kenji@ybbs.c.u-tokyo.ac.jp (Kenji Rikitake) (12/21/89)
I've been co-writing a USENET-news-compatible MS-DOS news processing
program with a staff member of my packet radio group
(Packet Radio User's Group), called "Terakoya" system.
Our news poster seems to be fairly compatible with Bnews 2.11.19,
except for control messages. The postmaster of my domain-master site
told me that control messages made by our programs did not work on
Bnews 2.11.19. He and I worked together to solve the problem. And found:
Bnews 2.11.19 requires following conditions to acceptable control messages:
1. Subject: field should appear before Control: field.
2. Approved: field should appear as the end of header fields
(i.e. after Lines: field).
Our news poster did totally opposite things to the above criteria.
Questions: do any implicit requirements exist on each header field?
If yes, why? And how?
-- Kenji
--
Kenji Rikitake, Sakurajousui R&D Center, Packet Radio User's Group (PRUG)
72407.524@compuserve.com/kenji%dctwcs.das.net@sun.com/kenji@Nikkei_MIX
kenji@ybbs.c.u-tokyo.ac.jp/AMPR:kenji@jj1bdx.ampr{.org|.jp}[44.129.16.82]tale@cs.rpi.edu (David C Lawrence) (12/21/89)
In <1452@ybbs.c.u-tokyo.ac.jp> kenji@ybbs.c.u-tokyo.ac.jp (Kenji Rikitake): > Bnews 2.11.19 requires following conditions to acceptable control messages: > 1. Subject: field should appear before Control: field. > 2. Approved: field should appear as the end of header fields > (i.e. after Lines: field). > Questions: do any implicit requirements exist on each header field? > If yes, why? And how? RFC 1036, the Standard for Interchange of USENET Messages, contains the information regarding headers. It does not make any mention of required header order. The C News authors point out that that processing of messages could be sped up even a little more if Control were required to be the first header line when present. Dave -- (setq mail '("tale@cs.rpi.edu" "tale@ai.mit.edu" "tale@rpitsmts.bitnet"))
henry@utzoo.uucp (Henry Spencer) (12/21/89)
In article <1452@ybbs.c.u-tokyo.ac.jp> kenji@ybbs.c.u-tokyo.ac.jp (Kenji Rikitake) writes: >Bnews 2.11.19 requires following conditions to acceptable control messages: >1. Subject: field should appear before Control: field. >2. Approved: field should appear as the end of header fields > (i.e. after Lines: field). > ... >Questions: do any implicit requirements exist on each header field? RFC1036 is silent about the matter, so RFC822 sets the rules. RFC822 is quite explicit: the order of headers is arbitrary. That's the theory. In practice... to the best of our knowledge, C News conforms to the standard and will work with any order. I can't answer for B News, I'm not sufficiently familiar with its insides. -- 1972: Saturn V #15 flight-ready| Henry Spencer at U of Toronto Zoology 1989: birds nesting in engines | uunet!attcan!utzoo!henry henry@zoo.toronto.edu
storm@texas.dk (Kim F. Storm) (12/29/89)
kenji@ybbs.c.u-tokyo.ac.jp (Kenji Rikitake) writes: >Bnews 2.11.19 requires following conditions to acceptable control messages: >2. Approved: field should appear as the end of header fields > (i.e. after Lines: field). Does that apply to control messages only, or is it the explanation why postings to comp.sys.sun ends up in junk here (Bnews 2.11.14)? Here is a sample header of one of the junk'ed articles, showing that an X- line follows Approved (which otherwise "correctly" follows Lines): Path: tidk!dkuug!sunic!mcsun!uunet!seismo!ukma!mailrus!cornell!uw-beaver!rice!sun-spots-request From: cher@ksr.com (Mike Cherepov) Newsgroups: comp.sys.sun Subject: Name Lookup Cache: BSD vs SunOS Keywords: Miscellaneous Message-ID: <1509@brazos.Rice.edu> Date: 6 Sep 89 19:38:31 GMT Sender: root@rice.edu Organization: Sun-Spots Lines: 27 Approved: Sun-Spots@rice.edu X-Sun-Spots-Digest: Volume 8, Issue 123, message 8 of 17 -- Kim F. Storm storm@texas.dk Tel +45 429 174 00 Texas Instruments, Marielundvej 46E, DK-2730 Herlev, Denmark No news is good news, but nn is better!
chip@chinacat.Lonestar.ORG (Chip Rosenthal) (12/30/89)
In article <431@texas.dk> storm@texas.dk (Kim F. Storm) writes: >[Re: location of Approved: in header] >Does that apply to control messages only, or is it the explanation why >postings to comp.sys.sun ends up in junk here (Bnews 2.11.14)? No. I believe it's due to the inews staleness check. If a message is too old, inews will junk it to prevent old articles which have already been expired from looping. This is an issue in moderated groups, because when you add up all the various delays, the "Date:" in the original mail message can place it beyond this threshold. That's why the "brkdig" program I just posted to comp.sources.misc (which breaks up digests into articles) checks the "Date:" field and strips it if too old.