CNAME records having MX

[Pete Ehlke]

* glen herrmannsfeldt <gah at ugcs.caltech.edu> said, on [011215 00:11]:
> 
> As my project for today has been debugging Java programs, I wonder
> when we will have a BIND written in Java.
> 
{ I swear this is not a troll. Really. I'm honestly interested in the
answer...}

What is it about Java programmers that compels them to rewrite
*everything* in Java? And worse, what is it about people with one or two
years of Java experience that compels them to implement, on their own
and from scratch, complex, protocol-driven clients and servers?  In the
past year, I've seen a java ftp client that failed if it wasn't talking
to a Sun in.ftpd, a java smtp server that didn't understand MX
preferences, and a different java smtp server that fell over dead if there
was more than one A record for its target host. IME with Java
programmers, there is an unfortunate tendency to reimplement the wheel,
even when pre-defined, tested, and debugged classes already exist for
the task at hand.

This on top of countless overly complex, fragile file-manipulation 
utilities that could have been done better and faster in under a 
hundred lines of perl.

So I guess my question is this: what would JavaBIND get you? Why would
you want to rewrite BIND in Java? What is the advantage? If you were
proposing this as a project to your management, what would be the
business case in favor of doing it?

-Pete