| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Comments:
<0> lets see if i understand this bot
<0> !m Asktool grant
<1> Asktool: (GRANT and REVOKE Syntax) : http://dev.mysql.com/doc/mysql/en/GRANT.html
<0> heh, cool
<2> blobaugh, yes, though I'm not sure what to search.... I'm guessing a special join :)
<3> !m emp__ union
<1> emp__: (UNION Syntax) : http://dev.mysql.com/doc/mysql/en/UNION.html
<3> See if that helps
<4> blobaugh: it's true,... I forgot grant
<3> Asktool, You should specify the user you are connecting as.
<3> Not sure if it matters...
<0> emp__: http://www.google.com/search?q=select+from+multiple+tables+mysql&start=0&ie=utf-8&oe=utf-8&client=firefox&rls=org.mozilla:en-US:unofficial
<3> Don't be mean
<0> well i got my problem partly figured out
<3> What's your problem?
<0> a local mysql db always seems to get corrupt once i upload it
<3> That would sukc.
<0> yup. that and i have a crappy webhost and they will not put anything newer than 4.1.9 or close on the server
<0> NEVER use westhost.com
<0> they told me i would have my own vps usermode style and lied. now im stuck with a crappy host i cant do much with
<0> i just wish i still had the emails to prove it to them to get my money back :S
<0> plus their stupid 'control panel' takes forever to do anything
<5> Okay, I must be stupid.
<5> I still can't see why msql -h 127.0.0.1 -P 3306 -p -u is failing with that error
<5> ... while mysql -p -u ... is working
<5> phpmyadmin is still working
<5> It'll let me in over the unix domain socket but it dies on TCP, and I know I don't any an iptables or ipchains and I can see netstat -nl ... showing me that the process really is listening.
<6> answerguy, does it work with your real IP?
<5> I've tried that, too. Same results
<6> ok, I'm out of ideas, since I don't use mysql much yet
<5> ERROR 2013 (HY000): Lost connection to MySQL server during query
<5> No syslog errors.
<6> are there any real applications to do mysql administration?
<5> The daemon is surely still running.
<6> I don't think I like my p***word being stored in the webspace
<6> and iirc, that's what phpmysqladmin did
<5> What is phpmyadmin doesn't store you p***word.
<5> It asks me for mine every time.
<5> (In my case I have it running on the same machine as my browser and my mysqld --- so it's all localhost
<6> I was hoping for something more like ms sql enterprise shell thing
<0> frb: huh? how do you expect to connect to mysql from any of your php sites if you dont save the p***word somewhere?
<0> try typing 'mysql' is that like it?
<6> blobaugh, having a user p***word stored in a site, and having the admin p***word stored is different
<4> Hey this was very helpfull for an SQL newbie:
<4> google: MySQL HowTo - amaroK Wiki
<6> I'm sure I'll fiigure it out if we choose to use mysql
<4> I'ts simple, specific, and pragmatic
<6> good night, and thanks for all the fish
<0> true, but still. as long as your server is setup properly a user will not be able to read any of the .php files except for what apache gives it
<7> hi, i have a specific question about mysql queries ...
<7> i have a database with a set of IP prefixes
<4> I would say to recomend that page for anyone with a proble like mine
<7> I want to perform fast longest prefix match on it
<4> thanks for all
<7> I have two options: 1) given an ip address, do someting like 'select prefix, mask from foo where (prefix = ip and mask = 0) or ( prefix = x&0xfffffffe and mask = 31) or ... (prefix = x &0x1000000 and mask = 1)
<7> 2) insert a field called prefix_max into the table for each prefix, so for something like 10.0.0.0/8, I'll have prefix_max = 10.255.255.255
<7> and the query will be select min(prefix_max - prefix) from foo where prefix<= x and prefix_max >= x
<7> which is likely to be faster?
<7> by 'x' in the queries, i mean the ip address i'm querying for
<7> oh, small mistake in option 1, it should be:
<7> I have two options: 1) given an ip address, do someting like 'select prefix, mask from foo where (prefix = ip and mask = 32) or ( prefix = x&0xfffffffe and mask = 31) or ... (prefix = x &0x1000000 and mask = 1)
<7> mask = 32, mask = 31 etc till mask = 1 (or 0)
<7> basically, my question is: is a min(simple expression) and two simple comparissons faster than a set of around 32 ORs, with comparisons/bitoperations within each clause?
<7> i need to be sure, because my professor thinks the latter is the `best' way :)
<8> arjenAU :-)
<9> lsmith: just replied. do better than coming up with bogus arguments. jeez
<8> i agree that you generally want short fixed length columns for your primary keys
<9> make that ints.
<9> having a char as a pk is generally silly. really.
<8> but inventing useless data that you are never going to query on is bogus
<9> no it's not.
