Re: Modular DNS resolution

[Bryce McKinlay]

On Sunday, Oct 5, 2003, at 23:25 Pacific/Auckland, Guilhem Lavaux wrote:

> I would like to introduce in classpath a concept which used to be in 
> kaffe but was recently removed because of the merging with classes 
> from classpath: DNS solver modularity. Previously it was possible to 
> specify to kaffe (using -Ddns.resolve=class or something like this). 
> That used to give an advantage: use a native method (using e.g. 
> gethostbyname) or a Java method to resolve internet name. The java 
> method is already implemented and we can get a lot of other useful 
> insights to resolved name than we can get with gethostbyname. As the 
> code to add if fairly simple (add an interface and a if branch at 
> class initialization), I think it can be added painlessly.

I think we want to stick with always using the OS's gethostbyname(). 
This ensures that the OS settings for name resolution get respected by 
Java - I suspect things like /etc/hosts, and other name resolution 
schemes like zeroconf/rendevous, etc may not get recognized by a Java 
resolver without additional Java-specific configuration.

The "additional insights" would certainly be useful for applications 
that need to integrate tightly with DNS, however I don't think it would 
be useful or neccessary to put this inside InetAddress itself - no more 
information can be obtained given the existing InetAddress interface, 
right?

Surely it would be better to put this code in its own class under gnu.* 
somewhere? Note that external resolvers can now easily create 
InetAddress objects using the JDK 1.4 getByAddress() methods, which are 
already implemented in classpath.

Regards

Bryce.