spaf@cs.purdue.EDU (Gene Spafford) (09/09/89)
This is a terrible kludge, but I am sooo tired of seeing those
bogus "Re^2:" subject lines.... They aren't the proper format
for follow-up subjects according to the RFC, and they don't track
with "rn" when trying to read news. I was trying to catch up on
a few weeks of back news and kept reencountering threads of
discussions broken by that obnoxious "Re^2:"
So, I whipped up the enclosed program and installed it in our news
library with the following sys file line:
complain:world,all:BU:/usr/lib/news/complain < %s
It sends mail to the author of the article complaining about the
non-standard, non-conformant use of that string. If enough of you
out there install it on your systems too, maybe the volume of mail
will convince people to fix the software.
As I said, it's a hack -- it works on Sequents and Suns, but no claims
that it will work anywhere else. It also assumes you have
/usr/lib/sendmail as your mail agent. Feel free to modify it to fit
your system.
#include <stdio.h>
#include <ctype.h>
extern char *malloc ();
extern int strlen ();
extern FILE *popen ();
#define NEWSTRING(ptr) {if((ptr = malloc(slen+1)) == NULL) \
return (-1); else strcpy(ptr, cptr);}
main ()
{
char *reply = NULL,
*sender = NULL,
*from = NULL,
*messageid = NULL;
register char *cptr;
register int slen;
char tbuf[512];
while (fgets (tbuf, 512, stdin) != NULL)
{
if (*tbuf != '\n' && isspace(*tbuf))
continue;
else if (*tbuf == '\n')
break;
/* find the : and convert header to lower case */
for (cptr = tbuf; *cptr && *cptr != ':'; cptr++)
if (isupper (*cptr))
*cptr = tolower (*cptr);
/* If no : found, we are out of the header */
if (!*cptr)
break;
*++cptr = '\0';
cptr++;
while (*cptr && *cptr == ' ')
cptr++;
slen = strlen (cptr);
if (cptr[slen - 1] == '\n')
cptr[--slen] = 0;
if (strcmp ("subject:", tbuf) == 0)
{
if (!isbadsub (cptr))
return (0);
}
else if (strcmp ("sender:", tbuf) == 0)
NEWSTRING(sender)
else if (strcmp ("reply-to:", tbuf) == 0)
NEWSTRING(reply)
else if (strcmp ("from:", tbuf) == 0)
NEWSTRING(from)
else if (strcmp ("message-id:", tbuf) == 0)
NEWSTRING(messageid)
}
/* If we get this far, we have a bogus subject line.
* We also have a reply address we can use to send
* mail to tell them about it.
*/
if (reply)
complain (reply, messageid);
else if (sender)
complain (sender, messageid);
else if (from)
complain (from, messageid);
else
return (-1);
return (0);
}
int
isbadsub (subjectline)
char *subjectline;
{
int i;
/* only check here is for first 4 characters being "Re^2" */
for (i = 0; i < 2; i++)
if (isupper (subjectline[i]))
subjectline[i] = tolower (subjectline[i]);
return (!strncmp (subjectline, "re^2:", 5));
}
#define OUT(ss) fputs(ss, mailer)
complain (addressee, messageid)
char *addressee,
*messageid;
{
FILE *mailer;
char hostname[64];
gethostname(hostname, 64);
mailer = popen ("/usr/lib/sendmail -t", "w");
if (mailer == NULL)
{
perror ("Cannot open mailer");
exit (-1);
}
fprintf (mailer, "To: %s\n", addressee);
OUT("Subject: Invalid format in Usenet posting\n");
OUT("Reply-to: nobody\n");
fprintf(mailer, "From: news@%s\n\n", hostname);
fprintf (mailer, "Your Usenet posting with message id %s\n", messageid);
OUT("has an invalid subject line, as defined by RFC 1036, the ");
OUT("Usenet format\nstandard document.\n\n");
OUT("Your posting software is either broken or producing articles\n");
OUT("that do not conform to the required standard format.\nPlease ");
OUT("fix your software, or discontinue its use.\n\nThis was an ");
OUT("automated message.");
pclose (mailer);
}
--
Gene Spafford
NSF/Purdue/U of Florida Software Engineering Research Center,
Dept. of Computer Sciences, Purdue University, W. Lafayette IN 47907-2004
Internet: spaf@cs.purdue.edu uucp: ...!{decwrl,gatech,ucbvax}!purdue!spafbob@tinman.cis.ohio-state.edu (Bob Sutterfield) (09/09/89)
For a bit more generality, in case multiple old-NN users get to chattering at each other: *** src/complain.c~ Fri Sep 8 15:23:44 1989 --- src/complain.c Fri Sep 8 15:25:23 1989 *************** *** 120 **** ! /* only check here is for first 4 characters being "Re^2" */ --- 120 ---- ! /* only check here is for first 3 characters being "Re^" */ *************** *** 126 **** ! return (!strncmp (subjectline, "re^2:", 5)); --- 126 ---- ! return (!strncmp (subjectline, "re^", 3));
fritz@unocss.UUCP (Tim Russell) (09/09/89)
spaf@cs.purdue.EDU (Gene Spafford) writes: >So, I whipped up the enclosed program and installed it in our news >library with the following sys file line: >complain:world,all:BU:/usr/lib/news/complain < %s >It sends mail to the author of the article complaining about the >non-standard, non-conformant use of that string. If enough of you >out there install it on your systems too, maybe the volume of mail >will convince people to fix the software. Boy, Spaf, I always thought you could be a jerk at times, but this really tops all. Nothing like wasting net bandwidth to send mail to an audience in which: 1) The vast majority probably have no idea that their software is even performing this "unspeakable" act. 2) Probably 99% have no control over the software. Not everyone has root access on their machine, Spaf, or is able to compile their own copy of nn just to support your rn-bigotry. Yes, nn had this "feature" at first, and we (the nn-using community) voted to get rid of it. But there are busy sysadmins out there who don't keep up, and it's hardly fair to bug their unsuspecting users with potentially hundreds of messages just because of this. I should know, our admin is that way, and if nn wasn't a user-supported program I have no doubt that we would still be at that patchlevel. One would hope that such flagrant abuse of the network would be avoided by someone as "professional" as you, but instead you're setting the worst kind of example, that of tying up network resources all over the country just because you're annoyed that you can't follow a message chain. Oh boo-hoo, Spaf, you poor poor man. I hope every single message your cute little program sends bounces and fills up your disk and you miss your news completely. You're a jerk. -- ---------------------------------+-------------------------------------------- Tim Russell, Computer Operator | Internet: russell@zeus.unl.edu Campus Computing | Bitnet: russell@unoma1 University of Nebraska at Omaha | UUCP: uunet!zeus.unl.edu!russell
rsk@boulder.Colorado.EDU (Rich Kulawiec) (09/10/89)
In article <1650@unocss.UUCP> fritz@unocss.UUCP (Tim Russell) writes: > I hope every single message your cute little program sends bounces and >fills up your disk and you miss your news completely. You're a jerk. No, the "jerk(s)" are those folks who coded a flagrant violation of the RFC into nn, and those folks that continue to run it in its broken state. And since *you're* the one who's so concerned about network resources, you might want to multiply the size of your article by the guesstimated number of sites on Usenet; then estimate how many thousands of letters Gene's program will have to generate before it even approaches that figure. Actually, my solution to the problem is rather effective, too: modify news 2.11 so that it silently discards non-compliant articles. -- Rich Kulawiec, rsk@boulder.colorado.edu, boulder!rsk
pst@anise.acc.com (Paul Traina) (09/10/89)
rsk@boulder.Colorado.EDU (Rich Kulawiec) writes: >No, the "jerk(s)" are those folks who coded a flagrant violation of the >RFC into nn, and those folks that continue to run it in its broken state. B---shit. Subject is a user controllable field. rn(1) is broken because it uses that field to follow message chains. It was a kludge to begin with; there is no responsible way of denying that fact. Nn did "the wrong thing" according to many people--I don't disagree with you, but the "blame" (and I use that term loosely) rests entirely on that design compromise in rn(1). Before bitching, why not fix your reader? Lazy bones. -- Reclaim those words you're afraid of. There's nothing wrong with being a pervert and/or slut. We're very special people. The best. Take pride in it. -- Hank B. (but I wish I had said it first)
peter@ficc.uu.net (Peter da Silva) (09/10/89)
My first reaction to Spaf's posting was "this has got to be a forgery". I mean, surely he of all people would understand that this sort of net.terrorism ia an amazingly bad idea. Until I hear otherwise I'll stick with this assumption. In article <11509@boulder.Colorado.EDU>, rsk@boulder.Colorado.EDU (Rich Kulawiec) writes: > Actually, my solution to the problem is rather effective, too: modify > news 2.11 so that it silently discards non-compliant articles. This is also Evil and Rude. People have been using all sorts of weird reply formats (Re:, re:, Re^2:, Re^N:, etc...) for longer than NN has been around. Having NN automatically violate the RFC is uncool, but dumping NN generated messages is worse. The NN problem has been going away as more and more people upgrade to the new software, anyway. All you do with tactics like this is make a bad situation worse. I haven't seen such a flagrant violation of net ethics since some idiot set up an auto-flame that would automatically generate a flame in response to certain users' articles. -- Peter da Silva, *NIX support guy @ Ferranti International Controls Corporation. Biz: peter@ficc.uu.net, +1 713 274 5180. Fun: peter@sugar.hackercorp.com. `-_-' "...the TV reporters, who are as intelligent as electric toasters" 'U` -- Clayton E. Cramer
tneff@bfmny0.UU.NET (Tom Neff) (09/11/89)
Well, if you want the OUTRAGEOUS suggestion, how "uncool" would it be to
mod B news to just FIX the mutant headers on their way through?!
