will it make a problem if i move ibdata1 to a new location then adjust mycnf so that ibdata1 will become of a
fucking mysql 'syntax' error
0') to pastebin, and under it include the new
lucox, and include the describe pun_0948_logupdates;
lucox, http://pastebin.com/d35391955 try this
this cannot work …
lucox, why?
because “ are used for marking constraint names not values
hello, is it possible to create a column with a boolean(TRUE/FALSE) value?
and always the same error
well, put the back ticks around the column names and leave the values as they were
lucox, how is it going?
what do you ask? nothing new …
lucox, ok I thought you wanted my help
No, however you can create a TINYINT(1) which will do the same thing however the value will be stored the same size as a TINYINT
Which isn't a major problem
I must have been mistaken…
yes i want
http://pastebin.com/d32ed6ac8
lucox, try this
the same
lucox, what version of mysql hosting are you using?
_Lemon_, ok thanks but as there are no TRUE/FALSE things how do i make it display like TRUE and FALSE checkbox in gtk2:impleList?
5.0.44
TRUE is just a non-zero value where FALSE is zero.
_Lemon_, ah ok
With very beefy hardware, and lots of optimisation, is a MySQL cluster suitable to host an MMORPG backend db?
Or would I be better with something else, if so, what?
I'll be caching data in memory so to reduce the amount of queries
lucox, try this http://pastebin.com/d337b10c
i think delimiter is only for mysql shell, and i have tried it several times
lucox, just humour me, please
always the same
lucox, I am sorry, but I have run out of ideas
me too 
lucox, you will have to ask someone more knowledgeable than I am
thanx anyaway
lucox, no probs
people any other help ?
now, has anyone got any idea about my problem. http://pastebin.ca/662122 when it is run at the MySQL 4.1.20 CLI, the created_at_IP column is created as a VARCHAR(15) rather than a CHAR(15). Does anyone know why, please?
If I was to do it, I'd probably use it to substitute hard disk storage (instead of any sort of flat files) that way, your program doesn't have to worry about being fault tolerant, so long as it updates the database - but I have no experience on this sort of level
For a few hundred players, I was thinking something like two replicating clusters of 4 servers.
replicating for zero downtime, switch to the other instantly on the other one having a problem.
Lots of HD writes/reads: BAD
hey
i have a highloaded mysql, the workload is 17++, but it only uses 200meg ram, it has 2GB ram total and is the server is dedicated to mysql
Yes but you won't be able to load from the database as the data is requested, yuo're gonna have to bring a load of it into memory preemptively
how can i speed the things up+???
So what I was suggesting was making sure that when the memory gets edited, updating the database asap!
edit configuration file?!??
PLEASE?!
_Lemon_, thanks a lot i've implemented it(in my perl project) and it works
my my.cnf http://pastebin.ca/662167
Awesome
_Lemon_, what's wrong with loading all basic char data on character select? The client will have to load it's local stuff after that anyway, giving the server a chance to do it's query and prepare it's reponse, in time to send it back for the loading
completion.
How fast is a simple select query that returns 1 row on a very beefy cluster?
Well, if you mean it loads during a client loading scene, then that's what I meant by pre-emptive, and is how it should be done (I assume)
I have a variant of http://blog.evanweaver.com/files/mysql/my.cnf.innodb_1024
It'll load upto 512Mb if I remmeber correctly, you should see what it does differently and tweak yours appropriately
http://blog.evanweaver.com/articles/2007/04/30/top-secret-tuned-mysql-configurations-for-rails
(I don't use rails)
OK, guys, I have worked it out. Apparently, if you have any VARCHAR columns in a table, then it will silently convert any CHAR columns to VARCHAR when you create the table. Do CHAR and VARCHAR CANNOT co-exist within the same table!! That's crazy!
It's not crazy, char is a fixed length which gives you static tables, while varchar gives you dynamic tables.
i am making a perl application acessing a mysql database and there is a TRUE/FALSE column: 0 is false and the rest is true…but i've not implemented yet the NULL entries so i've put the following line: if ($name and $mail and $problem and $resolution)
=that insert the values into the database and else {print "NULL entries are not programmed yet\n"} but if i want to insert 0 in the database it thinks that $res
olution isn't defined and it tells me that null isn't programmed yet…is there a trick to differentiate a NULL entry from 0?
NULL is zero
is not null?
oh differntiate
However mysql hosting differentiates from this and interpretes NULL as "absent" instead of zero
Hello I need someone to help me with a mysql problem…paid service
payment via paypal or ..anything
simple stuff 
any free lance around ?
State your problem and if someone has a clue, they might help… dun dun dunnn
well my problem is ..I need to reset the mysql root password. I found all in the manual but I cannot reset it ..so I would need someone to login and fix it for me an I pay
ygurugeek, aren't you from china?
Yeah he is.
good luck collecting
xD
Paypal.cn :3
nope
..CH
Switzerland
does it sound as a more..trustable country to you ?
!man reset password
ummm, you shouldn't be having any problems and need to pay - what's the specific problem?
Sorry - I have no idea what function you're talking about! but try http://dev.mysql.com/reset password
!man reset
see http://dev.mysql.com/doc/refman/5.0/en/reset.html
I did that
!man skip grant tables
see http://dev.mysql.com/doc/refman/5.0/en/server-system-variables.html
it just doesn't work …because
is on a mac mini box
some commands are different
e.g. doesn't even have mysql console
and by the time I figure it out..I'd rather pay someone pro to do it
ain't a Mysql expert
I remember doing this before, it was about startup commands than anything else…
yes
skipping grants table
but
You have to shut it down, then tell it to not load the user tables then you get teh defaults
for me doesn't work spent too many hours on it so I'd rather give access to soemone to do it
yes
been there done that
I did
I am probably too dumb or did something wrong
this is why
I decided to offer $$ for a fix
You've got more money than sense tbph.
Then, with mysql root access you can change the grant tables yourself
yes I do …and more money than time
I need to reset
the root password !
so I can't login and change anything
weird enough all worked last night
then …
most people helping out here are here because they want to learn something new, not earn a few bob for helping random people
at a restart looks like an old password took over
okay but I think someone might be interested in helping for money….so I try my luck
Hello.
