home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Internet Info 1994 March
/
Internet Info CD-ROM (Walnut Creek) (March 1994).iso
/
inet
/
ietf
/
nntp
/
nntp-minutes-91jul.txt
< prev
next >
Wrap
Text File
|
1993-02-17
|
9KB
|
209 lines
CURRENT_MEETING_REPORT_
Reported by Eliot Lear/Intelligenetics
NNTP Minutes
There were three meetings of the Network News Transport Protocol Working
Group.
There was also an area meeting which included the NNTP, LIST, 822-EXT,
and SMTP Chairs, along with the Area Director. This is to say, Eliot
Lear, David Lippke, Greg Vaudreuil, and Russ Hobby.
The following items were explored at the various meetings:
1. Differences Between Mail and News.
We should consider moving towards a common user interface between
mail and news. Similar message formats have made this possible in
the past. With the advent of a new message format for mail, news
will need to adopt some similar standard pretty quickly (ARE YOU
READING THIS, NEWS READER PEOPLE?????). There was discussion of
moving to unite the news and mail formats. While conceptually it
sounds like a good thing, the details need to be kinked out, and
the question needs to be discussed to death.
Russ asked what a document be put out that describes the current
news architecture. Erik Fair has volunteered to write that
document. That document should almost certainly include a safe way
to gateway news and mail. Conceivably another document will issue
from the area recommending a course of action.
2. News Reader Capabilities.
First, it was the consensus of the group that this topic is really
part of the Charter of the NNTP group; we're just considering
splitting new functionality into a separate protocol.
The current version of the draft contains very little enhancements
in the area of news reader capabilities. This is because the Chair
does not have any real concrete language in front of him from what
this group wants. The consensus, therefore, was to push on with
the transport document,and explore further the reader issues, and
in particular how this relates to Item 1.
If we do produce an NNRP document, we must be careful that by its
nature it would steer development away from useful areas (Ittai
Hershman's paraphrased comments). In this vein, if we do produce a
document, we should consider it an experimental effort rather than
a standards track effort.
Along the lines of a news reader protocol, Stan Barber brought
along a one page shopping list of items he would like to see in a
1
reader protocol. We discussed how to define a search command so
that it would be generally useful. Arguments for and against a
specific syntax and mechanism were heard.
3. Authentication
Theodore Tso is now the official ``stuckee'' for the SAAG in the
NNTP Working Group. Issues of Common Authentication Technology
(CAT) were discussed, particularly at the Thursday meeting. Text
needs to be written into the document to take advantage of CAT. We
are facing a problem with CAT because NNTP is one of the first
protocols to use it. Currently CAT can only be used to access
Kerberos and DEC SPX. Jeff Schiller suggested that a simple
challenge/response method would be acceptable if someone did the
footwork. Clear text, however, seemed to be right out, to the
point where it was thought that the SAAG might hold things up.
Jeff also discussed the evils of negotiating security methods.
It turns out that some of the logic that was applied to mail
standards can be applied to news. If we do, in fact, move the
transport document to proposed standard, the impetus for
authentication in the transport is greatly diminished.
4. Transporting Binary and Mixed Message Format.
It turns out that simply adopting the mail standards as news
standards may be a bit painful. With the introduction of binary,
there needs to be a new canonical form. This in itself would be a
minor irritation; however, the new mail format allows for mixed
binary and text. This means that it could be necessary to switch
between binary and text canonical forms in a single message. This
makes transport a nightmare, and is a good argument for encoding.
On the other hand, possibly the new binary canonical form might be
able to handle the problems. Interested parties are URGED to read
the draft mail documents and the archive of messages leading to
their production.
5. CCITT
Harri Salminen circulated a draft document that is CCITT's version
of netnews. The document may be retrieved from nic.nordu.net, via
anonymous FTP. Your comments are, of course, solicited.
6. Problems with the Current Document
Several people have sent notes pointing out formatting problems,
grammatical errors, and certain inconsistencies (like SIMPLE
authentication descriptions). Please mail all such complaints
directly to the Chair, and not to the list.
It was the consensus of the group that the IMAGE and BINARY options
be combined into a FORMAT option. Eliot Lear will write some text
in to this effect. It was also agreed that the COMPRESSION and
DATE commands would be removed, and that the NEWNEWS command be
extended to deal with DATE's purpose (which is to say that NEWNEWS
will both accept and deliver a cookie instead of a date). Text to
2
be written and argued.
State diagrams need to be completed.
Default behavior needs to be defined and mandated.
We discussed eliminating the OPTION command. The problem with
eliminating the OPTION command is that it gets hard to batch verbs,
and we concluded that batching such things was a good idea.
7. Making the IETF lists available to the IETF via netnews.
This issue was brought up both in the Working Group and in the area
meeting. Some action is expected in this area Real Soon Now (tm).
Social issues were discussed in the Wednesday meeting regarding the
perceived stigma from which news suffers.
8. News MIB
Russ Hobby stated that he would not require a News MIB from us.
However, several people have indicated some interest in managing
news objects, particularly Jim Thompson (not present at Atlanta).
Jim should proceed to take comments and write up a document. One
should be careful to study which functions are ubiquitous
throughout the Internet, and which are implementation specific.
9. Timetable
August 31, 1991 - We would like to see the NNTP document become an
Internet Draft. All this does is expose the document to the
Internet community. It can be changed from within the Working
Group after that point.
November, 1991 - Get architecture document out as an informational
RFC.
December, 1991 - After considering comments and making appropriate
changes, let the NNTP document proceed to proposed standard.
It is hoped that some code will be ready by December.
10. Multicasting Netnews
Brief mention was made on a research effort under way to explore
the possible use of multicast packets as a way for distribution of
news. Interested parties should contact the Chair directly.
11. Next Meeting
No next meeting date has been set as of yet. Depending on how we
proceed with a news message format, we may meet at Interop
(October).
12. General Information
If you wish to be added to the ietf-nntp mailing list, you should
send mail to ietf-nntp-request@turbo.bio.net.
3
Drafts and message archives can be gotten from turbo.bio.net via
anonymous FTP in the ietf-nntp directory. The format of draft
document filenames is documentname.format-type.
Attendees
Stan Barber
Robert Enger enger@seka.scc.com
Erik Fair fair@apple.com
Ned Freed ned@innosoft.com
Olafur Gudmundsson ogud@cs.umd.edu
Ittai Hershman ittai@nis.ans.net
Russ Hobby rdhobby@ucdavis.edu
Neil Katin katin@eng.sun.com
Eliot Lear lear@turbo.bio.net
David Lippke lippke@utdallas.edu
Joseph Malcom jmalcom@sura.net
Chris Myers chris@wugate.wustl.edu
Michael Patton map@lcs.mit.edu
Mel Pleasant pleasant@hardees.rutgers.edu
Jan Michael Rynning jmr@nada.kth.se
Harri Salminen hks@funet.fi
Theodore Tso
Gregory Vaudreuil gvaudre@nri.reston.va.us
4