--
Annex Canada now! We need the room, \) Tom Neff
and who's going to stop us. (\ tneff@bfmny0.UU.NETrsk@boulder.Colorado.EDU (Rich Kulawiec) (09/11/89)
In article <6115@ficc.uu.net> peter@ficc.uu.net (Peter da Silva) writes: >In article <11509@boulder.Colorado.EDU>, rsk@boulder.Colorado.EDU (Rich Kulawiec) writes: >> Actually, my solution to the problem is rather effective, too: modify >> news 2.11 so that it silently discards non-compliant articles. > >This is also Evil and Rude. People have been using all sorts of weird >reply formats (Re:, re:, Re^2:, Re^N:, etc...) for longer than NN has been >around. Having NN automatically violate the RFC is uncool, but dumping NN >generated messages is worse. And this solution has been around longer, too; it predates NN by a long time. I don't see how you can argue that it's "Evil and Rude" , unless you'd like to assert that all sites are under some sort of obligation to pass malformed articles, and/or that getting news 2.11 to enforce the RFC is somehow unacceptable. -- Rich Kulawiec, rsk@boulder.colorado.edu, boulder!rsk
skrenta@blekko.UUCP (Rich Skrenta) (09/11/89)
In article <1650@unocss.UUCP> fritz@unocss.UUCP (Tim Russell) writes: > 2) Probably 99% have no control over the software. Not everyone has > root access on their machine, Spaf, or is able to compile their > own copy of nn just to support your rn-bigotry. It's not rn-bigotry. It's rn, notes, or any newsreader that tries to follow threads based on the subject line. > Yes, nn had this "feature" at first, and we (the nn-using community) voted > to get rid of it. How nice of you, voting to make your newsreader obey the rules. > I should know, our admin is > that way, and if nn wasn't a user-supported program I have no doubt that > we would still be at that patchlevel. Sounds like you've got a great sysadmin. "Gee, this software is broken and barfing all over the net, but I don't think I'll upgrade it." Rich
peter@ficc.uu.net (Peter da Silva) (09/11/89)
[regarding silently trashing articles that have unusual subject lines (Re^n:, etcetera).] In article <11521@boulder.Colorado.EDU>, rsk@boulder.Colorado.EDU (Rich Kulawiec) writes: > I don't see how you can argue that it's "Evil and Rude" , unless you'd > like to assert that all sites are under some sort of obligation to pass > malformed articles, and/or that getting news 2.11 to enforce the RFC is > somehow unacceptable. Strictly speaking you're correct. Perhaps "evil and rude" is incorrect. A better description of this sort of activity is "stupid". People should not be literal minded idiots dancing to the tune of an admittedly flawed RFC. Subject lines are created by people, not a rigid peice of software. Zapping articles because a subject line is a slight variant of the RFC makes as much sense as zapping an article because the keywords are poorly chosen or because the signature is longer than 4 lines. -- Peter da Silva, *NIX support guy @ Ferranti International Controls Corporation. Biz: peter@ficc.uu.net, +1 713 274 5180. Fun: peter@sugar.hackercorp.com. `-_-' "...the TV reporters, who are as intelligent as electric toasters" 'U` -- Clayton E. Cramer
gsh7w@astsun.astro.Virginia.EDU (Greg S. Hennessy) (09/11/89)
In article <6118@ficc.uu.net> peter@ficc.uu.net (Peter da Silva) writes:
#Zapping
#articles because a subject line is a slight variant of the RFC makes as much
#sense as zapping an article because the keywords are poorly chosen or because
#the signature is longer than 4 lines.
#--
Why knows, that might be the next to go!
-Greg Hennessy, University of Virginia
USPS Mail: Astronomy Department, Charlottesville, VA 22903-2475 USA
Internet: gsh7w@virginia.edu
UUCP: ...!uunet!virginia!gsh7wpda@stiatl.UUCP (Paul D. Anderson) (09/11/89)
rsk@boulder.Colorado.EDU (Rich Kulawiec) writes: >No, the "jerk(s)" are those folks who coded a flagrant violation of the >RFC into nn, and those folks that continue to run it in its broken state. Actually, I'd like to see the RFC. I suspect that it deals with what may be in the header region, but not with the contents of the Subject: field. Would someone please post that RFC to this group, so that this may be judged fairly? Then we'll decide what to do... paul -- Pa3ul Anderson (w) (404) 841-4000 gatech!stiatl!pda (h) (404) 662-0799
brad@looking.on.ca (Brad Templeton) (09/11/89)
Is it so evil? Over 4 years ago, I think, some new features were added to the news software, namely the "References:" line and more recently the "Supersedes" line. In fact, those were about the last two significant improvements to the format. Years later these lines still can't be used properly because of all the people running old and non-compliant posting software. Possibly it will never be possible to use them. How you read and edit news on your own machine is your own business. But the format of what you post out to the rest of the net is the one thing on the net that's everybody's business. Actually, nn's "crime" isn't too bad. There are worse ones. But what other mechanisms are there to help in this direction? The method proposed is actually one I invented almost 7 years ago to stop the regular daily postings to net.test that happened whenever somebody brought up news on their machine. Usenet is a minarchy but there's one thing that everybody has to agree upon, and that's the basic file format. In some ways, the file format *is* usenet. -- Brad Templeton, Looking Glass Software Ltd. -- Waterloo, Ontario 519/884-7473
rsk@boulder.Colorado.EDU (Rich Kulawiec) (09/11/89)
In article <6118@ficc.uu.net> peter@ficc.uu.net (Peter da Silva) writes: >Strictly speaking you're correct. Perhaps "evil and rude" is incorrect. A >better description of this sort of activity is "stupid". People should not >be literal minded idiots dancing to the tune of an admittedly flawed RFC. Well, until someone writes another RFC to replace the existing one, it's all we've got, flaws or not -- and silently tossing an article into "junk" because of a noncompliant "Subject" line doesn't seem inherently different to me than junking it because it has no "From" line, a badly formatted "Date", or any other problem. This debate seems backward to me -- the folks who created NN are the ones who wired in the broken code, and I don't think their actions create an obligation on the rest of us to support it. >Subject lines are created by people, not a rigid peice of software. Zapping >articles because a subject line is a slight variant of the RFC makes as much >sense as zapping an article because the keywords are poorly chosen or because >the signature is longer than 4 lines. Actually the followup "Subject" lines *are* created by software, which is how we got into this in the first place. And, yes, I think junking articles with more than 4 lines of signature is an excellent idea. ---Rsk
scott@clmqt.marquette.Mi.US (Scott Reynolds) (09/11/89)
rsk@boulder.Colorado.EDU (Rich Kulawiec) writes: >...unless you'd >like to assert that all sites are under some sort of obligation to pass >malformed articles, and/or that getting news 2.11 to enforce the RFC is >somehow unacceptable. If you've read news.software.b lately you'd realize that there are quite a few people that consider the B 2.11 software, and not the RFC, to be the final authority. Of course, nobody has to pass _any_ articles through, but what does that accomplish? Why not do something constructive like offering the necessary nn patches to anyone who needs them? I've got them via anonymous UUCP if anyone wants them, e-mail me for details. There have got to be various archives to get them via ftp, also. -- Scott Reynolds rutgers!mailrus!clip!clmqt!scott Enterprise Information System scott@clmqt.marquette.Mi.US "Insufficient facts always invite danger." -- Spock
tombre@weissenburger.crin.fr (Karl Tombre) (09/11/89)
In article <6115@ficc.uu.net> peter@ficc.uu.net (Peter da Silva) writes: >My first reaction to Spaf's posting was "this has got to be a forgery". I >mean, surely he of all people would understand that this sort of net.terrorism >ia an amazingly bad idea. Until I hear otherwise I'll stick with this >assumption. 1. Gene Spafford announced himself that he went on a "speaking tour" in Europe. 2. He was announced here in France to hold a conference at INRIA Rocquencourt, Paris, on Sept. 27. It seems most likely to me that this conf is part of his "speaking tour", and hence the article posted in his name most probably is a forgery. Karl Tombre - INRIA Lorraine / CRIN EMAIL : tombre@loria.crin.fr - POST : BP 239, 54506 VANDOEUVRE CEDEX, France
flee@shire.cs.psu.edu (Felix Lee) (09/11/89)