Shut down mysql. Then start it up with "mysqld_sage –skip-grant-tables –user=root" and it should load without any users/pass
UPDATE mysql.user SET Password=PASSWORD('new_password') WHERE User='root'
Your password should be 'password' (no quotes)
dotgeek:~ davidcosta$ sudo /usr/local/mysql/bin/safe_mysqld –skip-grant-tables –user=root
Starting mysqld daemon with databases from /usr/local/mysql/data
dotgeek:~ davidcosta$ sudo /usr/local/mysql/bin/safe_mysqld –skip-grant-tables –user=root
Starting mysqld daemon with databases from /usr/local/mysql/data
it hands
I tried this
also tried….
with the init file as per manual
when I pass the login command it just hangs (the skip grant)
if you fix it
Just wondering. What would be the best table structure for a directory hierarchy?
would be much faster
sudo /usr/local/mysql/bin/safe_mysqld –skip-grant-tables –user=root &
Think all the way through and finish writing before pressing the enter button.
dotgeek:~/Sites/newgeek/vanilla davidcosta$ sudo /usr/local/mysql/bin/safe_mysqld –skip-grant-tables –user=root &
then it hang
table directory with: `directory_id`, `name` then a table file: `file_id`, `directory_id`, `name` ?
[1] 3226
hangs
dotgeek:~/Sites/newgeek/vanilla davidcosta$ Starting mysqld daemon with databases from /usr/local/mysql/data
wait
didn't hang
dotgeek:~/Sites/newgeek/vanilla davidcosta$
…. it wasn't hanging, just running
but didn't do much either hehe
yes
It's now running in the background
but then how can I pass the comman ?
ah
okay
and ? :X
But a directory containing a directory?
Now execute your statement
if I type mysql
it tells me command not found
so how do I execute it ( Thanks a lot for your help)
Table directory: `directory_id`, `parent_directory_id` (NULL), `name`
mysqladmin
Or… can't remember the program
okay I know how to run mysqladmin
but how do I run a statement there?
Just a place that will connect and send a query
Wouldn't that be inefficien
t?
Not really, you could probably remove the NULL part as well, as all directories will have a parent directory except for root
id 0 for root?
Yeah, should work
ygurugeek, absolutely can't remember - surely if you're trying to run mysql, you're expecting to run a query at some point?! Do it with that application
Are you sure that wouldn't come out bad in performance?
Otherwise install apache2/php5/phpmyadmin and do it that way
I'm not sure what you're planning to do with it… I can't see anything instantly wrong with it… there is no redundant data?
It's a quiet long story, not too long though.
Well, I'll assume you want to display some sort of directory tree to the user
I feel that you could fix it in no time..I feel that the force is strong in you
bI feel that you could fix it in no time..I feel that the force is strong in you/b
Partially.
Then look up binary trees, it's a common computing concept, try implementing a derivative of that somehow
Binary trees?
and unless you're containing 100,000 directories, I can't see anything going desparately wrong
I LOVE those
Binary trees look, in memory, like a directory tree except each node has only two links, to other nodes (which could be the last)
You've got mysql running with one user as root with no password, all you need to do is login and execute the statement I showed you earlier
I will try again
Your English isn't bad so you could easily use the documentation to finish it off
but my client says mysql is not running
Type ps -A
If you see it in the list then it is running
I'm developing a image uploading system, and I thought it would be a good idea allowing the users to sort their images by folders too, so the URL would become http://www.example.org/images/username/folder/subfolder/image.png
Otherwise you'll have to check the mysql error logs
Or something..
I get you
What you have there is very similar to a forum with boards (directories) within boards and then boards containing topics (files)
0.01 sudo /usr/local/mysql/bin/safe_mysqld –skip-grant-tables
Like Sub-category?
0.00 ps
I implemented that in a very similar way to the tables I suggested before - it works perfectly fine, all queries are doing great too
Ok, thank you for your time.
That means it's running, you're client isn't connecting to it properly then
No problem
via ssh how can I login then?
I mean to execute the query
"mysql –user=root"
I got an idea
Then you can type your statement
is there any book you like ?
Then type \g to execute it
A book? Naaah, I don't need payment :p
mysql works
let me see
You don't realise how close you are to finishing this
can you give me the update command ?
USE mysql; SELECT * FROM user;
UPDATE mysql.user SET Password=PASSWORD('new_password') WHERE User='root'
dotgeek:~ davidcosta$ UPDATE mysql.user SET Password=PASSWORD('1234') WHERE User='root'
-bash: syntax error near unexpected token `('
You still have to be logged into mysql
yes sorry :X
When you type "mysql –user=root" it should bring up a new thingy and look like mysql
ok done
now it just
says
:P
after the statement
safe_mysqld is the actual mysql server, dealing with giving you the data. mysql is the client that connects to the server
Type \G
should I try to connect now ?
You _are_ connected
I mean
with the application which needs to run
Did you type \G
(or was it \g?)
That will send the query you typed out
\G
Did it give you a confirmation or anything??
nope
just
hmmm
UPDATE mysql.user SET Password=PASSWORD('new_password') WHERE User='root'\g
Put that into it…
got a new line thing
;
semicolon
You're 100% positive? Nothing like:
Query OK, 1 row affected (0.00 sec)
1 Changed: 1 Warnings: 0
Write a semicolon there.
semicolon doesn't seem to be a problem… and it should generate an error :s
Write it.
In that new line thing you got.
do you have gtalk ?
is for an open source site
formerly dotgeek.org
PHP forums etc
so
ain't really doing it for bucks either !
I feel is something deeper because no one changed the root pw
and yesterday worked
then I updated the script..the php script
and didn't connect anymore to the server
That's odd.
will engrave your name on the new forum
yes is really like…may be there are 2 mysql running or
don't know how the password changed
Do you have a backup?
Then you've probably changed the password the script uses to connect to mysql with…
becase all was fine ..then all of a sudden..bom can't login anymore via admin etc
no
I didn't
ah
100% positive not
ain't so stupid lol even if is hard to believe
Where was this php script from? Was it malicious?
Do you have a backuo of your database?
no malicious
was an extension for a foum
I have all the db backup etc
Yeah but it still could have contained bad code if the source was a bit dodgy
If I were you I would remove MySQL completely from the server and set it up again.
sure but
I don't htink bad code
can change mysql root pw !
I'm assuming your php site was connecting with the root user?
yes
was using root user..because we never used mysql on this server so
yes
:X
bad practice
Terrible practice.
When this is sorted, create a new user account without grant access
yes
I would remove the MySQL completely from the server and put it up there again with the backup.
may be I was thinking to erase mysql and re install
Then even if that was bad code in the forum, it won't be able to mess with the mysql user tables :s
Didn't you just say you changed the root pw?
apparently not
cauz I can't connect
wouldn't you ssh ?
