|
Well,
I can't find out what kind of database uses .idx and .dat files for storage of data. I have tried several ODBC drivers but they won't work.
I hope someone has an answer to these questions,
thanx in advance,
Robert Smit
|
|
|
|
|
I've come across several in-house databases that have used those extensions, I've not come across a commercial one with these extensions. Most tend to be of the flat-file variety which usually means no ODBC support.
I have used those extensions for a flat-file database in the past but I doubt you'll be trying to connect to that 
Michael
Life’s not a song.
Life isn’t bliss.
Life is just this.
It’s living. -- Buffy the Vampire Slayer: Once more, with feeling
|
|
|
|
|
Micheal,
Thank you for your answer I think you could be right. It is only a small program that stores data to those idx and dat files. The database is used only for storing pupil en group results for math exercises. So it is not a complicated DB with a lot of stored procedures.
My question to you is how you read and write to those files in C++.
Robert
|
|
|
|
|
Without the file structures you'll have a hard task.
In the good old days when I used to try and integrate with Pegasus Accounts who used a similar system, I usually saved a record into the database using their app and then hex-edited the file to figure out what had been updated and changed.
Michael
Life’s not a song.
Life isn’t bliss.
Life is just this.
It’s living. -- Buffy the Vampire Slayer: Once more, with feeling
|
|
|
|
|
Hello,
like subject says, is there something more than setting a password for the database that I can do to make sure that people don't open an MDB file and mess with it?
Thanks!
---------------
Concentrating on Ideas
http://www.edovia.com
|
|
|
|
|
Hey LukeV,
You may choose to open an MDB in exclusive mode and set password to it. So that each time the MDB file is opened, it will ask password.
Deepak Kumar Vasudevan
http://deepak.portland.co.uk/
|
|
|
|
|
Hi.
I've got a basic ADO AddNew call, running in a loop, against Access 2002. Running on VC++6, Win2K, MDAC 6 sp2. See code below.
It seems to use memory constantly, even when the recordset object is released. It uses approx 10MB per 100,000 rows written. My system needs to be on line and writing data for a week at a time so this is going to be a problem for me. Any ideas?
Ta.
Gizz.
The following function is called from another big loop...
void fn()
{
using namespace ADODB;
_RecordsetPtr pRst = NULL;
_ConnectionPtr pConnection = NULL;
_bstr_t strCnn("Provider=Microsoft.Jet.OLEDB.4.0;Data Source=c:\\data.MDB");
try
{
TESTHR(pConnection.CreateInstance(__uuidof(Connection)));
pConnection->Open(strCnn,"","",adConnectUnspecified);
TESTHR(pRst.CreateInstance(__uuidof(Recordset)));
pRst->Open("tblMyTable",_variant_t((IDispatch *) pConnection, true),
adOpenKeyset,adLockOptimistic,adCmdTable);
_variant_t strGBID("GBID");
_variant_t strTSID("TimestampID");
static long i = 0;
static long imax = 1000;
for (;i < imax; i++)
{
pRst->AddNew();
{_variant_t vtFld;
vtFld.vt = VT_I4;
vtFld.lVal = i;
pRst->Fields->GetItem(strGBID)->Value = vtFld;}
{_variant_t vtFld;
vtFld.vt = VT_I4;
vtFld.lVal = i+1;
pRst->Fields->GetItem(strTSID)->Value = vtFld;}
pRst->Update();
}
imax += 1000;
pRst->Close();
pConnection->Close();
pConnection.Release();
pRst.Release();
}
catch(_com_error &)
{}
}
|
|
|
|
|
I don't know where is the problem,But about pConnection.Release(); and pRst.Release(); I have to say you don't need them because you use these inrefaces as an pointer and they will release automaticlly. But you problem MAYBE because of access,I know it has limitation on number of records,and after you insert 100,000 rows in a table in crash or ....You didn't which line cause memory leak!