In article <6833@stiatl.UUCP> pda@stiatl.UUCP (Paul D. Anderson) writes: > Actually, I'd like to see the RFC. RFC 1036, section 2.1.4: [...] If the message is submitted in response to another message (e.g., is a follow-up) the default subject [!] should begin with the four characters "Re:" [sic], and the "References" line is required. "default subject" is a curious phrase.. -- Felix Lee flee@shire.cs.psu.edu *!psuvax1!flee
peter@ficc.uu.net (Peter da Silva) (09/11/89)
In article <112@blekko.UUCP>, skrenta@blekko.UUCP (Rich Skrenta) writes: > It's not rn-bigotry. It's rn, notes, or any newsreader that tries to follow > threads based on the subject line. Following threads based on subject line is wrong. Subject lines are user- edited, and change continually. They very rarely have much to do with the subject, outside of moderated groups. A newsreader that depends in them is broken. You've got a problem with buggy software, and that's that. The right way to follow threads is the references line. If you really want to do a good deed get people to fix the software that breaks *that*. -- Peter da Silva, *NIX support guy @ Ferranti International Controls Corporation. Biz: peter@ficc.uu.net, +1 713 274 5180. Fun: peter@sugar.hackercorp.com. `-_-' "BORED OF THE RINGS was written by Sir James Burton, the famous actor. It was published by the Society for the Propagation of Religious Truth." - Tim Maroney
peter@ficc.uu.net (Peter da Silva) (09/11/89)
In article <11527@boulder.Colorado.EDU>, rsk@boulder.Colorado.EDU (Rich Kulawiec) writes: > Well, until someone writes another RFC to replace the existing one, it's all > we've got, flaws or not -- and silently tossing an article into "junk" > because of a noncompliant "Subject" line doesn't seem inherently different > to me than junking it because it has no "From" line, a badly formatted "Date", > or any other problem. But the From and Date lines are computer generated. Let's consider a potential case. Suppose I'm reading rec.arts.sf-lovers, and I am getting tired of the Annual Heinlein Flame Fest. I post an article with the subject: Re^32767: Flaming Heinlein. It's not a reply, it doesn't violate the RFC. It *looks* to a human like it's in violation, but the RFC doesn't mention subject lines beginning with "Re^" at all, so it's OK. A few (very few) remaining NN sites might still be doing this automatically, but the problem is going away without your terrorist tactics. Yes, that's broken... as it would be if it just copied the subject line without putting any "Re" at all. Should you junk any articles with a references line and no Re: in the subject line? Feh. -- Peter da Silva, *NIX support guy @ Ferranti International Controls Corporation. Biz: peter@ficc.uu.net, +1 713 274 5180. Fun: peter@sugar.hackercorp.com. `-_-' "BORED OF THE RINGS was written by Sir James Burton, the famous actor. It was published by the Society for the Propagation of Religious Truth." - Tim Maroney
fritz@unocss.UUCP (Tim Russell) (09/11/89)
tombre@weissenburger.crin.fr (Karl Tombre) writes: |It seems most likely to me that this conf is part of his "speaking |tour", and hence the article posted in his name most probably is a forgery. Upon comparing paths of the "offending" post and Spaf's monthly ones, I too have come to the conclusion that this is yet another NNTP forgery. I'm sorry for acting so hastily, and I should have known Spaf was better than to suggest such a flagrant abuse of net policy. -- ---------------------------------+-------------------------------------------- Tim Russell, Computer Operator | Internet: russell@zeus.unl.edu Campus Computing | Bitnet: russell@unoma1 University of Nebraska at Omaha | UUCP: uunet!zeus.unl.edu!russell
guy@auspex.auspex.com (Guy Harris) (09/12/89)
>Actually, I'd like to see the RFC. I suspect that it deals with what may be >in the header region, but not with the contents of the Subject: field. Wrong. >Would someone please post that RFC to this group, so that this may be >judged fairly? Then we'll decide what to do... Well, here are the passages on "Subject:" and "References:": 2.1.4. Subject The "Subject" line (formerly "Title") tells what the message is about. It should be suggestive enough of the contents of the message to enable a reader to make a decision whether to read the message based on the subject alone. If the message is submitted in response to another message (e.g., is a follow-up) the default subject should begin with the four characters "Re:", and the "References" line is required. For follow-ups, the use of the "Summary" line is encouraged. ... 2.2.5. References This field lists the Message-ID's of any messages prompting the submission of this message. It is required for all follow-up messages, and forbidden when a new subject is raised. Implementations should provide a follow-up command, which allows a user to post a follow-up message. This command should generate a "Subject" line which is the same as the original message, except that if the original subject does not begin with "Re:" or "re:", the four characters "Re:" are inserted before the subject. If there is no "References" line on the original header, the "References" line should contain the Message-ID of the original message (including the angle brackets). If the original message does have a "References" line, the follow-up message should have a "References" line containing the text of the original "References" line, a blank, and the Message-ID of the original message. The purpose of the "References" header is to allow messages to be grouped into conversations by the user interface program. This allows conversations within a newsgroup to be kept together, and potentially users might shut off entire conversations without unsubscribing to a newsgroup. User interfaces need not make use of this header, but all automatically generated follow-ups should generate the "References" line for the benefit of systems that do use it, and manually generated follow-ups (e.g., typed in well after the original message has been printed by the machine) should be encouraged to include them as well. It is permissible to not include the entire previous "References" line if it is too long. An attempt should be made to include a reasonable number of backwards references. It says "should" rather than "shall", but I couldn't find "shall" in the document, so I don't know whether the use of "should" signifies that the rules on "Subject:" lines are advisory or not. Nevertheless, given that there do exist newsreaders that use "Subject:" lines to construct threads - and that there is no hope that all postings will have correct "Reference:" lines, absent changes to *mail* readers on many systems (remember, some newsgroups are gatewayed to and from Internet mailing lists) - it should be considered Bad Form for newsreaders to put out subject lines with "improvements" like "Re^n:". (Whether messages containing such subject lines should be junked, fixed in transit, or passed through, is an issue I have no desire to become embroiled in....)
tale@pawl.rpi.edu (David C Lawrence) (09/12/89)
tombre@weissenburger.crin.fr (Karl Tombre) writes: Karl> It seems most likely to me that this conf is part of his Karl> "speaking tour", and hence the article posted in his name most Karl> probably is a forgery. Well, why don't we just sit back and wait for Spaf to tell us? Right, maybe he won't. Rick Adams never confirmed or denied here the cojecture regarding a possibly forged newgroup of talk.bizarre.nice. But just because Spaf has gone on a speaking tour doesn't mean he couldn't have made the posting ... In <1654@unocss.UUCP> fritz@unocss.UUCP (Tim Russell) writes: Tim> Upon comparing paths of the "offending" post and Spaf's monthly Tim> ones, I too have come to the conclusion that this is yet another Tim> NNTP forgery. I'm sorry for acting so hastily, and I should have Tim> known Spaf was better than to suggest such a flagrant abuse of Tim> net policy. The Path here indicates no immediately apparent foul play. It got here via purdue -> tut -> gem -> rpi. No other machines were involved, and there are well-established links between these. By the way, I've never seen any net policy document which would have labeled that a flagrant abuse. RFCs are about the closest we come anyway. Oh, you mean some sort of generally understood behaviour for USENET participants? Gosh, maybe we should set up some sort of cabal to decide just what that is. Anyone want to start a mailing list for the bigger sites, the ones that sort of form a backbone for this whole operation? It seems to be an idea whose time has come [ ... and gone.] Dave -- (setq mail '("tale@pawl.rpi.edu" "tale@itsgw.rpi.edu" "tale@rpitsmts.bitnet"))
bob@tinman.cis.ohio-state.edu (Bob Sutterfield) (09/12/89)
In article <7194@rpi.edu> tale@pawl.rpi.edu (David C Lawrence) writes:
The Path here indicates no immediately apparent foul play.
