Re: case insensitive LIKE with utf8 (multibyte)

On Fri, 30 May 2003, Jeremy March wrote: > The problem is that querying with a LIKE expression is treated case > and accent sensitive. [...] Has anyone else experienced this with > multibyte utf8? Yes. mysql-standard-4.1.0-alpha-pc-linux-i686 running with --default-character-set=utf8 gives me:

Re: MySQL 4.0 on Debian

Hello On Thu, 1 Aug 2002, Quinten Steenhuis wrote: > Has anyone successfully gotten MySQL 4.02 running on Debian? I'm > running into problems if I use alien to convert the RPMs. Yep, we're using 4.0.2 on Debian. You may want to grab MySQL .tar.gz instead of .rpm and install it somewhere into /

MySQL 4.0.1 Solaris 2.7 slave crash / mysqlimport

Hello MySQL 4.0.1 Solaris 2.7 official binary reproducibly crashes with sig 11 when run in a slave mode and when its master (4.0.1 official Linux binary) is populated via "mysqlimport" tool from a remote machine. Here are some interesting log files [edited]: , $ mysqld --user=mysql | 020612

Re: Testing MySQL 4.0

Hello On Tue, 19 Jun 2001, Sasha Pachev wrote: > If you have a spare development system, plug in 4.0 binary and try > to run your applications on it to see if it will crash. I just tried it on Debian GNU/Linux for i386 with up-to-date "testing" distro and up-to-date "bk clone". The compile pr

Re: Query optimization

Hello On Sat, 17 Mar 2001, Bård Farstad wrote: > I have a query in my program which is taking to long when I get data > in the database. I'm currently testing with ~40.000 articles in the > system. [...] Is it possible to optimize this query further? Have you played with EXPLAIN SELECT and fri

Re: Problems building 3.23.33 on Solaris 2.6

On Wed, 7 Mar 2001, Steve Sizemore wrote: > I'm trying to install the latest stable mysql on Solaris 2.6. [...] > ld.so.1: ../sql/mysqld: fatal: libz.so: open failed: No such file or > directory [...] /opt/local/lib/libz.so Apparently /opt/local/lib isn't in your LD_LIBRARY_PATH. Check it out v

Re: Some questions on indexes.

Hello > Is there a place or book where a newbie can gather information on > indexes? Section 12.4 of the manual is your friend. "EXPLAIN SELECT foo" will tell you what indexes are used in doing your "foo" query. cheers -- TS ---

Re: Slow Select count(*) - Second Post

Hello > Is there any way to re-write SQL to force the optimiser to use a certain > index? SELECT foo FROM bar USE INDEX baz WHERE ... See the manual at . cheers -- TS - Before posting,

Re: TEXT's are giving us headaches!

Hello > We use MySQL 3.23.32, so I recently added a FULLTEXT index for this > field on all forums and changed the source code to make use of the > MATCH() AGAINST() MySQL call. If you make FULLTEXT queries over many tables, and combine the results afterwards, you might get into problems with 3.2

FULLTEXT mismatch for tables containing 1 row only

Hello Another possibly related FULLTEXT problem: in 3.23.32, if the table contains one row only, then a fulltext match occurs regardless of the pattern you search for: DROP TABLE IF EXISTS t1; CREATE TABLE t1 ( id mediumint unsigned NOT NULL auto_increment, tag char(6) NOT NULL d

Re: 3.23.31 gives FULLTEXT problems on sparc64 Linux

Hello > > The error says: "Can't read indexpage from filepos: -1". [...] > > Thank you for reporting this. This will be fixed in a next release. > But it was really not a big problem - the table was not really > corrupted. [...] Thanks for the fix and the explanation! cheers -- TS -

Re: Increasing queries speed?

Hello > I wondered if there was a way to increase the speed of the query and > bring it to 2 or 3 seconds? Have you tried optimization hints from the chapter 12 of the manual? cheers -- TS ---

Re: isamchk and myisamchk inconsistently report errors

Hello > I assume simply doing a check is a read only operation so how is > this possible if the data file is not changing? The manual says: "in some case CHECK TABLE will change the table! This happens if the table is marked as 'corrupted' or 'not closed properly' but CHECK TABLE didn't find an

Re: 3.23.31 gives FULLTEXT problems on sparc64 Linux

Hello I wrote: > > After upgrading to 3.23.31, FULLTEXT selects give often "error -1 > > from table handler" for me. [...] Sergei Golubchik <[EMAIL PROTECTED]> wrote: > Thanks for a bug report! This would be fixed in a nex release. Thanks. In 3.23.32 the test case I submitted apparently wo

3.23.31 gives FULLTEXT problems on sparc64 Linux

Hello After upgrading to 3.23.31, FULLTEXT selects give often "error -1 from table handler" for me. The error log says "ft_read_first: Got error -1 when reading table foo". I think I have not seen those for older 3.23.x versions (x=28 or so), at least not that frequently. A small example to re

Re: Searching in different tables

Hello > I need to be able to search (WHERE column LIKE 'search') in heading, > body and/or author in all three tables, ORDER BY the latest date, > and LIMIT 20. MERGE table type is your friend, see section 8.2 of the manual. Another possibility is to use a temporary table (CREATE TEMPORARY TABL

Re: bugs

Hello > INSERT INTO new_table (campo1,campo2) SELECT (campo1,campo2) FROM old_table ^ ^ The 2nd set of parentheses should not be there. > Mysql dont suport sub-querys yet, we will be waiting It is not a subquery issue, see the section 7