I am begging you
You have to restart mysql without the –skip-grant-tables
You don't seem to respond.
I do respond
I tried it all and foesn't work lol
ygurugeek:
there shouldn't be a reason to reinstall, although a backup should be last resort…
I think I am omitting something
may be something stupid
hi! can anybody tell me whats wrong with that statement?
CREATE TABLE `association` (`idassociation` int(10) unsigned NOT NULL auto_increment, `name` varchar(30) NOT NULL default 'asdf', PRIMARY KEY (`idassociation`)) ENGINE=MyISAM DEFAULT CHARSET=utf8 AUTO_INCREMENT;
_Lemon would you ssh for me ? wouldn't take you much
I use the methods I know to work, instead of inventing others methods with trial and error.
for that I didn't need to come here
reinstall is easy no ?
ha, suppose but it's overkill for this problem
pm me the ssh login then
Yes, it is easy.
what's the problem with it? The error?
#1064 - You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near '' at line 1
I did pm you but may be I am in your ignore list :X
_Lemon_:
:P
I… am?
You… are even?
He's not registered at freenode.
I am but with another nick
Non registered users can't PM you unless you allow it.
now should work
CREATE TABLE `association` (`idassociation` INT UNSIGNED NOT NULL AUTO_INCREMENT PRIMARY KEY, `name` VARCHAR(30) NOT NULL DEFAULT 'asdf') ENGINE = myisam;
?
the problem is was the AUTO_INCREMENT at the end, thx!
np
hello i am doing a current SQL call being http://www.pastebin.ca/662221 but the problem is that group_categories and categories has mutliple match results but it only grabs the first one, i was wondering how i can grab all the results for the
categories
anyone know, its been bugging me for ages
hi people.. anyone there?
hi
What's the SQL command to view what users are added to mysql (and maybe show the privileges, if possible)?
as root
Is it possible to push back the auto-incrementing counter of a table?
I had 100 rows, and deleted 97 of them.
So now it has 1, 2, 3, 86.
I want it to be 1, 2, 3, 4.
ALTER TABLE thetable AUTO_INCREMENT = 5;
and change the 86 to 4
OPTIMIZE TABLE table_name;
oh, I didn't know that existed - that's pretty awesome
hey, i am running a rather big site, but the mysql is really slow how can i find out what bottlenecks the mysql?
it does not use all ram and workload is 17
or server load*
what's a prefix index in MySQL?
if you had a varchar(10) and you would only use the first 4 chars as the index?
hi, is it possible to change mysql db name ?
!man rename
see http://dev.mysql.com/doc/refman/5.0/en/rename-table.html
ummmm
http://dev.mysql.com/doc/refman/5.0/en/rename-database.html
!man rename database
Sorry - I have no idea what function you're talking about! but try http://dev.mysql.com/rename database
I wonder if the bot can get to it
http://dev.mysql.com/doc/refman/5.1/en/rename-database.html
it was added in 5.1.7
"RENAME DATABASE destroys any stored routines or events associated with the original database."
Ouch
rename_database–
if you had a varchar(10) and you would only use the first 4 chars as the index?
is that a prefix index in MySQL?
mmm..
it just create a new db
move table over
drop the old db
it makes sense, then
mysqldump old_db | mysql new_db
hrmm…
what would you gyys say the best way to do replication is?
(ground ball to second base)
haha
http://dev.mysql.com/doc/refman/5.0/en/replication.html
mm..it's not in ubuntu repo yet, 5.1.7
still beta, isn't it?
that's what topic claims.
is the midpoint insertion trategy and LRU used for only MyISAM … or is this for InnoDB as well?
Question. I have the command "SELECT `a`,`b`,`c` FROM `example`;", but say that I want to return those three fields but only when `a` is DISTINCT, how would I do that?
DISTINCT applies to the whole ROW
use GROUP BY `a`
HAVING count(0) = 1
thumbs, GROUP BY was causing certain entries to be ignored when I used it in conjunction with ORDER BY.
that numerical value in quotes refering to a column position in the table?
parens*
it's your only option
or chadmaynard's
count(0) = count(*)
however, I suspect grouping might be faster
any benefit from using count(0)?
I would be placing count(0) after SELECT without listing any of the columns that I want?
i meant GROUP BY a HAVING count(whatever) = 1; - returns a's that are distinct.
no, HAVING goes at the end
ah. Neat trick
so it looks like your only option is grouping.
I'm sure there are no real performance diffs
I'm pretty sure mysql substitutes count(*) to count(0) internally, though
lol
so, in my case, SELECT `a`,`b`,`c` FROM `example` GROUP BY `a` HAVING count(`a`)=1; ?
tias
how could I calculate the size of my non-leaf indexes for one table?
Try it and see, its quicker to type it on your system and try it than wait for one of us to tell you its ok
you ask questions like someone studying for a test.
lol
lol.
I've been reading -a lot-.
for a test?
threnody, I have an interview in two weeks
'what is the meaning of life?' 'how much does god weigh?'
42. No one knows.
so… how could I calculate the size of my non-leaf indexes for one table?
We are changing MySQL server hosts from one computer to another, and replication is involved. Old master (id=1) is still being seen as the replication source by slave (id=3). I've edited the my.cnf on slave to look to new master; what else keeps it looking to the old master?
I'd like to do something like "LOAD INDEX INTO CACHE t1, t2 IGNORE LEAVES;" … but I don't know how big my indexes would be, so I'm trying to figure that out.
Just spotted "CHANGE MASTER …" command in the docs.
I switched to using GROUP BY, but it's ignoring my ORDER BY, instead sorting first by GROUP, and then by ORDER. For instance, I have SELECT * FROM `example` GROUP BY `foo` ORDER BY `bar` DESC, it would return `foo`=2, `bar`=1 before `foo`=1, `bar`=2
I probably did not explain that well, because I'm not certain what is actually causing the problem, I'm sorry.
GROUP BY does not cause the sql engine to ignore ORDER BY
errr
^^
presumably, another factor is mangling your results
I have two columns, one that I'm trying to isolate from duplicates, sorted by the second.
except it needs to sort the second before it starts cutting the duplicates,and it seems to be doing it the other way around.
My replication slave still insists on trying to connect to the old master. CHANGE MASTER did not shift its attention to the new master. What else needs to happen?
if you're doing a group by, you are going to get one value for the associated columns. You can choose which value by using another aggregate function, like max() or min().
threnody, thus why I wanted to use DISTINCT over GROUP BY.
you can't use DISTINCT. Forget about it.
prehaps you want more like '…group by a desc, b asc'
threnody, I can try something like that lemme do that.
if you could tell us in english what results you want, it might help.