The only header lines that I could find that might indicate foul play,
when compared to the regularly-scheduled (mechanical?) postings that
came out around 3 Sep 89 22:27:43 GMT, are the uncharacteristically
present Reply-To: and the slight change to Organization:. Not enough
evidence either way. And when compared to non-regularly-scheduled
postings like <7792@medusa.cs.purdue.edu>, even those lines look fine,
because the look the same as normal (non-mechanical) discussion.
Note that, textual criticism aside, the issue of whether or not a
particular individual posted an article is immaterial to the morality
of pursuing a course the article might advocate. We are free {wo}men!msmith@topaz.rutgers.edu (Mark Robert Smith) (09/12/89)
According to the Path line here, the posting shows a UseNet link between purdue and an ohio-state.edu machine. The map entry for purdue does not advertise this link, and neither does the map entry for tut.cis.ohio-state.edu. Mark -- Mark Smith | "Be careful when looking into the distance, |All Rights 61 Tenafly Road|that you do not miss what is right under your nose."| Reserved Tenafly,NJ 07670-2643|rutgers!topaz.rutgers.edu!msmith,msmith@topaz.rutgers.edu You may redistribute this article only to those who may freely do likewise.
werner@utastro.UUCP (Werner Uhrig) (09/12/89)
aside from the question if Gene posted the original article (he is in Europe
at the moment and will not be back and able to comment on the matter until
the second week of October, probably)...
...I fail to see what is wrong with someone sending notes to sites and
authors who fail to conform with RFC-standards for message-headers!
...and I fail to see what is wrong with using the computer to automate and
support this venture!
...and I fail to see what is wrong in sharing the script or program with
other people!
quite honestly, I imagine that someone thought the Re^2: a cute
header and started using it without being aware of the consequence
and others started copying it (I admit that it is possible that I
have done so) - so we have a problem at hand, in that the software
with smarts to follow discussion threads cannot work anymore, right?
so how to deal with that? post an article to news.*? won't work,
it is not likely that many of the authors of the offending articles
read it. send a polite email message to the authors of offending
articles pointing out the problem they are causing others is the
only available path. doing this manually is adding injury to insult.
responding only to the ones one personally reads is not going to
quell this "cutsy habit" as fast as it is likely to spread.
automating a response to all offending news-articles is the only
way to go.
Quite honestly, I found some of the follow-ups to the original
article quite offensive; rather than sending a polite questioning
query to the author of the message that causes you grief, some people
have a habit of "shooting first and asking later". It is truely
depressing, especially when it happens to people with an established
track-record of years of "public service" to the net, as is the case
here.
--
-----------> PREFERED RETURN-ADDRESS FOLLOWS <--------------
(ARPA) werner@rascal.ics.utexas.edu (Internet: 128.83.144.1)
(UUCP) ..!utastro!werner or ..!uunet!rascal.ics.utexas.edu!wernerkarl@godiva.cis.ohio-state.edu (Karl Kleinpaste) (09/12/89)
msmith@topaz.rutgers.edu writes:
According to the Path line here, the posting shows a UseNet link
between purdue and an ohio-state.edu machine. The map entry for
purdue does not advertise this link, and neither does the map entry
for tut.cis.ohio-state.edu.
Well....true enough. I confess to a certain laziness with regard to
how often and how carefully I update the #U line in our map entry,
especially when compared against the relative fastidiousness with
which more ordinary (mail) connections are recorded. I suspect
similar thoughts apply to Purdue. I tend not to think much of #U
lines simply because they are not useful and change much too often.
--Karlpda@stiatl.UUCP (Paul D. Anderson) (09/12/89)
Has anyone gotten mail messages from Spaf generated by this (perhaps forged) mailing program? I have an associate that claimed last week to get some messages from him (before the 'posting'). How possible is it to have mail messages forged (as opposed to news messages)? Re: the contents of the Subject: colon field, the usenet RFC does not specify the *contents* of that field. It does say that whenever there are unresolved issues, you should refer to RFC 822, the internet mail standard. So everyone, off to the RFC archives for further info... paul -- Paul Anderson (w) (404) 841-4000 gatech!stiatl!pda (h) (404) 662-0799
msmith@topaz.rutgers.edu (Mark Robert Smith) (09/12/89)
Karl, My apologies. I do realize that the #U lines are often out of date. After this information, I can only believe that the article is either legit or from purdue. Mark -- Mark Smith | "Be careful when looking into the distance, |All Rights 61 Tenafly Road|that you do not miss what is right under your nose."| Reserved Tenafly,NJ 07670-2643|rutgers!topaz.rutgers.edu!msmith,msmith@topaz.rutgers.edu You may redistribute this article only to those who may freely do likewise.
tneff@bfmny0.UU.NET (Tom Neff) (09/12/89)
In article <6120@ficc.uu.net> peter@ficc.uu.net (Peter da Silva) writes: >Following threads based on subject line is wrong. Subject lines are user- >edited, and change continually. But this is because new threads and subthreads are continually born. Existing practice is to change the Subject field when the topic has truly drifted; so following by subject tends to work. Existing practice is also to include the "(was: " link for a while. > They very rarely have much to do with the >subject, outside of moderated groups. So that this thread, for instance, is a rarity? Hmm. It seems typical. >The right way to follow threads is the references line. If you really want >to do a good deed get people to fix the software that breaks *that*. But in general there is a limit on line length, and References quickly mount up. Moreover when one does wish to start a new thread with links to a previous discussion, it is frequently unclear how many old References to delete. Ref's just don't work well enough to be used universally. And they aren't the whole answer by any means even when they are working. Given one main discussion and several "limb" discussions hived off and active, how do you traverse? If you ignore the Subject field and pay exclusive attention to References, you may wander horizontally through several only distantly related discussions. At any rate, whether following threads by Subject is a good or bad thing, adding Re^n: contributes *nada*. It tells a newsreader program nothing about how to traverse, and a newsreading person nothing useful about the message being read. (Re^7? So what??) And unless everyone adopts it, even the count is unreliable. -- Annex Canada now! We need the room, \) Tom Neff and who's going to stop us. (\ tneff@bfmny0.UU.NET
mdb@ESD.3Com.COM (Mark D. Baushke) (09/12/89)
In article <6121@ficc.uu.net> peter@ficc.uu.net (Peter da Silva) writes: peter> In article <11527@boulder.Colorado.EDU>, rsk@boulder.Colorado.EDU (Rich Kulawiec) writes: rsk> Well, until someone writes another RFC to replace the existing rsk> one, it's all we've got, flaws or not -- and silently tossing an rsk> article into "junk" because of a noncompliant "Subject" line rsk> doesn't seem inherently different to me than junking it because rsk> it has no "From" line, a badly formatted "Date", or any other rsk> problem. peter> But the From and Date lines are computer generated. peter> Let's consider a potential case. Suppose I'm reading peter> rec.arts.sf-lovers, and I am getting tired of the Annual peter> Heinlein Flame Fest. I post an article with the subject: peter> Re^32767: Flaming Heinlein. peter> It's not a reply, it doesn't violate the RFC. It *looks* to a peter> human like it's in violation, but the RFC doesn't mention peter> subject lines beginning with "Re^" at all, so it's OK. peter> A few (very few) remaining NN sites might still be doing this peter> automatically, but the problem is going away without your peter> terrorist tactics. Yes, that's broken... as it would be if it peter> just copied the subject line without putting any "Re" at all. peter> Should you junk any articles with a references line and no Re: peter> in the subject line? Feh. If I were going to junk articles (which I do not advocate), those are exactly the articles which I would agree 'could' be junked. (Those articles which have a References: line, but do NOT have a Subject: starting with the four characters "Re: ".) The posting software should not allow an article to be posted which voilates the 'standard'. Someone trying to post a 'new' article with a References: line should have the newsposting software complain. If you wanted to use the Re^n: then you would have to modify the Subject: to have "Re: Re^n: " as the prefix. (As has been said, modification of the Subject string is allowed so putting in the extra Re^n would be 'legal'). However, I doubt that 'Re: Re^n: ' prefixes would find much favor. :-) -- Mark D. Baushke Internet: mdb@ESD.3Com.COM UUCP: {3comvax,auspex,sun}!bridge2!mdb
rsk@boulder.Colorado.EDU (Rich Kulawiec) (09/12/89)
In article <6121@ficc.uu.net> peter@ficc.uu.net (Peter da Silva) writes: >A few (very few) remaining NN sites might still be doing this automatically, >but the problem is going away without your terrorist tactics. This would probably be a good time to point out that at no time did I state that I had actually installed the code to do this. If you look back and re-read my postings on the subject, you'll note that I stated that modifying news 2.11 B to silently drop non-compliant articles was my solution to the problem; I did not say that I had implemented the solution on a site somewhere. Perhaps "terrorist tactics" is a bit premature and (in my opinion) a bit of a loaded phrase. ---Rsk
jim@eda.com (Jim Budler) (09/12/89)
guy@auspex.auspex.com (Guy Harris) writes: >>Would someone please post that RFC to this group, so that this may be >>judged fairly? Then we'll decide what to do... >Well, here are the passages on "Subject:" and "References:": >2.1.4. Subject > The "Subject" line (formerly "Title") tells what the message is > about. It should be suggestive enough of the contents of the > message to enable a reader to make a decision whether to read the > message based on the subject alone. If the message is submitted in > response to another message (e.g., is a follow-up) the default > subject should begin with the four characters "Re:", and the > "References" line is required. For follow-ups, the use of the > "Summary" line is encouraged. >... >2.2.5. References > This field lists the Message-ID's of any messages prompting the > submission of this message. It is required for all follow-up > messages, and forbidden when a new subject is raised. > Implementations should provide a follow-up command, which allows a > user to post a follow-up message. This command should generate a > "Subject" line which is the same as the original message, except > that if the original subject does not begin with "Re:" or "re:", the > four characters "Re:" are inserted before the subject. If there is > no "References" line on the original header, the "References" line Seems pretty clear to me. 1. The early version of 'nn' *was* incorrect in using Re^n: 2. The author corrected this *several* patches ago. 3. It's time to stop blaming the author of 'nn' who has done a very good job of responding to bug reports, including the report long ago concerning this particular implementation error. 4. It's time to start blaming the software maintainers who are way behind in the updates. To continue to blame 'nn' itself for this problem now, is like blaming 'rn' for the frequent question "Why do I keep seeing the same cross-posted article I've already seen?" when the problem really belongs to the failure of the maintainer compiling in Xref to news 2.11 long after it was available. jim P.S. This followup is via 'nn version 6.3.6'. If it says Re^n: you may very vocally blame me as I am the news maintainer here. -- Jim Budler address = uucp: ...!{decwrl,uunet}!eda!jim domain: jim@eda.com compuserve: 72415,1200 voice = +1 408 986-9585 fax = +1 408 748-1032
tneff@bfmny0.UU.NET (Tom Neff) (09/12/89)
One problem with automated complaint scripts is that, to the extent they
get around into general usage, you start levying email in flood volumes
comparable to that displaced by a news article itself. We all agree news
is more expensive than mail because it floods. The potential for automated
complaint mail overloading leaf sites is overwhelming.
