| |
| |
| |
|
Page: 1 2 3
Comments:
<0> damn - my laptop now has an uptime of 811 days
<1> YuppieOTL, 2 years and a half :D
<1> incredible
<2> YuppieOTL:: i would feel really bad was I your laptop
<2> s/was/were
<2> wth.
<3> yar
<4> I have a mssql table with two datetimes. I need the average difference between the two. Does anyone have an idea how to do that ?
<3> avg(abs(dt1 - dt2))
<3> erm, average
<3> group by abs(dt1 - dt2)
<4> abs does what ?
<3> absolute value
<4> "Implicit conversion from type datetime to float is now allow"
<3> so explicitly convert it then
<4> So I need to use CONVERT or something ?
<4> ok
<3> I hate the way MS-SQLServer handles dates and times
<4> ;)
<5> hi
<5> i need some help with a query: i need to do a search and get all students with all the grades in my search and with a cote equal or grater then the search cote for those grades. You can find the tables here: http://www.pc-architect.com/~ann/help.htm Please help...
<3> cote?
<6> I'm still trying to figure out how you manage to design a schema and not know how to get the data you need out of it.
<5> like a number, saing for that grade they got a good chance to score or not
<5> so i need all students that were predicted to p*** the exams with X chance kind off
<5> rmah: understand?
<6> Main Entry: cote
<6> Pronunciation: 'kOt, 'kt
<6> Function: noun
<6> 2 : a shed or coop for small domestic animals and especially pigeons
<7> @param_IDs, delete from [table] where [id] in (@param_IDs) when @param_IDs= "1,2,3,4" gives me an error
<6> I'm not aware of any rdbms that that will work right in.
<3> wtf is a cote?
<3> k00: you can't do that
<7> ahh i see where i went wrong
<7> i would need @param_IDS = " '1','2','3'" etc
<6> Why, is ID a string? ;)
<7> IDs
<7> meaning , more than one :-D
<6> Yes. If you ID field isn't a string, why would you think p***ing a list of strings would help when a list of numbers doesn't work?
<7> lol
<7> true.
<7> So..why does tsql favor strings like this
<3> because it ****s?
<7> thats one possibility
<8> so when the topic says "useless info here"...
<8> does it mean "here" as in the chan /topic... or "here" as in the chan? :P
<3> the world is a mystery
<9> the world is like chicken
<3> like a giant chicken?
<9> semi giant
<1> In sql server 2005 I have a stored procedure with a SELECT inside. I want to send the output only if this select return at least 10 items. Have I to write 2 SELECT (one with COUNT and one normal) or there is a way to do this with a single SELECT?
<3> BlackArro: at a guess, ***ign the results of the select into a variable and then return based on some attribute of that var
<1> for example? I have to use DECLARE var inside the stored procedure and use var = SELECT ...?
<3> I don't know
<1> probably I can use the SET NOCOUNT OFF. I try!
<6> @@cursor_rows, probably.
<1> ok, I found @@ROWCOUNT :)
<0> BlackArro : there is no way to do this with only one select from a stored proc
<0> there must be at least two
<1> the first is the SELECT COUNT and the second is the SELECT, isn't it?
<0> yes
<1> ok, thanks YuppieOTL
<1> I have a SELECT but I want to display before records with PK >= 150000 and after records with PK < 150000. Can I do it with a single SELECT? I tried using where pk > 150000 OR pk <150000 but it always returns fields from the begin.
<6> What?
<1> excuse me for my English
<1> :(
<6> If you want records with pk >= 150000 and with pk < 150000 ... well, that's all records.
<1> yes, but you want to see records higher than 150000 before and records lower after
<6> You need to build an order by to do it, there where clause has nothing to do with the order of the result set.
<6> Probably have to use case
<1> I tried with the UNION statement and it works well. :)
<10> I'm using MSSQL 2005. Does anyone know how to filter out indistinct records for only one field in a select statement where there are multiple fields returned?
