It should be divided into two groups of data as much as possible. 24) The database should be considered for the case where the amount of data is relatively large. 25) The database interface should be used in standard ODBC or ADO interfaces, try not to process the interface provided by the actual database DBMS because you may replace DBMS in actual use. 26) Small data can consider files, file paths should have to be designed to be relative paths. 27) In a function, you should try it immediately after using the file immediately after using it, so that other programs may use files. 28) Don't try to read all the files all over the memory, you should handle the big file. 29) Write programs should provide relevant test procedures to provide test methods. 30) Should consider the code, the usage of functions, do not exceed the range of functions that can be used.
-------------------------------------------------- ------------------------------
----- Excerpt from VMware.cn