We do not need to invent new species of rudeness, enough exist already.
Better to filter the venial sins silently and let software attrition run
its course since Re^n is known to have been fixed.
--
Annex Canada now! We need the room, \) Tom Neff
and who's going to stop us. (\ tneff@bfmny0.UU.NETrsalz@bbn.com (Rich Salz) (09/12/89)
In <6120@ficc.uu.net> peter@ficc.uu.net (Peter da Silva) writes: >Following threads based on subject line is wrong. ... >The right way to follow threads is the references line. We can't get people to change the subject line when the thread changes, and you wanna have them tweak the References line? Ha! /r$ -- Please send comp.sources.unix-related mail to rsalz@uunet.uu.net. Use a domain-based address or give alternate paths, or you may lose out.
brad@looking.on.ca (Brad Templeton) (09/12/89)
Oh, there is no question in my mind that the RFC is wrong on this point. Tracking threads based on the subject that the presence of "re:" is a stupid way of doing it, and RN's behaviour in this should never have been coded down into the proposed standard. But once it was coded in, the correct action for the NN authors would have been to discuss the change they wanted in news.software.b or some other appropriate place. Chances are they would have decided against the change if they had done this, but that's not important. Even if they then decided to go ahead, that would be fine, even if it broke RN, because at least they would have tried to work it out in advance. We can't restrict our software to the behaviour of old readers, but we should check the usefulness and consequences of innovation in advance. -- Brad Templeton, Looking Glass Software Ltd. -- Waterloo, Ontario 519/884-7473
ray@philmtl.philips.ca (Raymond Dunn) (09/13/89)
In article <6120@ficc.uu.net> peter@ficc.uu.net (Peter da Silva) writes: >In article <112@blekko.UUCP>, skrenta@blekko.UUCP (Rich Skrenta) writes: >> It's not rn-bigotry. It's rn, notes, or any newsreader that tries to follow >> threads based on the subject line. > >Following threads based on subject line is wrong. Whether or not it's "wrong", it's certainly no justification for arbitrarily making stylistic ad-hoc "improvements" to an agreed standard that causes much of the installed base to beak. Anyway, hasn't all that been covered before? Isn't the fix action already under way? > Subject lines are user- >edited, and change continually. They very rarely have much to do with the >subject, outside of moderated groups. A newsreader that depends in them >is broken. Look, the main problem with the Subject line is *not* that posters change it unnecessarily, but that they *dont* change it when they go off at a tangent. To say that they *rarely* have much to with the subject is just plain silly. In fact most of the time, even though there can be several sub-threads intertwining, the subject line is very relevant. Following the references line rather than the subject line is no panacea. It is not adequate if threads intertwine and wander back onto their original course as they often do, and even then it depends on posters not modifying the references line, an action specifically *allowed* in the spec. I've been mentally playing with an algorithm that would follow references, and frankly, it seems to always require much ad-hoc branching on little more than guesses. It is not clear whether depth first or breadth first scanning of the references tree is generally appropriate. In particular, branches will re-join without any immediately connecting references, as posters reply to several responses at the previous "level" together. Thus the thread must be followed not by the last reference, but by some earlier reference in the chain. Following the Subject line is just an extreme example of that, and is equivalent to following the original article's reference. When looking at the whole tree in hindsight it is fairly clear which paths should have been followed, but only because of their *content*, not their references, nor their subject, nor their originator!! Has anyone done any definitive work on this? Have any thread following algorithms already been suggested and tried out? -- Ray Dunn. | UUCP: ..!uunet!philmtl!ray Philips Electronics Ltd. | TEL : (514) 744-8200 Ext: 2347 600 Dr Frederik Philips Blvd | FAX : (514) 744-6455 St Laurent. Quebec. H4M 2S9 | TLX : 05-824090
ray@philmtl.philips.ca (Raymond Dunn) (09/13/89)
In article <6852@stiatl.UUCP> pda@stiatl.UUCP (Paul D. Anderson) writes: >Re: the contents of the Subject: colon field, the usenet RFC does not >specify the *contents* of that field. It does however specify the use and format of "Re: ". Clearly, "Re^n: " does not conform. Shouldn't we be directing followups out of news.admin? -- Ray Dunn. | UUCP: ..!uunet!philmtl!ray Philips Electronics Ltd. | TEL : (514) 744-8200 Ext: 2347 600 Dr Frederik Philips Blvd | FAX : (514) 744-6455 St Laurent. Quebec. H4M 2S9 | TLX : 05-824090
bnews@nixpbe.UUCP (Martin Boening) (09/13/89)
pda@stiatl.UUCP (Paul D. Anderson) writes: >rsk@boulder.Colorado.EDU (Rich Kulawiec) writes: >Actually, I'd like to see the RFC. I suspect that it deals with what may be >in the header region, but not with the contents of the Subject: field. I won't go off and post the whole RFC - but here is what it says about Subject lines ---begin quote---- ..... 2.1.4. Subject The "Subject" line (formerly "Title") tells what the message is about. It should be suggestive enough of the contents of the message to enable a reader to make a decision whether to read the message based on the subject line alone. If the message is submitted in response to another message (e.g., is a follow-up) the default subject should begin with the four characters "Re: " and the "References" line is required. For follow-ups, the use of the "Summary" line is encouraged. 2.1.5. Message-ID ..... ---end quote---- Interpretation by me: if a newsreader sets up a Subject line, it should always use "Re: " for followups. However, this is the default subject and users are free to insert anything they want here, even if this causes newsreaders following threads of discussion via Subject to loose out. Since the References line is mandatory for follow-ups, use that to follow threads of discussion. Just my opinion, of course. So have fun flaming around. Martin -- Martin Boening, c/o Nixdorf Computer AG, DS-CC2, Paderborn, West-Germany Email: | Phone: (+49) 5251 146155 USA: uunet!linus!nixbur!mboening.pad | Fax : (+49) 5251 146108 !USA: mcvax!unido!nixpbe!mboening.pad |
bnews@nixpbe.UUCP (Martin Boening) (09/13/89)
rsk@boulder.Colorado.EDU (Rich Kulawiec) writes: >Actually the followup "Subject" lines *are* created by software, which is >how we got into this in the first place. And, yes, I think junking articles >with more than 4 lines of signature is an excellent idea. >---Rsk Actually the DEFAULT followup "Subject" lines are created by software which is how we get DEFAULT followup "Subject" lines. Note the DEFAULT! According to the RFC, that's all it is. It is NOT mandatory(!) (as in mandatory :-)) subject line. It may be edited by users in any way they choose. The only thing to become mandatory for a follow-up article is the "References" line. THEREFORE it does not matter what news software DOES set up as Subject. (To use the four letters "Re: " to start the default subject on followups is a suggestion, as in: ... If a message is submitted in response to another message (e.g., is a follow-up) the default subject SHOULD begin with the four characters "Re: " and the "References" line is required... ) Note: The "Re: " is suggested (SHOULD), the "References" line is REQUIRED. So let's just say that a "Subject:" line has NO format and therefore articles MUST NOT be discarded because of them. This should be enough to convince inews to send this thing :-) My opinions, of course (why else should I write them down). Martin -- Martin Boening, c/o Nixdorf Computer AG, DS-CC2, Paderborn, West-Germany Email: | Phone: (+49) 5251 146155 USA: uunet!linus!nixbur!mboening.pad | Fax : (+49) 5251 146108 !USA: mcvax!unido!nixpbe!mboening.pad |
brad@looking.on.ca (Brad Templeton) (09/13/89)
In fact, my argument has always been that the ability to have an automatic subject line with "Re:" is a bug in the news posting programs. It discourages people writing descriptive subjects. What I want is a subject line that tells what is in the article it heads, not a subject line that tells what is in the article it's replying to, N times removed. Sadly, it's getting very hard to change this. Mind you, I think people would be more willing to go for radical change like this than for NN's change. NN's change was pointless, that's why it annoyed people. -- Brad Templeton, Looking Glass Software Ltd. -- Waterloo, Ontario 519/884-7473
henry@utzoo.uucp (Henry Spencer) (09/13/89)