I'm a young, dumb American, that might be difficult!
I wonder why you'd admit that fact
i have a very similar query and ordering doesn't work for me either
here's bad news - I'm an old dumb America. It doesn't get better.
group by ordering is a real issue.
american has n on the end. moron.
Well, let me rephase what I want. I have two columns, I want to sort by the second column, returning only unique results from the first column.
you're not the only one. http://forums.mysql.com/read.php?20,134952,134952
if you are only returning a where the count() = 1 then there should only be one b value associated with the group so max() = min() so ordering should work
Well, juding the book by the cover and just reading the title, I can tell that this problem looks the same as mine. I'll give it a read.
Got it — had to add a "master_host=" clause to the "change master" command.
Oh man, his final solution is not easy to extrapolate into my mind.
mydb.sql
when i give this i still get the password prompt
i am giving the password in the command line but it does not take it as an argument it still displays the password prompt
take the space out
-pxxx not -p xxx
in a bash script this works- '…-uroot -p'pwd'…'
chadmaynard!
seekwill!
oh, get a room!
Do you play DTD?
no
We can play multiplayer DTD
thanks
didn't you read?
Playing strip pokim?
ha
I could have predicted that!
hehe
: (
Well, there is PRIMARY and UNIQUE, but what is INDEX for?
to create an ordered set of data that the dbms can search through faster than the unordered set
For example?
um….
So, I'll set everything to INDEX. Wouldn't that be a nice performance boost?
maybe maybe not. Depends on the overhead of the index and if it is useful
Overhead?
is there a way to cap the disk usage a mysql user can use?
Yes.
i'm having a problem with a user taking up all the disk usage on a disk (accidentally)
how?
Well, changing the source always work. But I'm sure there is a better way, and unfortunately, I don't know how.
im trying to create a database to hold details on furniture ranges. i have a table with a list of ranges, i then have a table with a list of products (also has a range id to link it to ranges) the problem i have is each product can have anything from 1 to 40 sub products that i need to record
the dimensions of. has anyone any ideas as to how i could layout this table
set quotas
don't be silly
Could you explain it again with the INDEX thing? I didn't quiet understand.
I'm always silly
what don't you understand Polarina?
The thing it does…
Is it unique?
no
Can it be both?
UNIQUE is INDEX, but with the unique property added
what's a prefix index?
you'd have to set up this user databases physical filesystem files under O.S. cota control.
quota
If I set something to UNIQUE, it is also INDEX. Right?
technically it is just UNIQUE. but yes, UNIQUE indexes are indexed
Ok, thank you.
there is not point in doing UNIQUE(name), INDEX(name). that'd be redundant
anyone?
Ok.
I think I've found the solution, it's in the manual: http://dev.mysql.com/doc/refman/5.0/en/example-maximum-column-group-row.html. Is this what you were looking for?
anyone any ideas about my problem??
Hi all, I had to do a quicj change from server to server so i simply copied all my /var/db/mysql across, How can I make this run on the new server? I am not quite sure how to search for this so if i missed somthing please point in the right direction
Thanks
the files are all MYD and MYI and frm and similar
put it in the datadir and chown to mysql user
will i chown it to the name of the mysql user or to the actual mysql user account?
whatever mysqld runs as
ok thank you
Thank you so much chadmaynard you just saved me from a very angry boss
np
to be fair
chadmaynard++
chadmaynard–
oh, thats a bit annoying
you mean set a quota for there home directory, and change the path of the mysql data to go in there home directory?
I don't think there are bte-control controls under mysql
you could put a trigger to fail insert past a certain number of rows… but that would be work aroundable, I think, and even more annoying
yes.
that is what I meant
no path changing needed, you might just set upa soft link.
hmm
will it work with a softlink?
im not sure it will
dmb_ it should.
by softlink, you mean symlink right?
yes…a soft symlink, as opposed to a hard symlink
the later would nto work accorss filesystems, and myeb would not be subject to teh quota settings.
I willhave to leave now. (and I don't bother switching the nick very much) …
best wishes there!
bbl.
ok
afternoon
I'm looking for an application like pgadminIII but for MySQL… I know about the web-based ones, but any Qt/KDE ones?
gui
GUI tools can be found at http://dev.mysql.com/downloads/gui-tools/5.0.html phpMyAdmin at http://www.phpmyadmin.net/ and even navicat or http://www.webyog.com/en/
thanks
how do i select the latest record added from a table?
is select * from products order by max(timestamp) valid?
Any suggestions on what type of query I should run to get "all birthdays within in the next week" on a DATE column format?
Hi I'm tyring to Query the last 5 selection of a table, any idea on how to achieve that ?
will it make a problem if i move ibdata1 to a new location, then adjust my.cnf so that ibdata1 will become of a fixed size and move it to a new location (and autoextendable ibdata2 will be added in old location)? do i need to take care of ib_logfile's or ib_arch_log in some special way?
what would make mysql still be up, just not take anymore incoming connections? Nothing in the error log about it
Evil things like spiderpig?
I don't know, honest.
anything I can do to find out more about it?
Restart the MySQL server?
that fixes it, yes
but this happens a fwe times a day
Having it restart as a cron job would work.
yes, but no :p
What version are you using?
4.2 iirc
Upgrade, for gods sake!
easier said then done
What's the problem?
There is no 4.2…
sorry
4.1.14
Even worse!
besides, even with backups, doing database upgrades worries me
When it does not accept anymore connections, is the system still responsive?
yep
and existing connections continue to work fine
Installing a never version isn't going to be a permanent decission. You can always go back.
Log in as root, and do a SHOW PROCESSLIST. I bet you have sleeping connections. Those should be disconnected.
crap, just restarted
but next time I'll give it a shot
What connects to the database? PHP?
