writing code takes a lot of time because we need to write 
every character in the code. 
To my knowledge, since there is no OR, there is no central 
control over the script. If OR is htere, we can add the 
properties easily to the objects

It is not possible to implement in All areas of an 
Application that means simple Test case may be takes alot 
code.

->	QTP wont automatically provide the lines of code 
while recording. By default  QTP follows Object Repository 
method. 

->     Writing script for objects with multi fields proves 
the line of code to be very lengthy and difficult to 
comprehend.

->	DP by itself wont support smart identification, if 
at any case the object  identification properties wont 
exactly match.

->	It is not easy to specify regular expression for 
object identification if we are using DP.

->	DP is not supportive to insert checkpoints