From jik@cam.ov.com Tue Dec 13 13:35:33 EST 1994
Article: 2707 of news.software.b
Path: news.pop.psu.edu!hudson.lm.com!godot.cc.duq.edu!newsfeed.pitt.edu!gatech!bloom-beacon.mit.edu!pad-thai.cam.ov.com!usenet
From: jik@cam.ov.com (Jonathan I. Kamens)
Newsgroups: news.software.b,news.software.nntp
Subject: INN 1.4 misparses newsfeeds if ME entry has exclusions
Date: 13 Dec 1994 17:38:02 GMT
Organization: OpenVision Technologies, Inc.
Lines: 32
Distribution: inet
Message-ID: <3ckm5q$282@pad-thai.cam.ov.com>
NNTP-Posting-Host: gza-client1.cam.ov.com
Xref: news.pop.psu.edu news.software.b:2707 news.software.nntp:7381

My version of INN (RELEASE 1, PATCHLEVEL 4, DATE 20-Mar-93, with some security
patches) has a bug in it which prevents the newsfeeds file from being parsed
correctly if the ME entry has exclusions in it (e.g., if you have
"ME/cyber.sell.com:" instead of "ME:" :-).

The following patch seems to fix the problem:

*** newsfeeds.c	1994/12/13 17:27:12	1.1
--- newsfeeds.c	1994/12/13 17:28:57
***************
*** 501,507 ****
  
      for (sp = Sites, errors = 0, i = 0; i < nSites; i++) {
  	p = strings[i];
! 	if (p[0] == 'M' && p[1] == 'E' && p[2] == NF_FIELD_SEP) {
  	    if ((error = SITEparseone(p, &ME, subbed)) != NULL) {
  		syslog(L_FATAL, "%s bad_newsfeeds %s", MaxLength(p, p), error);
  		errors++;
--- 501,508 ----
  
      for (sp = Sites, errors = 0, i = 0; i < nSites; i++) {
  	p = strings[i];
! 	if (p[0] == 'M' && p[1] == 'E' &&
! 	    ((p[2] == NF_FIELD_SEP) || (p[2] == NF_SUBFIELD_SEP))) {
  	    if ((error = SITEparseone(p, &ME, subbed)) != NULL) {
  		syslog(L_FATAL, "%s bad_newsfeeds %s", MaxLength(p, p), error);
  		errors++;

I don't know if this is fixed in later versions of INN.

-- 
Jonathan Kamens  |  OpenVision Technologies, Inc.  |   jik@cam.ov.com