BIND 10 #1144: RR type implementation: DLV
BIND 10 Development
do-not-reply at isc.org
Wed Aug 10 21:46:32 UTC 2011
#1144: RR type implementation: DLV
-------------------------------------+-------------------------------------
Reporter: shane | Owner: jinmei
Type: task | Status: reviewing
Priority: major | Milestone:
Component: | Sprint-20110816
libdns++ | Resolution:
Keywords: | Sensitive: 0
Defect Severity: N/A | Sub-Project: DNS
Feature Depending on Ticket: | Estimated Difficulty: 3
Add Hours to Ticket: 0 | Total Hours: 0
Internal?: 0 |
-------------------------------------+-------------------------------------
Comment (by jinmei):
Replying to [comment:10 dvv]:
> Still working on comments and tests.
So do you want me to do full review, or are you asking for feedback on
the selected points? For now I assume the latter.
> I retained the template approach, but eliminated the extra inheritance
level.
I don't necessarily oppose to the idea of using template per se (and
in general this is one of the common cases where templates make
sense), but I'd point out some possible issues with this specific
case:
- As I already pointed out, including ds_like.h from {ds, dlv}_xxx.h
will cause a trouble.
- If you keep this style, using a separate DSImpl struct doesn't make
much sense any more because its definition is now part of the public
header file.
> Amongst other things, please let me know if using
RRParamRegistry::getRegistry().codeToTypeText() is appropriate from inside
a Rdata class.
It's not necessarily invalid, but in this case I'd construct the
corresponding RRType object:
{{{
isc_throw(InvalidRdataText, "Invalid " << RRType(typeCode) <<
" text");
}}}
That would also be better in that we can avoid including one
additional header file (rrparamregistry.h) from a public header file.
--
Ticket URL: <http://bind10.isc.org/ticket/1144#comment:11>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development
More information about the bind10-tickets
mailing list