BIND 10 #418: new query logic for in memory data source (1.5)
BIND 10 Development
do-not-reply at isc.org
Sun Dec 5 21:31:15 UTC 2010
#418: new query logic for in memory data source (1.5)
---------------------------+------------------------------------------------
Reporter: jinmei | Owner: jinmei
Type: task | Status: reviewing
Priority: major | Milestone: y2 12 month milestone
Component: b10-auth | Resolution:
Keywords: | Sensitive: 0
Estimatedhours: 0.0 | Hours: 0
Billable: 1 | Totalhours: 0
Internal: 0 |
---------------------------+------------------------------------------------
Changes (by jelte):
* owner: jelte => jinmei
Comment:
Code changes themselves look ok, one thing:
Should the bare pointer in ZoneTable::FindResult not be a ConstZonePtr?
(also have a comment on that name, but see below)
I do wonder about the general direction this is heading though; for
instance this change implies a new mandatory subclass for each datasource.
One or two of those may not be a problem, but it feels like a slippery
slope.
If the Zone class is only meant to be a specific subclass from
AbstractZone for the in-memory datasource, I would suggest we rename it,
to better reflect that.
If it is possible that we can provide a default class for 'general'
datasources that don't need specific features, that's what Zone should be
(if not, I think we should consider renaming AbstractZone to Zone
actually, to make calling code more intuitive; Zone::NXDOMAIN vs
AbstractZone::NXDOMAIN); and then recheck the shared_ptr typedefs, their
names don't match what they encapsulate right now.
Oh and a tiny comment on a comment; I'd say that the FindResult
struct/class is all state, not no state (comment at the definition of the
first one zonetable.h line 72) :)
--
Ticket URL: <http://bind10.isc.org/ticket/418#comment:4>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development
More information about the bind10-tickets
mailing list