$Id: TODO,v 4.9.1.4 1994/02/21 08:09:46 vixie Exp $

Things to do.  Each entry should contain the proposer, date proposed,
and an explaination of what's being proposed.  New ones are added at
the bottom.  Note that the author/coordinator of BIND does not 
neccessarily endorse all of the proposals listed herein; if you did
not get explicit "buy-in" then your changes may not be accepted even
if they appear in proposal form here in this file.

[vixie@vix.com 06Jan93]: syntax checking in db_load.c
	currently there is little or no syntax checking in db_load.c.
	you can see this in action if you add an HINFO RR with only
	one word of info -- the second field in the database is trash,
	and is sent out in zone transfers or responses as trash.  such
	records should be ignored as errors much as "cname and other data"
	is.  (kre@munnari.oz.au comments: This should turn into "rewrite
	db_load.c" - probably using yacc or something (though that doesn't
	matter) so that it approximates the syntax its supposed to implement.)

[vixie@vix.com 25apr93]: clean up #ifdef's and portability
	feature #ifdef's should be limited to whole functions, which will be
	called no matter what and would only be non-empty if the feature is
	enabled.  allow feature ifdef's in .h files, though.

	portability #ifdef's should be limited to whole functions, too.  add
	a new portability.c module that implements anything which varies from
	system to system.

	add a second portability.h-like file that is included _before_ all the
	system includes.  portability.h as it stands is included _after_ all
	system includes, which is convenient for most things but not all.

[sater@cs.vu.nl 26apr93]: sortlist improvement
	Improve the code around the sortlist area to better cope with parallel
	networks of different speeds. The -i hack I sent to you could function
	as inspiration only.

[kre@munnari.oz.au 26apr93]: add an INN style control interface
	to replace sending signals.  With that expand debugging to
	permit monitoring of actions taken on a single query
	(query through control port, full traced as it occurs)
	or all queries that reference some particular name or
	zone, or which are forwarded, or asked, of some
	particluar server.   Allow reloads & dumps of a single
	zone, rather 	than the whole universe.  Allow selective
	cache pruning (to edit away bad data that's been obtained
	from somewhere)

[kre@munnari.oz.au 26apr93]: add a syntax to zone files (non rfc
	standard, but I don't care) to permit RR's to age away
	at some particular time, and others to become active at
	some particular time (probably with a syntax something
	like 	"<[date]"  or  "@[date]"   preceding, or in the
	former case, replacing, the TTL field of the record).
	Approaching "date" in the "<[date]" case, the TTL's on
	the record would be decreased, so no data cached anywhere	
	will remain valid after "date", after "date", this RR
	would simply be inoperative (essentially identical to
	a comment).  In the "@[date]" case (or perhaps ">[date]"
	for symmetry) the RR would be ignored until "date" at
	which time the "@[date]" field would simply be ignored.
	Both annotations could be used together (with
	appropriate interpretations depending on which date is
	earlier than the other).   Annotations on RR's in a zone
	would cause the SOA parameters to be automatically
	adjusted in zone transfers (and SOA requests) so that
	secondary servers would also hand out the same values
	(dropping the TTL down low as a "<[date]" approaches,
	and forcing a new zone transfer at "date").

[steve@uunet.uu.net 26apr93]: TXT RR improvements
	- fix TXT records so that they can deal properly with multiple
	strings (e.g., ``foo	IN	TXT	"aaa" "bbb"'').  This
	results in a fair number of smallish changes throughout the
	code and also throughout various tools (e.g., nslookup).

[steve@uunet.uu.net 26apr93]: X25, ISDN, RT support
	- add X25, ISDN, and RT records, for sake of completeness (I already
	have code that should do this, at least for 4.8.3).  At least, I
	figure this should go in unless someone nuked these RR types while
	I wasn't looking... (-:

[steve@uunet.uu.net 26apr93]: core dump on invalid address formats
	- dotted quints (A.B.C.D.E) in A records may make named dump core.

[vixie@vix.com 02may93]: better nameserver tracking, plus data tagging
	Currently every A RR in the online database (cache or zones in
	memory) has some fields which are used to track attributes of the
	name server if this A RR happens to be that of a name server.  This
	is neccessary since we need to keep track of nameserver RTT and sort
	our queries appropriately to try "fast" or "close" nameservers before
	we try distant or slow ones.  However, due to the way C structures
	work, this information is maintained for other RR types and for A RR's
	which are not addresses of nameservers.  I propose adding a separate
	database that tracks neighboring nameservers; no field would be needed
	in a databuf (RR) to point to this since we can search it by address.
	Multi-homed nameservers (i.e., with nameservers on hosts which have
	more than one network interface) would be single-datum, multiple-key.
	Information to be retained about other nameservers would include RTT
	as well as error statistics, packet statistics, and so on.  This
	database would be dumpable in the same way that the main cache is
	now dumpable -- that is, using a signal.  At some point we can
	consider adding SNMP query capability to this database.  But most
	importantly, we can keep in each databuf (RR) the source address of
	the nameserver which sent us the data -- thus allowing the cache
	dump to include a tag on each RR that says where it came from.  This
	will help a lot in tracking down corrupt data.  Other than continuing
	to manage the source pool and integrate other folks' patches, this
	feature may be my only contribution to BIND 4.9.1.

