| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Comments:
<0> Is there a flag to mysqldump to reverse the order of options and not null in create statements
<0> so instead of column, type, not null, default it outputs column,type,default,not null
<0> the mysql syntax does not seem very portable
<1> Tenkawa: No, since it wouldn't understand that, it doesn't produce that.
<0> Therion: ok. wasn't sure if mysql understood both orders
<0> well darnit
<0> Therion: actually mysql _does_ accept the syntax I asked for
<0> I just tested it
<1> Tenkawa: If you're trying to produce SQL for some other engine, there may be some conversion utility you could use that would do it for you.
<1> Tenkawa: Woo I'm a liar
<1> Does it? :P
<0> `login` VARCHAR(45) default '' NOT NULL ,
<0> yes
<0> it acepted that (which s used by most dbs)
<0> Therion: been there tried that.. having to resort to building my own
<1> Amazing that it would accept things out of order
<1> It does not generally
<0> Therion: well by sql92 standards mysql's syntax is out of order I would say
<0> so maybe it was added in mysql both ways for compat
<1> If the standard mandates it in a different order then open a bug report
<1> and maybe we would change that
<1> Tenkawa: What engine are you tryign to port to?
<0> well its questionable if it would be a valid "bug" since mysql has enough non standard syntax as it is
<0> Therion: hsqldb and h2
<1> I wonder if the --compatible options cause it to switch
<0> I tried a few of those but the full list of them really isnt documented
<1> Tenkawa: Having non-standard syntax and being in opposition to the standard are different things
<1> Everybody has non-standard syntax :P
<0> had similar problems going from mysql to sqlite but I was able to write a parser for that
<0> :)
<0> yeah I know
<0> its more of a feature request than a bug I guess
<1> Well, if you can point to it as something the --compatible option or one ofour SQL modes should produce, then it makes sense as a bug report
<1> ie: if it doesn't produce SQL that can actually be imported by the target, then it's a bug
<1> Of course I do not thin kwe have any of those as targets
<0> Therion: is there a list anywhere of all the compatible options?
<1> mysqldump --help
<0> mysqldump --help has some but I'm not sure its all of them
<1> MySQL. Legal modes are: ansi, mysql323, mysql40,
<1> postgresql, oracle, mssql, db2, maxdb, no_key_options,
<1> no_table_options, no_field_options.
<1> hee
<1> With --ansi it still puts default last
<1> If that's wrong, then that's a bug. ;)
<1> You should try the migration toolkit
<0> oh intriguing
<1> If there is a jDBC driver for your target
<0> Mysql has it right.. its 4 other dbs that have it wrong
<0> according to sql92
<1> lol
<0> intriguing
<0> well that just ****s for me
<0> ROFL
<1> We're working very hard to be compliant with the standard. Probably harder than anyone except IBM
<0> thats goood
<1> Not there yet, but closer than some
<0> er good
<0> nod
<0> er #5
<0> the only thing that seems to validly throw some things off is the ``
<0> but tha canbe fixed with compact and ansi
<0> db #5 is ok
<0> but is missing jdbc compliancy
<0> Therion: oh well.. thanks for the info and help though
<0> guess its time for me to start writing a parser
<0> Therion: while I'm here... do you have any experience with MXJ?
<2> hi, any idea why my replication always dies when I delete a table or alter table name on master? why are these commands not replicated?
<2> Therion?
<0> Any of you got mysql embedded java server experience?
<3> Hello
<3> I've a simple question (not for me btw)
<3> I've a table with two fields, A and B
<3> also KEY A(A), and KEY B(B)
<3> the query: SELECT COUNT(*) FROM mytable WHERE a=100 and b=200 is *very* slow
<3> there are a lot of entries in the table where a=100 and b=200
<3> what I should do? Something like KEY A(B) maybe?