PHP, python, and a few C++ apps
Make sure you are not using pconnects (especially in PHP)
Hello
I need a help with creating 2 unique columns !!
yea, I guess I could go back to the old version if it fucks up
hi
I am thinking to use a bash script to query my database. I have seen there is shsql shell but I need to be root to install it
echo query | mysql
jbalint but I need to store the return data
I am trying to do something like $RESULT=mysql -u root -ppass database /tmp/query
ok
hey, how would i do something like "SELECT id,COUNT(SELECT something FROM table) FROM othertable"?
is that possible?
why are you trying to do it that way?
why not just do a join?
its a more advanced query of course, but i need something like that
to put a select inside of a count, that possible?
sorry, don't know, doesn't really make sense though
count(column)
not count(query)
yeah i was just wondering if it was possible to put a query in there
guess not
how exactly does JOIN work? i guess understanding that could help :-/
based on data from one table, you join another table with relationships
foreign keys and such
so like
table 1 has "mem_id", table 2 has "comment_mem_id"
Not really foreign keys
SELECT * FROM photos JOIN album_id on album? or something?
sorta
select mem_username FROM members, comments WHERE mem_id=comment_mem_id
or select mem_username FROM members INNER JOIN comments ON mem_id=comment_mem_id
same query
hrrrmmm
joins
http://hashmysql.org/index.php?title=Introduction_to_Joins - For more indepth info: http://dev.mysql.com/tech-resources/articles/mysql-db-design-ch5.pdf
thx
i can't figure out how to get the birthdays from a DATE() column. like, all the birthdays in the weeks 1 week
in the next* 1 week
DATE_FORMAT(minfo_birthday,'%d %m')='25 08' AND DATE_FORMAT(minfo_birthday,'%d %m')='18 08')
but that fails
ah, figured it out
%m%d instead of %d%m
what are the multimodal databases?
in case someone gives a damn BENCHMARK(1000000,ENCODE("hello","goodbye")) in debian mysql 5.x amd64 takes 0.41s to complete when using ondemand as cpu governor and 0.36s when using performance as cpu governor
using Q6600 cpu and ddr2 6400 ram
what is ondemand
cpu cooling feature, called "governor" in linux kernel
ondemand means that the cpu will work at it lowest freq until there will be some load and then it will throttle up
in my case using Q6600 2.4ghz cpu that means it will idle at 1.6ghz
but step up to 2.4ghz during crunching
ah
so the diff (0.36s vs 0.41s) is most likely the delay of detect higher load and then delay of stepping up the cpu into 2.4ghz speed
like amd's cool n quiet
using same technique in the linux kernel
I have 2 columns, I want to ORDER BY the 2nd column and then GROUP BY the 1st column. How can I ORDER BY before GROUP BY?
can someone explain why my query always returns ,1 ?
$query = "SELECT `notices_read` FROM `ibf_member_extra` WHERE `id` = ". $id;
there is only 1 row in the database, and the notices_read field is empty
running the query directly through mysql returns nothing, but PHP always sets the value to ,1
TheWikies, select a from (select * from foo order by b) as foo group by a
uTheWikies, select a from (select * from foo order by b) as foo group by a/u
TheWikies, the question is why do you want to do that?
I'm creating my own forum, and so I have a column which is the message ID and the second column is the date.
hwy
i mean hey
the message ID's go 1, 2, 3, etc. I want to sort by the date to display the newest posts, but not to show multiple posts of the same topic.;
ie, not to return multiple message ID's.
if i am selecting addresses from a very simple address database, and I want to select say 20, and I want to make sure all of my selection have different states, how would I do that?
DISTINCT nevermind
I guess I can just create another table, have the topics stored in one, and the messages stored in another. It just seems like I could have combined the two.
that's the way to do it TheWikies
using mysql cluster, I have a table that claims not to exist when i use describe, drop table, etc., but claims to exist when I try to create it
http://www.pastebin.ca/662522
5 [ERROR] NDB: failed to setup table nfackler.nuke_groups, error: 723, No such table
any ideas?
no files exist in the datadir for this table
any ideas what may complains about incorrect information in .frm files mean after moving ibdata1 to a different location and adjusting my.cnf accordingly?
hey, I'm trying to add a foreign key, the alter table doesn't give me any errors but the foreign key is not enforced… why?
Kronuz, myisam?
hmm
how can I tell?
probably not
can I convert the tables?
is mysisam slower?
or faster?
Kronuz, i think alter table can to that. myisam gives you the wrong answer faster
SHOW CREATE TABLE
nice
I used describe
hmm
there are no fulltext indexes in InnoDB ?
Correct
You can use an external fulltext engine though
hmm
So I have a bunch of dates in a table, and I want to plot them on a chart. I have too many to have an individual plot point per entry, so I want to group them by some section of time and count how many are in each section and plot that count. Am I on the right track?
yes
So the time frame could be months, or years, but I know I want let's say 40 data points… So maybe I should figure out how many seconds or something is in between my first and last date and divide and then group on that some how
Sembiance, sounds good
Now to actually write the SQL
Hrm.. so take first and last date. Subtract first from last. Divide the difference by the number of segments you want, then do a group on the rows date minus the first date dividide by the result of the difference/segmentCount calc.
Seems like that works
hi all, I have a lot of table-bin.xxx files in /var/lib/mysql, are they some kind of backup informations ? Can I remove the oldest ones safely ?
innodb binary log files, for use in replication and disaster recovery. search the manual for 'master purge' command.
thanks threnody , I didn't know where to look in the manual
http://dev.mysql.com/doc/refman/5.0/en/purge-master-logs.html
those puppies can fill a partition pretty quickly.
excellent, exactly what I was looking for
yep
so, as long as I keep them, I can go back in time and restore previous state of a database , right ?
yes
maybe I hould disable this feature completely then
is that possible ?
you can turn it off, or you can set it to automatically purge after a certain period.
I'd rather turn it off, because I'm already doing diff backups of the system with rdiff-backup
if you're faithfully backing up via some other method…
ok, so "SET SQL_LOG_BIN=0" should do the trick if I understand right
http://dev.mysql.com/doc/refman/5.0/en/binary-log.html
table1=(username, …); table2=(username,stuffA); table3=(username,stuffB) what I want is to create a report that says how many of StuffA and StuffB there are for any given user…
I'm having a hard time trying to visualize what the best way of doing that would be
like, if it was only one Stuff table, I could use group by after a left join… but with two tables…
the only thing I can think of is using nested selects
any other ideas?
might be a design issue.
users, possessions, stuffa, stuffb
??
I did a "PURGE BINARY BEFORE now()" , and yet the files are still there
an extra table linking users to the different stuff types.
threnody, what's the "possessions" for?
the manual says, 'PURGE MASTER LOGS BEFORE '2003-04-02 22:46:26';'
I did also with "MASTER"
threnody, do you think there's a way I can do that query without having to create that fourth table linking the users to the stuffs ?