[kyle@uunet.uu.net 16may93]: need an option to die if primary zone file missing
	as of 4.9, a server will not forward a query if it is itself on the
	NS list for the relevant domain.  this means that if a primary server
	cannot load its zone file, it will not be able to answer queries in
	that zone -- it won't even forward them.  this is arguably correct,
	since it prevents bad forwarding loops when two or more servers are
	all unable to load the zone (primary or secondary, with secondary
	failures being the more common).  what is needed is real loop detection
	such that reasonable non-looping queries can be forwarded.  what we're
	likely to actually get is an option that causes named to just syslog
	and die if it can't load a primary zone file.  note that at present,
	named is running somewhat bare-assed since an expired zone in a
	secondary (or missing zone file in a primary) will cause that named
	to return SERVFAIL for all queries to that zone.  if your screwed up
	primary/secondary server is also the forwarding server for a collection
	of hosts, those hosts will get SERVFAIL's back from queries to the
	affected domains, and depending on the age of their resolvers, they
	might not try other servers after they get the first SERVFAIL.
	[ this entry was written by Paul Vixie after getting a problem report
	  from Kyle after uu.net disappeared in a brief but ugly way.  --vix ]

[paul@vix.com 19feb94]: things on the list that i'm expecting to fix in 4.9.3:
	-> svr4 changes from wisner@well, marc@cam, istewart@datlog
	-> too many open files on ultrix, sunos (orion@iastate, g.michaelson)
	-> sequence number arithmetic (kre@oz, sra@epilogue)
	-> convex patches from ukkonen@csc.fi
	-> switch completely to posix-style signals
	-> xfrnets directives should aggregate
	-> syntactic sugar to use "mtime" of file as soa serial number
	-> resolvsort fixes from kre@oz
	-> better support for "firewalls" (zohar@ibm, minnich@dupont)
	-> AIX SIGDANGER support
	-> attributes in TXT RR (cpw@lanl)
	-> fix database consistency problems during zone reloads
	-> preliminary support for variable width subnet masks
	-> CLEANCACHE code from marka@csiro
	-> corrected hostname(7) man page from ash@hp
	-> specify ip source address on udp packets (thakur@harvard, ash@hp)
	-> handle 0 serial numbers (leres@lbl)
	-> failover isn't working very well for hesiod queries (gshapiro)
	-> quite a few cleanups to res/gethnamadr.c
	-> promote bryan@umich's contrib/umich/query to tools/dnsquery
	-> res_send fix for sunos (maybe others) from mark andrews.

## ++Copyright++ 1993
## -
## Copyright (c) 1993
##    The Regents of the University of California.  All rights reserved.
## 
## Redistribution and use in source and binary forms, with or without
## modification, are permitted provided that the following conditions
## are met:
## 1. Redistributions of source code must retain the above copyright
##    notice, this list of conditions and the following disclaimer.
## 2. Redistributions in binary form must reproduce the above copyright
##    notice, this list of conditions and the following disclaimer in the
##    documentation and/or other materials provided with the distribution.
## 3. All advertising materials mentioning features or use of this software
##    must display the following acknowledgement:
## 	This product includes software developed by the University of
## 	California, Berkeley and its contributors.
## 4. Neither the name of the University nor the names of its contributors
##    may be used to endorse or promote products derived from this software
##    without specific prior written permission.
## 
## THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
## ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
## IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
## ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
## FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
## DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
## OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
## HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
## LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
## OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
## SUCH DAMAGE.
## -
## Portions Copyright (c) 1993 by Digital Equipment Corporation.
## 
## Permission to use, copy, modify, and distribute this software for any
## purpose with or without fee is hereby granted, provided that the above
## copyright notice and this permission notice appear in all copies, and that
## the name of Digital Equipment Corporation not be used in advertising or
## publicity pertaining to distribution of the document or software without
## specific, written prior permission.
## 
## THE SOFTWARE IS PROVIDED "AS IS" AND DIGITAL EQUIPMENT CORP. DISCLAIMS ALL
## WARRANTIES WITH REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES
## OF MERCHANTABILITY AND FITNESS.   IN NO EVENT SHALL DIGITAL EQUIPMENT
## CORPORATION BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL
## DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
## PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
## ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS
## SOFTWARE.
## -
## --Copyright--