Mazy
"And the carpet needs a haircut, and the spotlight looks like a prison break
And the telephone's out of cigarettes, and the balcony is on the make
And the piano has been drinking, the piano has been drinking...not me...not me-Tom Waits
|
|
|
|
|
Hi,
I am new to database programming in c++ and come across some quest. which i can't find in books.
I am programming in c++ mfc platform, using ADO to access to the sql database.
However, my database is using MS windows 2000 Terminal Services Web Connection
and i don't know how to link to the database. In Short, how is the connection string parameters to be like.
Thanks
|
|
|
|
|
Creating a database this way:
if( database.OpenEx(sSql,CDatabase::noOdbcDialog) )
{
sSql.Format( "CREATE TABLE Sounding ([Pressure (mb)] FLOAT,[Height(m)] FLOAT)");
database.ExecuteSQL(sSql);
sSql.Format("INSERT INTO Sounding VALUES('%f','%f')",mPressure,mHeight);
database.ExecuteSQL(sSql);
....
If mHeight is 17.60 it is inserted 17600000, if I change to '%.f' it is inserted 17, and if '%.2f' 1760. What is the solution to insert the right decimals?
Thanks.
|
|
|
|
|
There are two point here:
1.Are you sure that mPressure and mHeight are float value or maybe you define them as int.
2.It could be in your database.You set the Scale of it tp 0.Its is zero by default so you forgot to change it,So SQLServer or access omit them.
Mazy
"And the carpet needs a haircut, and the spotlight looks like a prison break
And the telephone's out of cigarettes, and the balcony is on the make
And the piano has been drinking, the piano has been drinking...not me...not me-Tom Waits
|
|
|
|
|
Yes, mPressure and mHeight are float values.
How could I change the default Scale?
Thank you.
|
|
|
|
|
Do you use Access or SQLServer?
Mazy
"And the carpet needs a haircut, and the spotlight looks like a prison break
And the telephone's out of cigarettes, and the balcony is on the make
And the piano has been drinking, the piano has been drinking...not me...not me-Tom Waits
|
|
|
|
|
Well, I'm using it into an MFC app, with Visual C++, and getting the ODBC Microsoft Excel Driver. It´s the first time I include a database utility in an applicattion, so I don´t know exactly what I'm using. If it helps, I use Cdatabase class with function member ExcecuteSQL.
Thanks.
|
|
|
|
|
José Luis Sogorb wrote:
ODBC Microsoft Excel Driver
I never used it,but I don't know how do you want to create table or other things in it. May I know why do you choose Excel driver?!!!!!!!!!!
Mazy
"And the carpet needs a haircut, and the spotlight looks like a prison break
And the telephone's out of cigarettes, and the balcony is on the make
And the piano has been drinking, the piano has been drinking...not me...not me-Tom Waits
|
|
|
|
|
Hi,
I'm trying to write a binary data from a recordset field into a file. I'm
creating an ADODB.Stream object (adTypeBinary), write the rs field into it
and then save it to a file.
The problem is that the file is not written well; it's size is double than
it should be.
Can somebody help me?
Thanks!
Here's a part of the code:
Dim objStream
Set objStream = CreateObject("ADODB.Stream")
objStream.Type = 1 'adTypeBinary
objStream.Open
objStream.Write rs.Fields.Item("Object").Value
objStream.SaveToFile filePath, 2
objStream.Close
Set objStream=nothing
|
|
|
|
|
How will Yukon change the way we develop database apps? We are in the requirements gathering stage of a new product development effort and we're thinking that since Yukon will be out before we are shipping, that we may want to take advantage of it's new and supposedly revolutionary capabilities.
I'm particularly interested in how the inclusion of the .NET CLR into the database will be implemented. I get the impression from the articles that I've read that you'll be able to bypass some of the mid-layer technologies necessary to now do data access. Certainly we won't need TSQL, but what else will we be able to bypass? SQL itself? ADO/ADO.NET? OLE DB/ODBC? I wonder, will we be able to implement a data-connected class in C# running on Yukon almost exactly as we would implement an internal type within a windows app and have Yukon automatically implement the actual data storage in the database? In other words, will we be able to treat data just like objects? If so, this is a huge pardigm shift and it bears investigating before we start developing with traditional techniques.
