-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Steinar H. Gunderson wrote:
> On Wed, Oct 10, 2007 at 08:53:14PM +0200, Bram Matthys (Syzop) wrote:
>> Personally, I'm more unhappy with the 'Report number of timeouts back to the
>> user code' API change.
>
> Mm, but I guess that's a different discussion, no?
Indeed. I commented on both issues, though.
> I guess the main cost of breaking an API is in breaking the first function.
> As long as the rest of your changes are done at the same time, it has less
> impact.
>
> So, the question is really: Do we want to break the API to get timeouts _and_
> TTL values?
I don't understand, why does this has to be an 'x AND y' question?
Isn't it possible to threat them differently?
As I stated I personally see a big difference between breaking the API for
some timeout statistics in the general resolver routines (which are likely
to be most used) versus returning the parsed ttl in a *record parse function*.
Anyway, I don't expect any change, I justed wanted to state my concerns /
opinion so you will give this more thought in the future.
Breaking the API of a library shouldn't be thought of lightly. OTOH it
shouldn't be completely avoided either if there's a good reason/addition. I
still maintain my view that timeout statistics do not fall in that category,
though.
Will be my last reply... ;)
Bram.
>
> Obviously, this is not a discussion I can make on my own, as I'm biased here. :-)
>
> /* Steinar */
- --
Bram Matthys
Software developer/IT consultant syzop_at_vulnscan.org
PGP key: www.vulnscan.org/pubkey.asc
PGP fp: 8DD4 437E 9BA8 09AA 0A8D 1811 E1C3 D65F E6ED 2AA2
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (MingW32)
iD8DBQFHD2AE46ioc5305a8RAjgrAJwK9GDc1femjJiuSGW7gFMb3pSmUACgqRw5
Sjc2PNRoWFPOfMYKaKXm4o8=
=f89D
-----END PGP SIGNATURE-----
Received on 2007-10-12