Re: [boost] Policy-based designs considered harmful?

From: Jeremy Siek (jsiek_at_[hidden])
Date: 2002-01-21 11:10:12

Actually, this problem may not be as bad as it sounds. I had a similar
problem to overcome for testing the MTL (and for the BGL adjacency_list).
I tested all possible combinations, but I was able to write many of the
tests in a generic fashion so that they would be applicable to any
combination. Yeah, the tests took a long time to compile and run, but at
least I didn't have to type in an exponential amount of test code! I'm not
sure just how generic the tests can be for smart ptr, but I wouldn't be
surprised if it makes the job manageable.

On Mon, 21 Jan 2002, Peter Dimov wrote:
pdimov> Everyone that believes in policy-based designs ought to consider the
pdimov> following
pdimov>
pdimov> Assignment: Write the unit test for Loki::SmartPtr that probes
pdimov>
pdimov> a) all possible combinations of the supplied policies;
pdimov> b) user-defined policies.
pdimov>
pdimov> ;-)
pdimov>
pdimov> --
pdimov> Peter Dimov
pdimov> Multi Media Ltd.
pdimov>
pdimov>
pdimov> Info: http://www.boost.org  Send unsubscribe requests to: <mailto:boost-unsubscribe_at_[hidden]>
pdimov>
pdimov> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
pdimov>
pdimov>

----------------------------------------------------------------------
 Jeremy Siek                          http://php.indiana.edu/~jsiek/
 Ph.D. Student, Indiana Univ. B'ton   email: jsiek_at_[hidden]
 C++ Booster (http://www.boost.org)   office phone: [phone]
----------------------------------------------------------------------