In article <12715@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: >In fact, my argument has always been that the ability to have an automatic >subject line with "Re:" is a bug in the news posting programs. It >discourages people writing descriptive subjects... It also discourages subjects like, say, "x", or "followup". Brad, I'm afraid you seem to have this fixed notion that people will consistently put forth the effort to compose sensible and descriptive subjects/keywords/etc. if they are told to. I doubt that very much. -- V7 /bin/mail source: 554 lines.| Henry Spencer at U of Toronto Zoology 1989 X.400 specs: 2200+ pages. | uunet!attcan!utzoo!henry henry@zoo.toronto.edu
werner@utastro.UUCP (Werner Uhrig) (09/14/89)
In <12715@looking.on.ca>, brad@looking.on.ca (Brad Templeton) writes: > In fact, my argument has always been that the ability to have an automatic > subject line with "Re:" is a bug in the news posting programs. It > discourages people writing descriptive subjects. I would rather trust a machine-generated Subject-line containing "Re:" to indicate a thread than a user's hand-knit Subject line which is more likely to obscure that a thread exists. the attentive author adds a "Summary"" line, or, if he wants to make clear that it is only "formally" a follow-up, he can append at the end of Subject-line a "really: I'm taking the discussion elsewhere" > What I want is a subject line that tells what is in the article it heads, > not a subject line that tells what is in the article it's replying to, > N times removed. what we all want is better computer support to help in identifying and extracting information that interests us while ignoring the rest. By custom, the Subject-line has a "formal definition" of contents, and I find the Summary-line, given the proper contents by the author, satisfies my needs at the moment. I don't consider the situation as bleak as some people paint it, on the other hand, I also don't mean to say that things could not benefit from some changes. Tools exist to benefit from "Re:" so lets use them until some truely better tools are available. I look forward to the clipping-service as a truely remarkable step forward. kudos for that initiative, Brad ... how is it coming? I have not heard any news for quite a while (or am I not looking into the right "nooks and crannies"?!) -- -----------> PREFERED RETURN-ADDRESS FOLLOWS <-------------- (ARPA) werner@rascal.ics.utexas.edu (Internet: 128.83.144.1) (UUCP) ..!utastro!werner or ..!uunet!rascal.ics.utexas.edu!werner
kim@kannel.lut.fi (Kimmo Suominen) (09/14/89)
In article <12715@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes:
From: brad@looking.on.ca (Brad Templeton)
Date: 13 Sep 89 07:41:31 GMT
Organization: Looking Glass Software Ltd.
In fact, my argument has always been that the ability to have an automatic
subject line with "Re:" is a bug in the news posting programs. It
discourages people writing descriptive subjects. What I want is a subject
line that tells what is in the article it heads, not a subject line that
tells what is in the article it's replying to, N times removed.
I find it very handy to kill articles on some subjects and as far as those Re:
subjects are true, killing works fine. I'm sure I'll loose something that
way, but that's not my problem.
So if anyone has something important to say, (s)he'd better change the sub-
ject. Personally I'm for the "New stuff (was: Old stuff)" style of subject
lines.
I've been using NN and I'm now using GNUS and both allow me to edit the sub-
ject line before posting (they do offer me a default as well).
Kim
--
,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,
( Kimmo Suominen Electronic Mail on Internet: kim@kannel.lut.fi )
( "That's what I think!" on Funet: KUULA::KIM )
''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''brad@looking.on.ca (Brad Templeton) (09/15/89)
In article <1989Sep13.162419.22921@utzoo.uucp> henry@utzoo.uucp (Henry Spencer) writes: >It also discourages subjects like, say, "x", or "followup". Brad, I'm afraid >you seem to have this fixed notion that people will consistently put forth >the effort to compose sensible and descriptive subjects/keywords/etc. if >they are told to. I doubt that very much. Perhaps. But the idea of Re: oldsubject adds *no* information to the article, assuming the References: line is updated properly. (And for the point of this debate, we are) Every article, in fact, has two subjects, one the general topic of the thread, and one the subject matter of the individual article. I would like to know both. Having a good Subject line for the individual article saves an immense amount of time. Anything to encourage it is worthwhile. If I go away for a while, there might be 100 articles in clari.news.gov.international, but since they all have (as an extreme) a professional editor's headline, I can pick the ones I want in a few seconds. With a typical USENET group, I have no choice to but wipe out whole long discussions with the 'k' key, with no idea of what's really going on. I don't expect a professional's headline on every article, but it would be nice to have more than 0 information. People might put on 1 word subjects, but that will be their problem, since if it happens enough I will just code to discard those articles in advance. I don't know about you, but I don't have the time in most groups to read the *body* of the article to find out if it's worth reading. -- Brad Templeton, Looking Glass Software Ltd. -- Waterloo, Ontario 519/884-7473
brad@looking.on.ca (Brad Templeton) (09/15/89)
In article <4352@utastro.UUCP> werner@utastro.UUCP (Werner Uhrig) writes: > I look forward to the clipping-service as a truely remarkable step > forward. kudos for that initiative, Brad ... how is it coming? > I have not heard any news for quite a while (or am I not looking > into the right "nooks and crannies"?!) You have mistaken me for Geoff Goodfellow, who is running the pro-moderatd newsgroups. My venture sells electronic newswire material and so on. It, after many long delays, is finally up and running. -- Brad Templeton, Looking Glass Software Ltd. -- Waterloo, Ontario 519/884-7473
brad@looking.on.ca (Brad Templeton) (09/15/89)
Oh yes, too much software depends on Re: to change it now. I am just saying I felt it was a mistake to put a big piece of non-info in the header like that. The software should have relied on References: and References should have been made to work. But don't expect that on USENET, where there hasn't been a usable advance in the capability of the file format in around 5 years. To do things today, we would have to make the 'Summary' line the *real* Subject of the individual article, and leave the Subject: line as a piece of non-information. -- Brad Templeton, Looking Glass Software Ltd. -- Waterloo, Ontario 519/884-7473
henry@utzoo.uucp (Henry Spencer) (09/15/89)
In article <13772@looking.on.ca> brad@looking.on.ca (Brad Templeton) writes: >... But the idea of Re: oldsubject adds *no* information to the >article, assuming the References: line is updated properly. (And for the >point of this debate, we are) However, it does not *subtract* information, which is also important. All the more so since a References: line entry for an article which has expired is not very useful. -- V7 /bin/mail source: 554 lines.| Henry Spencer at U of Toronto Zoology 1989 X.400 specs: 2200+ pages. | uunet!attcan!utzoo!henry henry@zoo.toronto.edu
guy@auspex.auspex.com (Guy Harris) (09/16/89)
>The right way to follow threads is the references line. If you really want >to do a good deed get people to fix the software that breaks *that*. You mean like the software that gateways some newsgroups bi-directionally to and from Internet mailing lists, and the mail-reading software on the systems on which people read those mailing lists? Make 'em all generate References: lines, or else? Would everybody who advocates not using the subject lines to link threads please either: 1) come up with a way to make your favorite thread-linking mechanism work on gatewayed groups such as "comp.unix.wizards"; 2) come up with a way to convince people that gatewaying groups like that is a Bad Thing because of the lack of "References:" lines or some other thread-linking mechanism, and therefore should be ceased (when I say "people", I mean people in a position to do something about it, and people who are currently reading those groups as mailing lists - please don't waste your time or mine trying to convince *me*, as I'm not in any position to do anything about it, and since I don't read them as mailing lists I don't think I have the right to say to those who do "you don't deserve to be able to read them in this fashion"); or 3) cease telling us all what a Bad Thing it is to use the Subject: line to link threads, because in some cases of messages arriving from Internet mailing lists it's all you have. Hindsight is 20/20; if you want to draw the lesson from all this that "netnews should have been set up so that References: lines *always* worked, even if that meant not gatewaying from mailing lists", feel free (but be prepared to get toasted by those who find such gatewaying useful), but don't draw the lesson that there's anything you can do about it on USENET as it currently exists.
