| |
| |
| |
|
Page: 1 2 3
Comments:
<0> http://www.b0g.org/wsnm/article/16261
<1> bleh, writing specs is ****ing boring ...
<1> I wanna code damn it!
<2> nlol
<2> MuteThis ?
<2> WHEN @OrderBy = 'NumericID'AND @Descending = 0 THEN RIGHT('000' + CONVERT(VARCHAR(3),NumericID), 3)
<2> i use that code to right aligh
<2> replace 0 with whatever you want
<2> change the number of 0's and the number 3 to fit your appropriate length
<3> Arrakis: i'll take a look thanks
<3> Arrakis: nice, thanks
<2> np
<4> is it unrealistic or uncommon to do connection based transaction from a series of web pages?
<2> unrealisitic?
<4> when i do a connection based transaction, it locks all the tables ***ociated with the transaction, and the tables cannot be accessed until that transaction is finished. how does one overcome this problem?
<5> Transactions issue locks.
<5> The scope and severity of the lock depends on the actions taking place in the transaction.
<4> yah.
<4> um.
<4> im doing 5 inserts.
<6> simultaneously?
<5> Which RDBMS?
<4> say i do insert into contacts values('1'), then sql server wont even let me do "select contacts" until that transaction is finished
<4> * from
<5> select * from contacts with (nolock) would work, tho
<5> Or setting the connection to read uncommitted
<4> oh.. so i'd have to modify all select?
<4> basically, im invoking 5 sprocs that do inserts, and i dont know what the best approach is...
<5> By default a SELECT statement will attempt to issue a shared lock for the records that it is reading in order to ensure that they don't change while being read. NOLOCK/READ UNCOMMITTED disables this, which might allow dirty reads.
<4> thank you
<7> sessm : client-side transactions are bad, however - why not wrap the 5 sprocs in a 6th?
<5> The transaction that ran the INSERT statement has an exclusive lock on the new row in the table in question which prevents a table-wide shared lock from being issued.
<4> YuppieScm: see. im no db guy, but.. common sense tells me that client side transaction ... just doesnt sound reliable
<5> issueing transactions from client side is reliable, just slower.
<4> k. but. it ****s that i have to modify all my select statements..
<5> You don't, SET TRANSACTION ISOLATION LEVEL READ UNCOMMITTED
<4> oh.
<4> thats much better
<4> thank u
<4> YuppieScm: right now i have a 6th sproc that takes all the params. 70 params. unmanagable.. ugly
<4> thank u~
<2> lmao
<2> whoa 70 ?
<1> wtf?
<1> store those ****ers in a table somewhere, and query it ...
<8> Is it not possible to have something like this in a stored procedure on SQL 2000?
<8> ORDER BY @order_by_clause
<1> compi ... being as you don't order by arbitrary values, no ...
<1> if you want to do something like that, you need to generate dynamic sql ..
<1> declare @sql varchar(2000); set @sql = 'select ' + @select_list + ' from ' + @from_clause + ' where ' + @where_clause + ' order by ' + @order_by_clause; exec @sql;
<8> i was afraid of that
<8> Why cant it be done with the other syntax?
<1> because @order_by_clause is most likely a 'string', which would turn the query into:
<1> ... ORDER BY 'column, list';
<8> the order by variable looks like this: fname
<1> you can't expect the query compiler to figure out in what manner you plan to use a variable ...
<1> compi, omfg ... fine, then:
<1> ... ORDER BY 'fname';
<1> which is STILL invalid ... as 'fname' is a string, not a column name ...
<8> removequotes(fname)
<8> :)
<8> whew
<8> a little jumpy today, arent we
<8> how fitting is the topic :)
<1> i hate when people say how stupid ****ing functionalities should exist in their dbms
<1> if you want non-standard functionality that does more or less, nothing at all ... then use mySQL
<8> i hate when my fav chinese restaurant puts squirrels instead of chicken in my orange "chicken"
<9> I love how mysql will accept 31-feb-2006 as a valid date :P
<9> That makes date math SO much easier
<1> Jarett_, alot of dbms's will accept that, as it's easily translateable into a valid date
<1> postgres accepts textual date's as well ...
<8> and that's not stupid ****ing functionality
<8> heh
<1> compi, who said it was?
<9> no DB should accept that as valid...
<9> one that does shouldnt be trusted
<1> automagically determining whether or not the variable is used as a literal string or a string is stupid ... the developer should be smarter
<1> Jarett_ their's nothing wrong with acceping '31-feb-2006' ... as their's only one way to translate it ...
<1> I like the way postgres works with dates though ...
<9> the db shouldnt be the one to determine that you mean 31-feb-2006 as 03-mar-2006
<1> current_date - '10 days 12 hours 2 minutes 3 seconds' ... is valid in a postgres query
<9> thats because current date is a real date in time
<1> how do 31-feb-2006 and 03-mar-2006 mean the same thing?
<9> 31-feb wont ever be
<9> zoul, you tell me.. thats the problem
<1> '31-feb-2006' will be converted to '02/31/2006' as it should be ...
<9> zoul, do you see the point that feb has 28 days, sometimes 29
<1> now, if mysql does the conversion incorrectly, then thats because your p***ing it an invalidly formatted string
<0> why are you helping with mysql?
<1> Jarett_ I missed that point ...
<9> ShrikeX, we arent
<0> heheh
<9> we are discussing its stupidities
<1> in that event, 31-feb-2006 should return an error, not 03-feb-2006
<0> ah good
<5> Point is that MySQL ain't bright enough to realize that the date is flat-out invalid.
<0> hahah
<0> it accepts feb 31st?>
<1> 02/31/2006 should error, simple as that ...
<5> It did in 4.x
<0> oracle throws a **** fit
<9> not sure if the newer builds do but it did for a long time
<5> 5.x probably still does.
<0> I remember the first time some jack ball sent me a file with that in it
<0> and it killed my etl scripts
<0> man was I mad
<5> There's a site out there which lists all of the bizarre implicit behaviors MySQL exhibits.
<1> ShrikeX your etl scripts were poorly written then, they should account for data entry errors ;p
<5> In fact, I think it's in the bots.
<1> @mysql
<0> @ mysql
<10> No match found for mysql
<0> @/ sql
<10> Search on sql returned: religion sqlserverauth sqlporting learnsql msde_license rdbms mysqlgotchas pginstaller nopaste whynomysql
<11> good thing too
<0> @ mysqlgotchas
<10> mysqlgotchas = http://sql-info.de/mysql/gotchas.html
<9> @ whynomysql
<0> @ whynomysql
Return to
#sql
or
Go to some related
logs:
#politics
#politics
charicatuer
#politics
get high on adderall
#windows
sify
girl2006+pics
#c
#politics
|
|