<9> at least not as a general rule.
<8> i use surogate keys alot too
<9> I never bother debating "surrogate keys", I just always work with auto-incs where appropriate. at that level I don't worry about a few bytes storage, as I know that the overall picture is better and so performance will improve.
<8> but should they introduce integer keys for their "scopes", their "names"?
<9> lsmith: you need to zoom out. this was just a tiny snapshot of a huge mess.
<9> dl horde if you want and take a peek.
<8> heh .. well you were the one that zoomed in :-)
<8> and all i said is from what i see in your post .. i do not see the big problem
<9> indeed, because I really couldn't be bothered listing the whole nonsense. it would have confused ppl. this was clear enough.
<8> i do not think that this table is obviously utterly always wrong
<9> well, we disagree and so far I don't think you've got any supporters.
<10> I was here earlier with a problem, but I didn't find a solution, so my problem is:
<10> error: 'Can't connect to local MySQL server through socket '/tmp/mysql.sock' (2)
<9> !perror 2
<1> System error: 2 = No such file or directory
<8> so it goes
<9> redblade: netstat -l | grep mysql
<10> and?
<9> lsmith: I still think it's an interesting debate, and possibly a good research topic. does portability help or hinder, and what are the costs.
<8> i guess mysql is just missing a way to get to the internal ROWID and you are covering this up by sprinkling autoincrements ;-)
<9> redblade: that netstat will show where the socket is. if nowhere, then you don't have your server running ;-)
<10> unix 2 [ ACC ] STREAM LISTENING 10322 /var/run/mysqld/mysql
<9> lsmith: rowid is a hack of the same.
<9> lsmith: so not covering anything.
<8> i was just teasing .. sorry :-)
<9> lsmith: plus, you... portability fan... are now suggesting to use a highly vendor specific trick??
<8> as for the cost analysis
<8> might indeed be interesting .. but very difficult
<9> but possibly essential.
<9> and why difficult... you can spec a project both ways. or grab an existing project, spec it optimally using the info originally known at the start of the project, and then compare it with the current situation which was specced generically.
<9> what I also didn't mention is that most if not all cases of ppl moving away from mysql....
<8> well the question is how much of a bargain advantage you get for your licenses
<10> arjenAU, does the ^^^^ mean that it is running?
<8> by being portable
<9> the result was generally worse.
<8> etc
<9> redblade: yes.
<9> redblade: and it shows the path to the socket.
<10> so what might be wrong then
<10> how can I change it?
<9> redblade: don't be a monkey. you saw your errormsg showing one path, you see where the real socket is. they differ. that's why you get an error.
<8> of course portability seems to be more important to people investing into proprietary rdbms
<10> arjenAU, That part is clear.
<9> redblade: so make sure your app uses the proper path. or change the /etc/php.ini file. or whatever. you didn't give enough info about your environment to allow anyone to give you more specifics
<8> however even if you are say on firebird and want to move to ingres for some odd reason
<8> because ingres has some feature you need
<8> then there would be costs involved
<9> lsmith: I see most startups not caring about portability, just about money - or the lack of it. they have no money to invest, so it's a moot point.
<10> environment... hmm
<8> or maybe you want cheaper support contracts with another rdbms
<10> what would be most helpful
<10> ?
<9> redblade: the hint I gave a few minutes ago ;-)
<8> arjenAU: thats true .. thats why most startups end up with open source and no support contracts :-)
<9> lsmith: take it as a fact, dude. that's the market.
<9> ok, gotta go.
<8> laters
<10> arjenAU, there is no php.ini file
<10> arjenAU, basically what I don't understand is, why isn't the path /usr/local/mysql/bin/mysqladmin ?
<10> because (/usr/local/mysql/bin/mysqladmin status) causes the error message to jump up
<11> how do I store an md5 using a binary data type?
<11> insetad of varchar(32) ?
<11> type BINARY i guess
<12> Is there a cleaner way to do this "SELECT * FROM artists WHERE (artists.artist BETWEEN 'Z' AND 'ZZ') to get artists starting with Z?
<13> WHERE artists.artist LIKE 'Z%'
<12> bigfoot-: That is what I was using, then I realized I wanted a range
<3> Why do you want a range?
<12> will: To allow the user to search between different characters
<3> Well, then, no..
<14> is Z always the last alphabet?
<12> No, I just didn't know if there was a way to grap the end of the character set, like some distinguished terminating character
<14> dcvkldcvjtjke: it gets complicated if it uses the collation and case insensitive
<14> dcvkldcvjtjke: for case sensitive of binary, then there could be ways
Return to
#mysql
or
Go to some related
logs:
#ldap
can't mmap package info file
#kernel
#fedora
evill33t svn
shockwave livna
#ubuntu
sql qieries
__buildins__
progname Audio::Scrobbler
|
|