Has anyone played with Yukon yet? Is it included in the latest MSDN subscription?
|
|
|
|
|
I am VERY new to C# and ADO.NET. Can someone tell me why I get a namespace error with the following code....
using System.Data.ADO;
I have tried downloading a couple of ADO.NET examples, but there always seems to be missing references...
ie) System.Configuration
System.Diagnostics
Is there a project setting to change, or maybe missing dlls?
Thanks for any help ... Paul 
|
|
|
|
|
If you want to use ADO.NET only add System.Data . Not System.Data.ADO.
Paul Silvernail wrote:
System.Configuration
System.Diagnostics
These are somethings different from ADO.NET.You don't need them.
Mazy
"And the carpet needs a haircut, and the spotlight looks like a prison break
And the telephone's out of cigarettes, and the balcony is on the make
And the piano has been drinking, the piano has been drinking...not me...not me-Tom Waits
|
|
|
|
|
Thanks Mazy,
I have sucessfully opened a jet database with OleDbConnection ... were cooking on the front burner now!!
|
|
|
|
|
I am designing a distributed application which is implemented using remoting. Basically the server side remote object is used to retrieve information from the database.
Currently in this object i am using a single dataset to store all data, which runs fine as we have not yet tested against large volumes of data.
Now the thing is that i would like the application to be scalable for which i might have to use things like object pooling, now if i pool my objects then each object will have a differnet instance of the dataset and there will be no synchronization between them. But i have a feeling that directly manipulating the database will degrade my performance.
Anyone have any stats, ideas or suggestions or any experience with a similar problem?
May the Source be with you
Sonork ID 100.9997 sijinjoseph
|
|
|
|
|
Avoid using large datasets in remoting applications. Even if you specify binary serialization, the contained objects (datatables, datarelations) still get serialized to HTML diffgrams (very verbose) and consume immense ammounts of memory and bandwith if many rows are involved. My experience is that this also leaks memory (although this may just have been due to costruction in the "large object heap" which seems only to be garbage collected if the moon is full). This MSDN magazine article <a href="http://msdn.microsoft.com/msdnmag/issues/02/12/CuttingEdge/default.aspx">http://msdn.microsoft.com/msdnmag/issues/02/12/CuttingEdge/default.aspx</a>[<a target=_blank title='New Window' href="http://msdn.microsoft.com/msdnmag/issues/02/12/CuttingEdge/default.aspx">^</a>] goes into much more detail, and suggests some workarounds.
We ended up loading a custom class instance (basically an arraylist of arraylits), serializing that, deserializing and reconstructing a dataset on the client side. Cut memory usage by factor of 4 and increased performance by factor of 3. No more memory leaks.
When all else fails, read the directions...
OldRob
|
|
|
|
|
Thanks a lot mate, but actually i don't actually pass the dataset over the wire at any time but still a very informative link.
Could you tell me what your experience was in using a dataset i mean for large databases with the enitre database in memory do you think it will be a good solution?
May the Source be with you
Sonork ID 100.9997 sijinjoseph
|
|
|
|
|
Depends on what you mean by large ... If you plan to keep several million rows of two or three hundred columns in memory as a read only db, I would think this would strain even a large box. Not too sure the class is thread-safe for multiple readers(enumerators...), pretty certain it is unsafe for multiple writers. We were working with a 30GB database,so no thought of keeing it in ram. Datasets/DataAdapters are not real performant for heavy insert/update use, sqlcommands with stored procs work best for this, and sqlDataReaders are much more performant for populating read-only caches...
The DataSet is a rather complex structure, carring a lot of overhead (duplicate copies of all the data, state information for every row, complete schema information... really is an in-memory db of sorts)
Also, the remoting issue is just as severe when passing data between processes in the same box (appdomain to appdomain). We were fairly convinced that this resulted in real memory leaks with datasets of 10000 rows x 300 columns or so (worse if multiple datatables adding up to this..)
OldRob
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