brad@looking.on.ca (Brad Templeton) (09/16/89)
It's actually not that hard to fix. Just a SMOP. (Small matter of programming) If incoming data from a mailing list has Subject lines linking messages, map those into proper references. The gateway, or sites that get from the gateway, or even a large collection of major sites, simply have to keep a database of current subject lines. When a message comes along (ie. from a mailing list) that doesn't have a References line, you look up the subject in the database. You find the reference and generate a references line. It's not a *good* references line, unfortunately. It will only reference the root of that tree, but it's better than nothing. The reverse would also be possible in a usenet that didn't even bother moving redundant 're:' subjects. Look at the root and extract the subject to feed back to the mailing list. A real tree based system is *great* folks, and if we had one, people would use it, and if outsiders kept injecting messages without proper threading, then more and more people would just start doing automatic reject of such messages. I know I would except in the groups for which I have the keenest interest. I just don't have the time to read messages that people don't classify well. In many groups, I think I'm like many. I type '=', see 30 messages, and if I don't see a subject that grabs me, I type 'c' to skip 'em all. Or I read the group in "new messages only mode." In that mode, rn only shows me non-followups, and I have to deliberately un-kill to see the followups. Then I only see non-followups and followups of messages I tagged as interesting. The result? There may be 7,000 people reading some part of news.admin, but I bet most messages only get fewer than 1,000 actual readers. People post on usenet to get an audience. If the audience goes away, people will have to learn. messages only get a very small number o -- Brad Templeton, Looking Glass Software Ltd. -- Waterloo, Ontario 519/884-7473
pokey@well.UUCP (Jef Poskanzer) (09/16/89)
In the referenced message, guy@auspex.auspex.com (Guy Harris) wrote: } 1) come up with a way to make your favorite thread-linking } mechanism work on gatewayed groups such as } "comp.unix.wizards"; My comp.unix.wizards directory currently has: 463 articles 158 articles without References: 15 articles without References: and with Re: in the subject I suggest that one reasonable way to have reference-following handle gatewayed groups is to ignore them. Let it break on 3% of the articles. Who cares, it would still work loads better than subject-following. Or, you could fall back on subject-following when the references line is missing. However, this is only a suggestion. How you read news is your business, not mine, and I'm happy with how I read news. --- Jef Jef Poskanzer pokey@well.sf.ca.us {ucbvax, apple, hplabs}!well!pokey "It has been said that physicists stand on one another's shoulders. If this is the case, then programmers stand on one another's toes, and software engineers dig each other's graves."
guy@auspex.auspex.com (Guy Harris) (09/17/89)
>If incoming data from a mailing list has Subject lines linking messages, >map those into proper references. > >The gateway, or sites that get from the gateway, or even a large collection >of major sites, simply have to keep a database of current subject lines. >When a message comes along (ie. from a mailing list) that doesn't have >a References line, you look up the subject in the database. You find >the reference and generate a references line. Gee, I guess you must not agree with those folks who say that following threads by comparing subject lines is wrong, then, since you just suggested, in effect, that subject lines be used as one way of tracing threads.... :-) In other words, while this would make it more likely that References: lines are present - and I agree that this would be a Good Thing, assuming a low enough rate of false hits - it doesn't obviate the need for some constraints on the contents of subject lines. It merely means that some software other than newsreaders is doing the subject-line-based tracing; given that, at present, there isn't any software that I know of that does so *other* than newsreaders, it is a Good Thing that the newsreaders do it, so that there's at least *some* way to keep from seeing threads in which you're not interested - it's not perfect, but it's sure better than nothing, and absent some way of ensuring that all articles, or at least a large enough majority of them, have valid References: lines, relying on References: lines may be close enough to nothing for all practical purposes.
guy@auspex.auspex.com (Guy Harris) (09/17/89)
>I suggest that one reasonable way to have reference-following handle >gatewayed groups is to ignore them. Let it break on 3% of the articles. >Who cares, it would still work loads better than subject-following. >Or, you could fall back on subject-following when the references line is >missing. I would vote for the latter; I would want articles from a mailing list folded into a thread, even in the absence of a References: line (I'd also want articles folded into a thread even if the References: line isn't properly maintained - the quality of the posting software used doesn't necessarily correlate with the quality of the posting :-)). In other words, I'd really like a newsreader that used Reference: lines if present, but that can cope with them if they're absent - or if they don't list every single article prior to them on the thread. For the purposes to which I put subject-following with my present newsreader - namely, killing threads in which I have no personal interest - subject-following seems to work pretty well indeed. I don't think Reference-following alone would be enough to make a significant improvement; what I want is a newsreader that can list the *threads* in a newsgroup, and let me browse through them, and for that the References: line would be useful - as a way of sorting articles so that followups are shown after the article to which they're replying, if nothing else. With a newsreader like that, the Subject: line *does* contain information - it identifies the thread. (No, you can't just say "get the subject from the original article in the thread"; remember "Orphaned Response"? Followups can arrive before the original article, assuming the original article even arrives in the first place....) A Summary: line would also be useful, to identify the contribution of that particular article to the thread - assuming people take the trouble to prepare a good one, which they may or may not do; at this point, I'm unwilling to take for granted predictions about what people will do until I see whether they actually do it or not. The References: line basically doesn't strike me as being that much more useful than the Subject: line, in the present environment, for grouping articles into a thread. Its main use seems to be in 1) sorting articles based on a (partial) order, so followups are shown after the article to which they're responding if possible, and 2) following a reference back to the article to which the current article is following up. Unfortunately, my current newsreader ("rn") doesn't do either one, as far as I can tell, and a replacement newsreader that *does* do both of them would have to do at least as good a job as "rn" of discarding uninteresting threads to be an acceptable replacement for it. (It'd probably also have to let me look at the *threads* in a newsgroup, as indicated above - with the option of showing me only threads with articles I've not seen yet, or threads with articles I've either not seen or marked as "unread" - and then let me open up each thread to see it, in order to be enough better to provoke me into making the effort to learn it.)