<3> I've no clue about how index works mysql
<1> d22: Presumably your configuration is responsible
<1> antirez: unless you are using 5.0 or later, you need index on (a,b)
<1> antirez: not just index on (a) and index on (b)
<0> Therion: did you see my earlier q?
<1> Tenkawa: Haven't used MXJ to a degree to have anything useful to say about it, honestly
<0> Therion: is it fully compliant from a "sql" level with iits equiv mysql version?
<0> so basicly could it replicate the schema/data with no issues
<4> hello again
<4> connection lost, I was not able to read any reply
<0> well here goes my mxj testing
<0> It would be great to use as long as its compat
<1> Tenkawa: don't even know that much about it, but it's embedding an actual mysqld server so it's "fully compliant" with that...it's just a regular server, packaged to appear to be a java object
<0> excellent :)
<0> just what I need
<5> What are the disadvantages of per-table innodb tablespaces?
<5> I can't come up with a good reason for not enabling them, but I'm sure there must be a reason there's a switch. :)
<2> Therion my configuration is default I believe
<1> d22: no replicate-* options?
<6> Hmm. I was reading through mysql-storedprocedures.pdf. The labels part really confuses me. What exactly are labels supposed to be for?
<1> mendel: They use more file descriptors and some extra processing during startup/shutdown, but that's about it
<5> Ok, thanks
<1> mendel: And if you have a crap OS or filesystem taht limits per-file sizes (not done in recent memory), you can run into problems with it that the shared tablespace would not have
<1> Narada: For GOTO! ;)
<6> hehe :)
<6> Yeah I need to get my head around this concept which is new to me. It is not like a programming syntax so ... need time.
<1> If you mean on the BEGIN/END, I dunno
<1> Clarity, nested blocks, something
<1> Do not know anyone who uses it
<1> Not that I know everyone
<6> Heh yeah if it's one of those minority features then great. I'm not missing anything important :)
<6> Is that a mysql extension?
<1> Should be straight out of the ANSI specification
<2> Therion I replicate only given databases
<1> d22: replicate-do-db?
<2> and when I change something like add/delete table this is not replicated
<2> yes
<6> I'm talking about labels on page 24 of this document: http://dev.mysql.com/tech-resources/articles/mysql-storedprocedures.pdf.
<2> Replicate_Do_DB: ppc,ppc2,feed,banners
<6> Just to clarify.
<1> d22: Are you "in" the target database? ie: a USE statement to go into it, before issuing the statement?
<1> d22: and not doing cross-database queries?
<2> what is cross-database query?
<1> USE test; DROP TABLE otherdb.table;
<2> no I dont do this
<2> I edit database directly with phpmyadmin for example, I add table XXXX and this table is not added on the replicated server
<1> phpMA may not be setting an active database
<1> Check your binlog to see if there is one
<1> Otherwise, use replicate-wild-do-table and not replicate-do-db
<6> lol taken from the mysql manual: "We don't mention GOTO in the MySQL Reference Manual.". Haha!
<6> I mean from the mysql pdf.
<0> hmm.. mxj doesnt appear to have a standalone mode
<0> uggh
<7> computed goto is the best
<7> you know.... the goto target is determined by a function call
<8> or jumps
<7> and such things do exist
<7> visual dialog script
<7> oh, got a great idea from that... let me toss a link
<7> http://qmsk.paivola.fi/stuff/vdd-html-db.txt
<7> it's like... so cool
<7> anyways, visual dialog sciprt uses computed gotos :P
<6> bleugh ***embler
<7> ***embler is nothing compared to vds
<8> haha html as a table type
<7> database storage engine
<7> yeah
<8> what are they on in Finland
<7> hmm?
<9> salmiakki flavoured vodka!
<10> haha
Return to
#mysql
or
Go to some related
logs:
#perl
ubuntu detect wlan0 lspci -ndiswrapper
gmail PERM_FAILURE: Parse Error:
compiz t41p
#centos
#linux
#linux
GnosticApex
#debian
group by variable restraints sql
|
|