The current log messages from the CocoPowder and
related classes are cluttered up with large quantities
of information that can only be deciphered by a
product developer. For example, the following is just
a tiny part of a large log file generated from a small
program.
callFields=[]
callOtherFields=[]
callFields=[]
callOtherFields=[]
callFields=[]
callOtherFields=[]
Done getting clause info.
Closing cursor.
Getting Meta Data handle.
Call Fields=[]
...
This type of output goes on for pages and pages of output.
Is this information intended to be useful for end users?
I suggest having a level of diagnostics for your developers
and another one for end users (e.g., focussed on SQL
statements being executed, how rows are mapping to objects,
etc.). The large amount of noise in the log files makes it
difficult to debug mapping errors quickly. Just my opinion...
|