BIND 10 #583: random Qid

[BIND 10 Development]

#583: random Qid
-------------------------------------+-------------------------------------
                 Reporter:  jreed    |                Owner:  stephen
                     Type:  defect   |               Status:  reviewing
                 Priority:  major    |            Milestone:  R-Team-
                Component:           |  Sprint-20110308
  resolver                           |           Resolution:
                 Keywords:           |            Sensitive:  0
Estimated Number of Hours:  1.0      |  Add Hours to Ticket:  0
                Billable?:  1        |          Total Hours:  0
                Internal?:  0        |
-------------------------------------+-------------------------------------
Changes (by jelte):

 * owner:  jelte => stephen


Comment:

 Replying to [comment:5 stephen]:
 > '''src/lib/asiolink/qid_gen.h'''[[BR]]
 > The "#ifdef" sentinel is !__QID_GEN_: for consistency with the other
 header files is should be !__QID_GEN_H.
 >
 > Convention is to have private members suffixed with an underscore.
 >

 fixed

 > A little documentation as to what the mt19937 class is would be useful.
 >

 added

 >
 > '''src/lib/asiolink/qid_gen.cc'''[[BR]]
 > Does the !QidGenerator have to to be allocated dynamically? Why not make
 it a static object and call seed() in the constructor? That way there will
 be no complaints from memory checkers.
 >

 Right. Originally i made it dynamic so we could enforce the use of the
 singleton instance; we could make the constructor private. But thinking of
 this again, I agree that having it automatically cleaned up is more
 important.

 > return values are not enclosed in brackets.
 >

 doh. fixed

 >
 > '''src/lib/asiolink/tests/qid_gen.cc'''[[BR]]
 > OK, will accept the argument for now.  However, apart from the fact that
 I think that the Dilbert cartoon in http://www.random.org/analysis might
 indicate problems with this approach :-) I have added a task (#668) to the
 general backlog concerning the need to do proper randomness tests.
 >

 Right.

 > Other: Ticket #669 has been created for checking the QID in the
 response.

 cool :)

 Changes in commit 058f40e75eb59aea1d3e18ffcdfdacc8386aafa3

-- 
Ticket URL: <https://bind10.isc.org/ticket/583#comment:6>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development