I cannot, but perhaps another here can think of some multiple join that would accomplish it.
hmm… I can think of two ways of doing it with nested selects
I don't know if this is the place to ask, but with phpmyadmin, there's a "binary log" button, I don't what it's supposed to do, but when I click on it , mysqld start to run wild and do lot of hard disk access, but nothing appears
select table1.username, (select count(*) from table2 where table2.username=table1.username) as StuffA, … from table1;
select username, count(stuff_a), count(stuff_b) from (table1 left join (select username, stuffA AS stuff_a, '' AS stuff_b from table2) using (username)) left join (select username, '' AS stuff_a, StuffB AS stuff_b from table3) using (username);
(oh, and group by)
or rather a union, I think
in your stuff tables, one row per item?
threnody, what do you mean?
one stuff = row
but there are two kind of stuffs (and they both have other different "properties" or columns)
threnody, the tables mostly only have in common the username
well, I usually use linking tables for these sorts of situations, but it might not be for everyone.
(actually only have that in common)
hey, is SELECT … FROM a LEFT JOIN b LEFT JOIN c LEFT JOIN d; the same as SELECT … FROM ((a LEFT JOIN b) LEFT JOIN c) LEFT JOIN d ?
LEFT JOIN a, b, c, d ?
??
brackets confuse everything. Try not to use them.
but is it the same ?
join a table, then join another, then join another, then …
if I have a table with 2 primary keys, what happens if one of those keys is auto_increment?
and what about what hlms said?
ok, I can't remove the log files with purge, I don't understand why, instead I disallowed the log-bin option in my.cnf, so now is that ok to delete the files manually ?
there is no point in joining 4 tables at the same time
why? I suppose it depends
there is no point in joining 4 tables at the same time
besides, a JOIN joins TWO tables at the same time.
!man join
see http://dev.mysql.com/doc/refman/5.0/en/join.html
SELECT id, name, (SELECT COUNT(*) FROM bar WHERE fooid = foo.id) as numbars FROM foo; can that be done without subqueries?
it can
in mysql, if I'm managing files, should I just use the filesystem or store them with mysql
what's the consensus?
with the database I get the added benefit that permissions are easier to enforce
store file *contents* on the DB? don't…
select id, name, count(fooid) as numbers FROM foo inner join bar on fooid = foo.id group by id, name
^^
PovAddict, yeah
is it much much worse?
it's been a pain dealing with File calls in my rails app
storing files in the databased is a TERRIBLE idea
okay
that's the consensus
I will not use the BLOB
performance wise, it'll kill your app.
yeah
I'd better use the tried and true fs…
if I have a table with 2 primary keys, what happens if one of those keys is auto_increment?
I don't think you can have 2 PKs.
sure you can. its called a composite key
a PK made of two columns?
so, is that ok to delete the table-bin.xxxxxx files if I don't use the binary log features anymore ?
I mean manually, I can't make the "purge" command work
im wondering if the auto inc key will start at the default, then just keep going up no matter what.. or if it might start over if the other PK gets changed
Then you'd have a single, composite PK.
now that I think about it, it probably just counts up and up
hmm I'm on a similar question as d03boy
did you see my reply?
yes and thanks; but the PK thing is a different problem
ah.
I have a 'group' column, and an 'id' column; the id has to be actually "id within group"
add a WHERE clause, then
ya thats basically wha ti wanna know
i have a recipe (rid) and the steps within the recipe (step), both primary keys
where id in (…)
i'd like the step to count up starting where the last one left off just for that rid
same
same what?
same as me
I need the same as him the message number has to be "local" to each newsgroup
ok then.
do I need to code that myself instead of using auto_increment?
I'll keep out of it.
we both just have the same question, thast all
http://dev.mysql.com/doc/refman/5.0/en/example-auto-increment.html
\o/
which section?
see from the second example
"For MyISAM and BDB tables you can specify AUTO_INCREMENT on a secondary column in a multiple-column index."
hello how to restore only one table from bih mysqldump backup? i saw a solution with limiting privileges somwhere, but i cannot find it now
you could extract the statements related to that table, place them in a separate file, and restore that.
i have a question about mysql stored procedures.
i'm getting an error when i attempt to enter my header
DELIMITER //
CREATE PROCEDURE update_charge(
IN p_user_id,
IN p_task_id,
IN p_num_hours,
IN p_charged_date
)
BEGIN
DECLARE v_count_charges int;
DECLARE v_charge_id int;
oh god
SELECT count(*) INTO v_count_charges
PASTEBIN
try http://pastebin.ca or http://pastebin.mysql-es.org
FROM charge
WHERE user_id = p_user_id
AND task_id = p_task_id
AND charged_date = p_charged_date;
0 THEN
UPDATE charges
SET num_hours = p_num_hours,
WHERE user_id = p_user_id
it is 8GB file
AND task_id = p_task_id
AND charged_date = p_charged_date;
ELSE
stop pasting
INSERT INTO charges
(
user_id,
task_id,
num_hours,
try using a pastebin
charged_date
) VALUES (
p_user_id,
kick him already…
p_task_id,
grep it?
p_num_hours,
p_charged_date
);
END IF;
but hwo to edit then
END;
/
END!
sorry
kahotep, pastebin.ca
i meant to only post the header
he can't "stop" without disconnecting
use a text editor?
from his end, he sees he already pasted all
there are binary files
here we see one line per second…
if i have a simple table which has just two columns which are both the primary key, how can i do an insert into that table without risking being told that i have a duplicate primary key?
the dump is binary?
but ok, i will try to recall this from memory
basically, i want the 'duplicate' error to be swallowed
partially
PovAddict, i still dont see that example.. not sure if im lookin in the righ tplace
thats teh procedure
if the row exists already, good, i don't care - don't want an error though
i've got it 
http://hashmysql.org/index.php?title=Single_table_restore
here's one with the error on the bottom http://pastebin.ca/662611
ah.
but thanks
(this is for logging IPs against user id's)
http://dev.mysql.com/doc/refman/5.0/en/example-auto-increment.html
there's two boxes, one showing MySQL commands and one with the result, then three paragraphs, right?
oh, i see it, thanks
interesting stuff starts @ 3rd paragraph
cool
uh - forgot to ask
thanks for the alternative without subqueries but… which one is actually more efficient?
a JOIN is always more efficient.
but please, do benchmark it yourself
I was more asking by curiosity… I may have like a hundred rows
How could i increase @i value for every row? Update cats set position = @i WHERE parent_id = '$parent_id' order by position asc
a JOIN, especially a INNER JOIN on indexed columns, is always faster than a subquery
what idiot was just posting stored procedures in here
i hate those idiots
it was an accident. get over it
those need to be kicked at the 3rd line pasted; not banned, let them join back; but at least that way they stop
get a flood bot in here!
ghm
??
http://mysql.pastebin.com/m77affe2 - there is only one auto_increment column and it IS a key
so what's up with that error?
to make a composite key do I have to do PRIMARY KEY (col1, col2) or can I just put PRIMARY KEY in each of the columns lines
d03boy, the first
mmmmm
what wrong with this query? Update cats set position = (SELECT MAX(position) +1 ) WHERE parent_id = '$parent_id' order by position asc
it produces same positions
the inner select doesn't have FROM
you probably need:
Update cats set position = (SELECT MAX(position) +1 FROM Cats) WHERE parent_id = '$parent_id' order by position asc
any idea on MY error?