brad@looking.on.ca (Brad Templeton) (09/17/89)
In article <13666@well.UUCP> Jef Poskanzer <pokey@well.sf.ca.us> writes: >I suggest that one reasonable way to have reference-following handle >gatewayed groups is to ignore them. Let it break on 3% of the articles. >Who cares, it would still work loads better than subject-following. >Or, you could fall back on subject-following when the references line is >missing. No! No no no no no no no! While first of all my own survey showed 6% of articles on the net in the month I surveyed had "Re:" and no references line, the problem is far worse. Each article that breaks the reference chain breaks the chain for all the articles that follow it up. So 6% break points can translate into 30% or more messages not linked to their root. I wrote software to let me kill a tree based on the References line. When I see a message I don't want to follow, and I note the message-id in my database, I find that the followups still keep showing up, again and again. Yes, you can keep killing and killing but it ruins the whole point of it. -- Brad Templeton, Looking Glass Software Ltd. -- Waterloo, Ontario 519/884-7473
brad@looking.on.ca (Brad Templeton) (09/17/89)
The nice thing about the References line is that it represents a *tree*, not a stream. The "Re:" subject line only connects all messages with the root of that tree, unless the subject is edited. It would be very useful to use the full tree. You can get rid of, or more to the point, deliberately follow, interesting subtrees without the other subtrees getting in your way. For example, this discussion, "Tired of bogus subject lines" has discussions about NN, References, possible forged usenet messages, mailing lists and more. It would be good to be able to select. Yes, a clever reader can make guesses when it sees an article that has the references line damaged, but it can't ever do a perfect job. You will either be forced to see damaged messages to see if they match the subtrees you are following, or ignore them all, and miss things in the trees of keenest interest. Of course, even smart mailing list interfaces with lots of databases of subjects can't fix this. But the groups with ML gateways are known, and this can be dealt with. -- Brad Templeton, Looking Glass Software Ltd. -- Waterloo, Ontario 519/884-7473
woerz%isaak@isaak.uucp (Dieter Woerz) (09/18/89)
In article <13666@well.UUCP> Jef Poskanzer <pokey@well.sf.ca.us> writes: > ... >I suggest that one reasonable way to have reference-following handle >gatewayed groups is to ignore them. Let it break on 3% of the articles. You could even reduce the number of articles without reference-line in gatewayed groups, when the gatewaying software tries to use the In-Reply-To header line, which I think most UAs generate, in mailed postings. You should have mailed the posting with the original Message-Id line (the one used by news) and then you get the message- id of the article to which the mailer has replied. If it is needed you could even try to look up the referenced article and include the references from this one into the article to be added. >Who cares, it would still work loads better than subject-following. >Or, you could fall back on subject-following when the references line is >missing. > ... Dieter Woerz ISA GmbH, Azenbergstr. 35 D-7000 Stuttgart-1 W-Germany UUCP: {pyramid!iaoobel,uunet!unido}!isaak!woerz BITNET/EARN: woerz@ds0iff5
chip@ateng.com (Chip Salzenberg) (09/19/89)
According to brad@looking.on.ca (Brad Templeton): >It's actually not that hard to fix. Just a SMOP. (Small matter of >programming) Quite: 1. Make the news->mail gateway keep the news article Message-ID. 2. Make the mail->news gateway use the In-reply-to:, if any, to generate the References: line. >If incoming data from a mailing list has Subject lines linking messages, >map those into proper references. This is to be used if and only if In-reply-to: is missing and the Subject begins with "Re:". -- You may redistribute this article only to those who may freely do likewise. Chip Salzenberg at A T Engineering; <chip@ateng.com> or <uunet!ateng!chip> "If you push something hard enough, it will fall over." -- Fudd's First Law of Opposition
kjones@talos.uucp (Kyle Jones) (09/19/89)
Jef Poskanzer writes: > I suggest that one reasonable way to have reference-following handle > gatewayed groups is to ignore them. Let it break on 3% of the articles. Dieter Woerz writes: > You could even reduce the number of articles without reference-line > in gatewayed groups, when the gatewaying software tries to use the > In-Reply-To header line, which I think most UAs generate, in mailed > postings. In-Reply-To is not a reliable reference source because 1. It is an optional header. 2. It is not required to contain the message-ID of the message being replied to. I've have received many pieces of mail that have an In-Reply-To header that does not contain a message-ID. The real solution is to for mail user agents to support the more rigid RFC 1036 definition of References. (References has the same status as In-Reply-To in RFC 822, and this is not enough.) So if any authors of MUAs are listening, please make your mailer support References in a future release. Every little bit helps. kyle jones <kjones@talos.uu.net> ...!uunet!talos!kjones "Aye, and if my grandmother had wheels, she'd be a wagon!"
guy@auspex.auspex.com (Guy Harris) (09/20/89)
>You could even reduce the number of articles without reference-line >in gatewayed groups, when the gatewaying software tries to use the >In-Reply-To header line, which I think most UAs generate, in mailed >postings. It might be nice to use headers like that, but: 1) Berkmail (Mail/mailx), for instance, doesn't generate them - I don't know whether most users run a UA that generates them, or not; 2) they don't always contain the Message-ID of the message to which the user is replying.
karl@ficc.uu.net (Karl Lehenbauer) (09/21/89)
In article <15787@looking.on.ca>, brad@looking.on.ca (Brad Templeton) writes: > It would be very useful to use the full tree. You can get rid of, or more > to the point, deliberately follow, interesting subtrees ... I have used tree-based messaging systems. One was, I believe, Communitree, a forth-based system. The other was an experimental package on PLATO. I felt they were confusing and easy to get lost in. I thought that PLATO's solid, "standard" base-message-and-list-of-responses was less confusing. -- -- uunet!ficc!karl "The last thing one knows in constructing a work is what to put first." -- Pascal
peter@ficc.uu.net (Peter da Silva) (09/25/89)
I said: >The right way to follow threads is the references line. If you really want >to do a good deed get people to fix the software that breaks *that*. In article <2454@auspex.auspex.com>, guy@auspex.auspex.com (Guy Harris) writes: > Would everybody who advocates not using the subject lines to link > threads please either: > 1) come up with a way to make your favorite thread-linking > mechanism work on gatewayed groups such as > "comp.unix.wizards"; Put the references lines (and other magic lines) *after* the header, the way Fidonet apparently does it, and the way the sources groups do it. i.e.: + From: no.such.user@cup.portal.com 25 Dec 1969 00:12:45 + Subject: Re: Re^n: Heinlein was a Bozo + Message-ID: <2^n-1@cup.portal.com> + + Newsgroups: rec.arts.sf-lovers,comp.sys.ibm.pc.jerry.pournelle + References: <24.dec.1969.23.20.12.verbose.10096425176284@verbose.BITNET> + + Heinlein was god, you scum-sucking fascist. -- Peter da Silva, *NIX support guy @ Ferranti International Controls Corporation. Biz: peter@ficc.uu.net, +1 713 274 5180. Fun: peter@sugar.hackercorp.com. `-_-' "That is not the Usenet tradition, but it's a solidly-entrenched U delusion now." -- brian@ucsd.Edu (Brian Kantor)
tower@bu-cs.BU.EDU (Leonard H. Tower Jr.) (10/11/89)
[ Excuse the lateness of this, I'm just catching up on
news.admin after a very heavy month. ]
nn could pass these header lines to inews:
Subject: Re: ORIGINAL TEXT
X-NN-Re-Count: 2
or whatever the count is. The X- is recommended to be used by
RFC-822, which promises that no version of RFC-822 will ever have a
header line starting with X-. The whol header line title is
just an example. It could be shortened or made what ever the nn
people like.
When nn sees an X-NN-Re-Count: header line. It would NOT display it,
but would display the Subject line as:
Subject: Re^2: ORIGINAL TEXT
or whatever the count is. Note that this is NOT the Subject line in
the news spool file.
Yes, it would cost a few more bytes per article posted by nn.
Yes, some news readers would require people to add X-NN-Re-Count: to
a list of header lines to ignore.
But, it would allow this kind of experiment to be tried and let the nn
users have the kind of information they want displayed, while not
violating the standards and bothering users of other news readers
constantly.
Yours for solving problems at the proper level of abstraction.
thanx -len mcb@ncis.tis.llnl.gov (Michael C. Berch) (10/11/89)
In <40005@bu-cs.BU.EDU> tower@bu-it.bu.edu (Leonard H. Tower Jr.) writes: > nn could pass these header lines to inews: > Subject: Re: ORIGINAL TEXT > X-NN-Re-Count: 2 > or whatever the count is. [...] > > When nn sees an X-NN-Re-Count: header line. It would NOT display it, > but would display the Subject line as: > Subject: Re^2: ORIGINAL TEXT > or whatever the count is. Note that this is NOT the Subject line in > the news spool file. Why not just have NN count the number of Message-IDs in the "References:" header? -- Michael C. Berch mcb@tis.llnl.gov / uunet!tis.llnl.gov!mcb
amanda@intercon.com (Amanda Walker) (10/11/89)
In article <521@ncis.tis.llnl.gov>, mcb@ncis.tis.llnl.gov (Michael C. Berch) writes: > Why not just have NN count the number of Message-IDs in the > "References:" header? But, Michael, that would be *cheating* :-). On a more serious note, I've been working on the "broken References:" problem, and one workaround (read: "hack") that seems to help a lot in reconstructing References: chains is to look in the beginning of an article for "In article <msg-id>", and using anything that appears to augment the References: line if necessary. This does, of course, only work for articles with quoted material, but any information is better than none. Has anyone else done any experimentation with this sort of thing? -- Amanda Walker amanda@intercon.com
fritz@unocss.UUCP (Tim Russell) (10/12/89)
amanda@intercon.com (Amanda Walker) writes: |In article <521@ncis.tis.llnl.gov>, mcb@ncis.tis.llnl.gov (Michael C. Berch) |writes: |> Why not just have NN count the number of Message-IDs in the |> "References:" header? |But, Michael, that would be *cheating* :-). Actually, the latest release (I believe it's 6.3.7) does exactly that. But the other day, I followed up to an article, and the references line on it was "References: <numerous>". I really hope that was done by the user and someone's software out there isn't doing it. -- Tim Russell Univ. Of Nebr. at Omaha russell@{zeus.unl.edu | unoma1.bitnet} "I'm against any law that I wouldn't break if I could get away with it." -- A. Whitney Brown, SNL