C Do transactions have to be in the database?

It is possible to write things directly in c.

This kind of business processing is not expected by the design of DataGridView.

DataGridView is designed as a data source-binding mode, which is to hope that when the database data changes, we can execute the binding method to synchronize the page display.

The basic idea of adding, deleting, modifying, and querying is as follows: Get a new value -- operate the library -- rebind.

After deleting a record, you need to bind the data source to the datagridview again so that consistency can be achieved.

Definitely different, the transaction of C ** is a transaction of business logic, while the transaction in SQL is a transaction of data logic, maybe the transaction of business logic teaches the scope of transactions of SQL is larger, and the designer needs to control the transaction.

Depending on whether you intend to control the transaction in the logic of the database or in the logic of the business, in fact, the implemented transaction will also be converted to a database transaction.

There is a lot of controversy on the Internet, and personally, I prefer to agree that transactions should be done by our programs, not through SQL statements, because SQL statements are dead, while business is alive.

For example, when the business side wants to add users, the users must enter a certain group, in fact, this is a business requirement, and it is better to implement it in **.

There are also some transactions that are difficult to complete with a database because there is a lot of data that is dynamic and has to be done through **.

I think the runtime environment is different, the transactions in C run in the .NET Framework hosting environment, and the transactions in SQL statements run in the data engine of SQL Server.

sqltransaction is a transaction object in class C, which can maintain the data uniformity of the database.

This is required when multiple SQL statements are executed at the same time.

For example, if the first two statements are successfully executed, but the third one is incorrect, you can use transactions to undo the first two results.

If you don't use it, it's likely to be inconsistent with the business.

First of all, you need to understand what a rollback is.

For example, I have an order, and there is information about a hundred products in the order, and I want to insert this order into the database, so if I don't use transactions, insert them one by one, and when I insert the 50th, I wipe it, and the network is disconnected... Then you are now in an embarrassing situation, the 50 pieces of data is not a complete data, maybe others will query this order and say: "Hey, no, I obviously have 100 products in this order, why only 50".

If you use a transaction, it is to submit the insertion operation of these 100 commodities at one time, if there is an exception in the middle, none of these 100 commodities will be inserted into the database, and you can also call rollback to roll back the database to the state before the insertion operation.

In addition, to tell you secretly: using transactions to commit operations at once is much more efficient than inserting delete modifications one by one in a loop (if you don't believe it, you can experiment with it).

connectionstring="datasource=.\sqlexpress;

initial

catalog=cecdom;

integrated

security=true;user

id=sa;password=lsyuan"；

Here's an example of a connection string that doesn't have the server name in it, i.e. "datasource."

That paragraph. If you're not familiar, you can use the wizard to create a connection.

followed by sqlcommand

is a stored procedure, pay attention to check whether the parameters correspond.

Transactions are used when multiple SQL statements are required to complete a task.

It means that the database execution object is bound to a specific SQL, so you need to create a new cmd object each time.

See if the database engine of the database is innerdb