this is my query
query="SELECT studenti.Nume, studenti.Prenume, cursuri.denumire_curs, note.nota, note.data " +
"FROM studenti INNER JOIN " +
"(profesori INNER JOIN " +
"(cursuri INNER JOIN [note] ON cursuri.id_curs = note.id_curs) " +
"ON profesori.id_profesor = cursuri.id_profesor) ON studenti.id_stud = note.id_stud " +
"WHERE profesori.user_name = '" + user + "'";
i want to give a name to the result table so i can put it in a DataSet
ntext: Variable-length Unicode data with a maximum length of 2^(30 - 1)(1,073,741,823) characters. Storage size, in bytes, is two times the number of characters entered. The SQL-92 synonym for ntext is national text.
If your news content will always be less than 8000 characters you can use varchar or nvarchar. Otherwise you will have to use text, ntext or image datatype.
If you ask me no chance to do this in a user defined function. You would need dynamic SQL with exec( @stmt ) and that is not allowed in UDFs. Can't you use a stored proc? You can use dynamic SQL there.
Visit my blog at http://www.cubido.at/rainers
You have trusted_connection = true and yet you are not wanting to use a trusted connection as you supply a user name and password. As trusted connections are considered more secure it tries to log you on using a trusted connection, but it fails.
I need to modify the excel export feature in Reporting Services... The issue is that client needs to have chart in excel format after exporting so that user can manually change the values in excel sheet to check the effect in the chart dynamically. Can anybody tell me how I can do that? I know it would require the creation of chart in BIFF format. But could not find out how?
If someone knows any .NET library that can help me in that please let me know. I don't want to go for SoftArtisian Office Writer as it has certain limitation as it doesn't supports the grouping of data in RS.
I'm considerind adding database capabilities to my software. First step is selecting the proper database. The database should have 0 maintenance, be light weight. I would also appreciate a single file binary database and not BDF/Paradox like ASCII files.
At the moment I think SQLite could make it.
Once you made a choice - especially with small companies like mine - it is difficult to switch to another one. So since I'm not a database specialist, I would appreciate advices or suggestions on other candidates, and possibly pros and cons.
SQLite is a proprietary format embedded database .. that sort of works against 'switching to another one' unless, you use a database abstraction layer in between your code and the database layer.
The most obvious one I can think of is ODBC, and there is an (?experimental) ODBC driver for Sqlite. You're still not off the hook with ODBC, because of SQL variations (Ie SQLite doesnt support all of the SQL-92 Standard) , but it could make it 'easier' to use ODBC
There is a database abstraction layer (cant remember the name off-hand, will look it up and try and post back), that MAY (memory is a bit dim) let you select between different database 'back-ends' - my guess is you would still have to stick to fairly basic SQL unless its translated somewhere for each potential database back-end.
.. but thats the approach I would use if there was any hint of having to change the database back-end ..
(if it counts for anything, I use SQLite in some projects - I also use SQLite and Oracle 9i in other projects)
ps .. your p3DO explorer looks great - wish I knew how to do some of that stuff - not enough time ....
Thanks for your detailed answer Garth,
Since my needs are rather basic I will most probably only use SQL basic commands. So I think I can cope with those SQLite limitations.
Also although I will design it for multi-users use it'll mostly be used by single user.
For the driver, yes it is a choice in itself but the thiner layer, the better.
> your p3DO explorer ... - wish I knew how to do some ... - not enough time
Yes it takes a lot of time, fortunately (or not I'm single and I love 2D/3D graphics. This gives me the will to go ahead.
Looking for the proper library/component is a hard part of the project (I made severall costly mistakes in the past).
Thank you for your advice. I already used Interbase in the past, it is a good database but not light enough for what I intend to do.
I would most probably choose it for professional clients running on a network (I've read SQLite is not good for transactionnal network access), but my needs are too basic even for Interbase.
Note - I don't say for FireBird since I don't know it.
I have a problem with and app that is connected to an SQL 2000 database
The user got a new server dual CPU 3.2 GHz Pentium 4 with 2G memory and the app is slower compare with the old machine a dual CPU 755 MHz Pentium 3 with 512 memory.
If you use the TCP/IP connection in ODBC driver, the new machine is very slow.
If you use Name Pipe in ODBC, both machine a doing well.
The old machine it is working well with both TCP/IP and Pipes.
Does anybody have any idea why it is such a big difference between TCP/IP and pipes for the new machine? The different is from 1 minute to 10 minutes for an operation.
SELECT q.cust_no, earliest_date, item
FROM tbl_order as t
SELECT cust_no, MIN(date) as earliest_date
GROUP BY cust_no
) AS q ON t.cust_no = q.cust_no AND t.date = q.earliest_date
1.The above code is working fine for insert statements. How to use the above code in case if we are trying to update database with more than 1 updated row of dataset. e.g. instead of using "insert into TEST values(30,'900')";, use "update test set emp=100". Here we donot know how many rows will be updated in dataset.