You can't specify target table 'cats' for update in FROM clause
ghmm
store the value from SELECT MAX(position) +1 FROM Cats in a variable, and use it in your next update statement instead
this should change to mssql channel
why?
cause i cant read. i thought it was when i joined
come again?
thought this was MS SQL Server channel
my mistake
thumbs, maybe you could tell how could i increase value of numeber (+1) for every row in query?
then i will not need select max value from database
Update cats set position = @i WHERE parent_id = '$parent_id' order by position asc
@i should be increased +1
position list should be reordered
http://mysql.pastebin.com/m77affe2 - any idea what's wrong?
the primary key must be id_INT only
actually.. that error message made me change my thinking.
i think the primary key requirement is gone…. and that query as you have it worked on my mysql server here.
InnoDB does not support that kind of PK
i'm using mysql server 5.0.32
You can't have the PK on ngroupid as well
how do you know he's trying to create an innodb table?
I'm smart
PovAddict is my brother
He's sitting in the room right next to me
Pass me a beer while you're up
is he your brother?
I think so
you think so?
Sometimes
thumbs knows why he's using InnoDB as well
Ask him
mysql isam has been all i've ever used
i don't even know why you'd want to use something else
I use isam, generally
I'm not really qualified to help anyone with innodb
ISAM is not even in 5.0
it's not?
Nope
i'm using mysql 5 and myisam though?
Yeah, MyISAM
But not ISAM
ISAM was disabled in 4.1 and completed out by 5.0
(Or I could be off on a version)
so you're calling me outdated, then?
Yeah, way outdated man
is ther a way to do a case insensitive order by ?
Kronuz, there is always the good old ORDER BY lower(foo)
hmm
a little on the slow side.
evening all
hi
I've looked for a bit and can't find what the best method to store a MAC address in to a data base is… with IP address it's inet_aton/inet_ntoa.
you could use a varchar
example?
varchar(length)
I guess I was thinking that there was a function similar to when IP addresses are used.
no to my knowledge
I'm guess I'm a little clueless of this "varchar(length)" method that you speak of.. I'm totally a DB noob.. learning as I go
!man varchar
see http://dev.mysql.com/doc/refman/5.0/en/storage-requirements.html
aren't Mac addresses just numbers like IP addresses? IP's are 4 bytes, not sure on the MAC's though
a mac address contains ':' and alphanumeric characters
therefore, a varchar is the most efficient way to store it.
But those alphanumeric are just hex, a representation of a number
and as Zero123 said in a pm, are only 6 bytes
! bigint or VARCHAR(6) sounds like the way to go to me…
that's what I was thinking
Though you'd have to parse it :/
eeh
What language are you in?
C
inet sounds like C
Yay! lol
I'm sure it's been done before, no need to reinvent the wheel…
exactly.. but can't find an example
Are bit shift operators C++ only?
You could just take each bit, shift it along or something clever
hmmm, yeah I'd probably spend a few hours on this, glad I'm not writing it - sorry :p
haha
hmmm … best way to store currency in a db? remove the decimal and convert to cents?
this is beginning to sounds like a math riddle.
how about using the DECIMAL type?
doesn't really matter to .. that's what inet_aton does.
that's fine
ive heard there are yucky issues that deal with rounding
Then use a FLOAT
yes, FLOAT has them. It's much less of an issue with DECIMAL
or NUMERIC
oh
FLOAT is LESS precise than NUMERIC
ahh interesting
for instance
NUMERIC (10,2)
Is this the problem intrinsic to computers and FLOAT calculations? If so, you could be having issues before you get to the database…
that would be ten digits, including the fractional value, with a two degree precision on the fractional part
good info, thanks thumbs
so you could do (20,8)
it depends on the precision you're seeking.
you think there would be a neat little function like ether_ntoa/ether_aton.. huh? go figure
there might be
I looked briefly… nothing pops up
year-month-date?
..
curdate() wont work :/
DATE_FORMAT ?
yea figured it
thanks
hello all!
You're enthusiastic.
anyone have any idea how to compile/install mod_auth_mysql?
a word of warning - ignore DARKGirl
i can't find libmysqlclient.so
heh
she's a little clue-impaired
Is it in the folder /usr/lib/ ?
and do you have read access?
yes i do
this is on osx, btw
so the folder structre might be different
Consider it found?
no, its not in there
Where is it then?
Is it installed?
You gotta find it then tell the compiler where it is, I assume…
yeah
i dont think its on my computer though
I did a locate of it
and came up with nothing
good luck
Then install it…?
not sure how
lol thanks thumbs though… 3am here for me so I'm oblivious to anything
ummm…
I'm warning you, you're wasting your time
but go ahead anyway
Does OSX have anything like apt-get install? or apt-cache search?
what are multimodal databases?
looks like i have to recompile it
because it doesnt include that file
for some ODD reason
ummm, that file is a pretty key file to connect to mysql
I really doubt you should have to…
but if it solves it - go ahead!
trust me, its not on osx mysql
ive done locate and everything
trust her, she has no idea what she doing.
"For the people maintaining the installation packages and gems ? has nobody ever noticed that this key libmysqlclient.so file is missing from the MacOSX? binary of MySQL?"
i installed from binary
lol
Shows how many people use MacOSX then :p
hello. theres any way in MySQL querys wherei can find only coincidences having the first character of the textfield
?
Hello
hi GoSox
GoSox, you can help me, please =?
unlikely
im here to GET help
hmn.. ok
=/
name,street,city,state . I'm trying to select say 20 rows at a time, but I want every state in my selection to be unique. How can I do this
select DISTINCT …. LIMIT 20
Error 'Can't create database 'sa_bayes'; database exists' on query. Default database: 'sa_bayes'. Query: 'create database sa_bayes'
Do anybody understand that? Im trying to use a master-master replication scheme
that was for you ^^
thumbs i tried that, but i want the whole row, with only the state column being distinct
then use grouping
"database exists"
how would grouping help?
select MIN(`a`), `state` from …. GROUP BY `state` LIMIT 20
it would return ditinct states
what is `a` in that query?
the other columns.
