DragonFly kernel List (threaded) for 2003-11
Re: Bind update
:This question comes up everytime we look to patch or update some vendor
:type of software
:in the cvs tree. I don't think we are any were near close enough to look
:at hacking up the
:contrib dir, at least not for several months.
:I can't find a pressing reason to remove it as it would require a large
:amount of work for a proper removal. I know, I know bind-8 has some
:serious, unfixable issues withit the biggest of which being that NS glue
:and additional-record returns are not properly separated out from
:offical glue and official record data in internal structures and can
:poison the DNS cache.
:I also don't want to rely on using the bind9 from ports as there is not
:real support under DragonFly for the ports tree at this time. Only hacks
:that we've added on to help make life better. At this FreeBSD is going
:in a different direction than DragonFly and the ports tree is
:being tuned for FreeBSD and hence may end up breaking builds of software
:cleanly from the vendors website. I've also noticed that the bind-9 port
:does not properly
:install the new encrypted command/management system.
Plus all of these mechanisms are eventually going to move to an IPC
messaging API. e.g. password lookups, group lookups, resolver lookups,
etc. Once these mechanisms are moved to an IPC messaging API we will
be able to change the backend (the IPC service representing the
resolver, for example) without having to recompile or reinstall any
other programs. There is no reason to complicate our lives now when
we will eventually be going to this more modular model later. Once
the new model is in place it will be far easier to rip out portions
of the system and replace them.