org wrote:
> What the subject says and I
What the subject says and I get this error:
/tmp/cco8QBGJ.o: In function `fts5Bm25GetData':
sqlite3.c:(.text+0xb521b): undefined reference to `log'
(using the latest 3.9.1 release)
Any ideas?
Thanks.
BTW, MySQL also seems to do it 'correctly'.
Plus, it apparently recognizes if the random() expression in the ORDER BY is
the same as the SELECT one and again sort correctly (without re-evaluating)
and without needing an alias.
-Original Message-
From: Domingo Alvarez Duarte
Sent:
I would like to request this improvement to the shell .import command. (Wish
list, I guess.)
Current .import behavior is that when the imported table does not exist, it
will be created using the first row from the data file as column names for the
new table.
This is great but it has one
Me too, for about the last 7-10 days or so!
-Original Message-
From: R.Smith
Sent: Friday, February 27, 2015 9:38 PM
To: sqlite-users at mailinglists.sqlite.org
Subject: [sqlite] List duplication
Hi all,
Just a quick question, I've been getting duplicated mails via the forum.
It
20 matches
Mail list logo