GoSox, why you cant use DISTINCT ?
he wants each state to only appear once, I guess
yes
then grouping is required.
i'm trying to get a somewhat even sampling, so i want no double states in each selection
to even it out
please, somebody can help me ?
then use my solution
cause in the table, there are TONS of cali and texas entries, and they are dominating my results
select MIN(`a`), `state` from …. GROUP BY `state` LIMIT 20
thumbs i don't understand your MIN(`a`) thing?
when grouping, you need to use an aggregate function on the other non-grouped fields
!man group
see http://dev.mysql.com/doc/refman/5.0/en/group-by-hidden-fields.html
so if `a` is not in the GROUP BY list, you need to apply functions such as MIN() or MAX() on them.
this applies for any column
i'm still not getting what the max or min stuff is for
pls, i will have problems if i cant solve this
you can't just use select a, b, c, d from … GROUP BY d
you need to apply functions to a, b, and c in that context
i just did and it looks like it worked believe it or not
SELECT state, name, city, id FROM chefmoz_list GROUP BY state ORDER BY RAND() LIMIT 20
you didn't group by state
you grouped on a function not using any of the selected columns
that's different
this will return multiple rows with the same state
I was under the impression that you only wanted one row per state.
i do
and thats what i'm getting with teh query i just showed you
your query is non-deterministic
but use it if you're happy with it
i have no idea what you are saying
my problem is very easy !, or not ? O.o
XP
suspiciously, all my cities are starting with the letter A
but using RAND() on a non-existing field, the resultset returned will vary from run to run
how do I untar a .tar file that is not tar.gz?
what is the command?
No mean?
its not 'tar' or gunzip
tar -xvf i believe
you might get several rows with the 'TX' state, for example
no the one thing it definitely is going is only selecting one of each state
SELECT state, name, city, id FROM chefmoz_list GROUP BY state ORDER BY RAND() LIMIT 20
doing*
that's different, now.
no thats what i send you the last time
note that this is syntaxly wrong.
you need to alter it this way:
can InnoDB have two primary keys?
SELECT state, MIN(name), MIN(city), MIN(id) FROM …
since you're grouping by STATE
and how will my results differ if i do that?
ln -s FULL-PATH-TO-MYSQL-VERSION-OS mysql
they will be deterministic
and what does that mean?
i.e., consistent across several runs
what is FULL-PATH-TO-MYSQL-OS
ok and what does THAT mean?
they won't change on each run.
but ummm i want them to change
i'm trying to select 20 random items out of a 120,000 item table
quick question — is it possible to run a second sql query within another sql query?
oh i see
the ORDER BY will do that
BUT
i know order by rand() is slow
thats ok
without properly GROUPING, you might get more than one row for a given state
there is no eay to tell ahead of time.
s/eay/way/
symbolic link
the solution is to apply a function, like MIN or MAX, to the columns not in the GROUP BY list
you should teach the_wench ''quick question' does not reduce your bandwidth footprint'.
heh?
quick question — is it possible to run a second sql query within another sql query?
ok i did your query, but i'm getting all cities that begin with 'A' still
like
its not being very random
that is one issue of grouping, you can't guarantee the ordering.
i need the selection to be totally random
'quick question', as in 'drop everything, this shouldn't take any time…'
there was an article that could have helped you, however.
oh hehe
ahaha
WHERE COUNT(*) = 1
this will improve your ordering.
ok i think i'm goona give up on this tonight, this isn't working and i don't have more time to get into it
thanks anyway
I hate when they drop like this.
can InnoDB have more then one primary keys? if so, is each primary key considered a clustered-index?
he is.
lol.
today is phase 2?
yes
I've been studying non-stop
so you're taking a course?
nope.
self-teaching myself.
'Why are protons green?' 'What is the mohr number of igneous sediment?'
lol
42, I don't know.
lol, good memory
wait, I gave you that answer last night too
Can you get some sort of information about replication in mysql?
like whats being replicated and when etc?
so what do you say? can InnoDB have more then one primary keys? if so, is each primary key considered a clustered-index?
how do I log in as the mysql user?
I may be a dummy, but having two PRIMARY KEY is like having two second basemen.
well
The notion of a Primary Key does not admit anything beyond one PK.
it's not possible under MSSQL, nor with mysql, AFAIK
I quote Xaprb
Hence "Primary".
You can mistakenly has a two column composite PK…
he is, I believe
have*
"And in tables where the primary key is several columns and it’s desirable to have the table “clustered two ways” by using the indexes as I’ve explained, not all of the columns need to be added to additional indexes."
wtf?
you can index as many fields/columns as you want.
no wonder your questions are so bizarre.
http://www.xaprb.com/blog/2006/07/04/how-to-exploit-mysql-index-optimizations/
well that question is either misleaded, wrong, or intentionally misleading.
the bottom line is that you can't have multiple primary indexes for a table.
s/indexes/keys/
I should not have said 'You can mistakenly has a two column composite PK…'
I should have said 'You can have a multi-column composite PK…'
but just one.
Cannot load /Library/Apache2/modules/libphp5.so into server: Library not loaded: /usr/local/mysql/lib/mysql/libmysqlclient.15.dylib\n Referenced from: /Library/Apache2/modules/libphp5.so\n Reason: image not found
i can see where the problem is, but I dont know how to fix it
you'll never manage to fix it.
consult a competent admin
its not supposed to be "/usr/local/mysql/lib/mysql/libmysqlclient.15.dylib\n". its supposed to be "/usr/local/mysql/lib/libmysqlclient.15.dylib\n"
how do I change the reference?
hmm
DARKGitl is like The Terminator.
she's damn annoying
DARKGirl is like The Terminator.
DARKGirl–
terminator?
there was this movie, ou might have heard of it…
indeed
your typing could improve.
WITHOUT A DOUBT
heheh
how am I like the terminator?
she's simple minded like him
oh, ask apache config questions in #mysql, and then repeat them endlessly when you get no response.
we ignored her in #apache a long time ago, too
(I had to llok down at the keyboard and hit each key like a terminator. I'm off to Best Buy tomorrow!)
oh thats an apache question sorry, I was reinstalling mysql so I thought to post it here
If you make a field not null is it not required to give it a default value?