<11> (sqlserver2k ent) if I make sqlserver pick a random number between 1 and 4 over and over again, will it equal 2 25% of the time?
<12> no, it'l be 2 50% of the time
<11> 50%?
<11> why 50%?
<12> ya, there's only 2 numbers between 1 and 4, those are 2 and 3
<11> I'm saying 1 through 4 inclusive
<11> 1,2,3,4
<11> it will equal 2 25% of the time?
<12> heh, well, you gotta specific
<12> what do you think?
<11> I'm hoping yes ;)
<6> Over enough iterations, yes.
<6> It is random, after all.
<6> Which means it will occasionally be streaky, but over long runs it'll be fine.
<11> Drk`Angel: ok cool thx
<13> Hello. I'm reading some SQL for PostgreSQL, and i have come across the abbreviation "oid". Can someone please tell me what this might mean? thanks.
<9> did you check google first?
<9> if you did you'd find the answer all over the place
<9> oid is a unique id in postgresql
<9> every row is implicitly given
<9> object identifier
<13> yes, I see the abbreviation all over the place, but nobody says what it means. I see that it is a postgres type, but ... Thanks for the help, Darken.
<9> it does say what it is
<9> google "postgresql oid"
<9> and you find it all over the place
<9> i verified
<9> every single link tells you what it is
<9> even in foreign languages
<9> i don't mean to bite your head off
<9> just check a bit better on simple questions :)
<13> thanks, Darken|.
<1> is there a way to improve a LIKE %keyword% query in mssql 2005?
<14> BlackArro : yes
<1> ok. What is it?
<14> BlackArro : depends
<14> BlackArro : first you need to ask a more detailed question
<1> I have a select ... where field like %keyword% and the field is an uk field
<1> so there is an index on the column where i do the LIKE
<14> 'uk field'?
<1> unique key
<14> what is the data type?
<1> is nvarchar 150
<14> do you insert to that table much?
<1> no
<14> 150-unicode char unique values? heeheehee
<1> yes
<14> which rdbms is this for?
<1> there is a PK int and a uk nvarchar 150
<1> i'm using sql server 2005
<1> the like is on the uk field
<1> do you think a Create statistics on this field can help me to improve performance?
<14> if you're doing %value%, then the uk isn't going to help at all - you need to use the full-text stuff
<14> no - create stats won't help you a damn bit
<1> what are statistics for?
<14> what?
<14> that question indicates a number of holes in your understanding of sql as a whole...
<1> For what statistics are useful?
<14> the stats are used by the query optimiser/parser engine to decide where to start looking in the index
<1> Yes, I'm not an expert and my English is not good. I'm sorry :(
<14> any worthwhile database will have a scheduled job to update the statistics - more often for a heavily updated index
<14> but this has nothing to do with your problem
<14> the uk will only help your lookups if you are doing field like value%
<14> but you're not
<14> you're doing like %value%
<14> so the index is almost totally useless
<1> ok, I understand. So i have to use a full text index
<15> hi, is there anything in MSSQL server that's similiar to the SET data type in MySQL?
<14> face- : nfi. what *is* the SET datatype?
<14> face- : did you not understand the question?
<15> allows you to store multiple values in one column, in a somewhat clean way
<1> but when I use CONTAIN(field, 'keyword') it returns me different results than like %keyword%
<14> no ****. that's another good reason never to use mysql - using a single field to store multiple values is a sign of a ****ty design
<14> that mysql actually supports it is the sign of a ****ty rdbms
<1> are there a full text command equal to the like %keyword%?
<14> so, the answer is 'no, there is nothing in ms-sql that is the equivalent'
Return to
#sql
or
Go to some related
logs:
#windowsxp
#debian
#linux-noob
sunce chat
daily q.n225
wrong # args: should be while test command
XBMC bangbros plugin.
#computers
#visualbasic
#politics
|
|