mju@mudos.ann-arbor.mi.us (Marc Unangst) (10/01/90)
> 1) Why would SCO ship brain_damaged binaries such as this? Because they're SCO. Next question? ;-) > 2) I thought it was 'rn' that had the .UUCP hardcode bug. Does > bnews also? I believe that SCO rn has been modified to get the domain the same way B News was (see below). > 3) Is it any wonder so many sites have this flaw (I see them > daily) if this is the case? [danger: rhetorical question] Not really. > 4) How would this user, *with the least effort* get his own > version of bnews compiled? (i.e. did SCO do weird things to news > to make it work on their box?) You don't have to do that. I'm not sure if this is documented or not, but if you put your domain name, with a leading dot, in /usr/lib/news/localdomain, SCO's news will use that instead of .UUCP. For example, for my site (mudos.ann-arbor.mi.us), I would put ".ann-arbor.mi.us" in /usr/lib/news/localdomain. FYI, I believe there's a similar file in /usr/lib/news called "control" or "forward" or something like that, which tells B News whether to automatically execute control messages, or whether an administrator should be notified. The default is to automatically do everything, which is probably okay for somebody who just wants to type "custom", config the software, and forget about it -- but probably not okay for people who actually know something about Usenet. I can look around and see what the exact filename is if people are interested. This works on the version of B News and Xenix that I happen to have around. The usual disclaimers apply, of course, and if you *can* compile News yourself you might want to do so anyway. -- Marc Unangst | "da-DE-DA: I am sorry, the country you have mju@mudos.ann-arbor.mi.us | dialed is not in service. Please check the ...!umich!leebai!mudos!mju | number and try again." -- Telecom Kuwait
wb8foz@mthvax.cs.miami.edu (David Lesher) (10/01/90)
I recently exchanged mail with a user down under about his bogus .UUCP suffix, instead of the correct .oz.au that he showed in his .sig files. He told me that SCO had shipped them only binaries for bnews, with the .UUCP hardcoded into it. 1) Why would SCO ship brain_damaged binaries such as this? 2) I thought it was 'rn' that had the .UUCP hardcode bug. Does bnews also suffer? 3) Is it any wonder so many sites have this flaw (I see them daily) if this is the case? [danger: rhetorical question] 4) How would this user, *with the least effort* get his own version of bnews compiled? (i.e. did SCO do weird things to news to make it work on their box?) comments/flames/free_beer? -- A host is a host from coast to coast.....wb8foz@mthvax.cs.miami.edu & no one will talk to a host that's close............(305) 255-RTFM Unless the host (that isn't close)......................pob 570-335 is busy, hung or dead....................................33257-0335
cat@tygra.ddmi.com (CAT-TALK Maint. Account) (10/01/90)
In article <1990Sep30.220547.7205@mthvax.cs.miami.edu> wb8foz@mthvax.cs.miami.edu (David Lesher) writes: }I recently exchanged mail with a user down under about his bogus }.UUCP suffix, instead of the correct .oz.au that he showed in }his .sig files. } }He told me that SCO had shipped only them only binaries for }bnews, with the .UUCP hardcoded into it. } All he has to do is put the line ".oz.au" as the only line in the file /usr/lib/news/localdomain and make the file world-readable. I too have the pre-compiled SCO Bnews and thats how I got around it.
alan@mq.com (Alan H. Mintz) (10/02/90)
In article <1990Sep30.220547.7205@mthvax.cs.miami.edu>, wb8foz@mthvax.cs.miami.edu (David Lesher) writes: > I recently exchanged mail with a user down under about his bogus > .UUCP suffix, instead of the correct .oz.au that he showed in > his .sig files. > > He told me that SCO had shipped only them only binaries for > bnews, with the .UUCP hardcoded into it. > > 1) Why would SCO ship brain_damaged binaries such as this? I don't know about _brain-damaged_. Not everyone has a registered domain address. Such sites are usually distinguished by the .UUCP suffix, which is known to be bogus. SCO provides a service to people without access to a developement system by shipping a compiled Bnews distribution. Documentation regarding the configuration used to compile it is sparse, but I suppose it wouldn't mean much to those requesting it anyway. Exploring around in the binaries let me patch mine to change the uucp suffix to com. In poking around just now, I discovered that, if there is a file named "localdomain" in your /usr/lib/news directory, Bnews will create your system name by appending whatever is in the "localdomain" file to the end of the system name as contained in /etc/systemid. Since /etc/systemid appears in the text portion of the binaries, I suppose that you could patch it to look elsewhere for the system name as well. > 2) I thought it was 'rn' that had the .UUCP hardcode bug. Does > bnews also? The above seems to apply to all the bnews stuff (UUCP can be found in most of the binaries.). You can patch them to .COM (or whatever) if appropriate, or use the "localdomain" file method. > comments/flames/free_beer? Yes. No. YES!!!! -- < Alan H. Mintz | Voice +1 714 980 1034 > < Micro-Quick Systems, Inc. | FAX +1 714 944 3995 > < 10384 Hillside Road | uucp: ...!uunet!mq!alan > < Alta Loma, CA 91701 USA | Internet: alan@MQ.COM >
wb8foz@mthvax.cs.miami.edu (David Lesher) (10/02/90)
In <95@mq.com> alan@mq.com (Alan H. Mintz) writes: >I don't know about _brain-damaged_. Not everyone has a registered >domain address. Such sites are usually distinguished by the .UUCP suffix, >which is known to be bogus. SCO provides a service to people without >access to a developement system by shipping a compiled Bnews distribution. >Documentation regarding the configuration used to compile it is sparse, but >I suppose it wouldn't mean much to those requesting it anyway. Exploring >around in the binaries let me patch mine to change the uucp suffix to com. Alas, you cannot fit "oz.au" in the space "UUCP" provides. The fellow tried. I did pass along the localdomain stuff, and I'll wait to see if that helps. I've got nothing against SCO shipping binaries. I do object to SCO shipping ONLY binaries. -- A host is a host from coast to coast.....wb8foz@mthvax.cs.miami.edu & no one will talk to a host that's close............(305) 255-RTFM Unless the host (that isn't close)......................pob 570-335 is busy, hung or dead....................................33257-0335
bernie@DIALix.oz.au (Bernd Felsche) (10/02/90)
In article <95@mq.com> alan@mq.com (Alan H. Mintz) writes: >In article <1990Sep30.220547.7205@mthvax.cs.miami.edu>, wb8foz@mthvax.cs.miami.edu (David Lesher) writes: >> I recently exchanged mail with a user down under about his bogus >> .UUCP suffix, instead of the correct .oz.au that he showed in >> his .sig files. Yup, it's me... I've had to bash the Sender: and Reply-To: by hand. See the bruises? :-) >> He told me that SCO had shipped only them only binaries for >> bnews, with the .UUCP hardcoded into it. >> >> 1) Why would SCO ship brain_damaged binaries such as this? > >I don't know about _brain-damaged_. Not everyone has a registered >domain address. Such sites are usually distinguished by the .UUCP suffix, >which is known to be bogus. SCO provides a service to people without >access to a developement system by shipping a compiled Bnews distribution. This is very amusing, since the development system was ordered at the same time from the same distributor. >Documentation regarding the configuration used to compile it is sparse, but >I suppose it wouldn't mean much to those requesting it anyway. Exploring >around in the binaries let me patch mine to change the uucp suffix to com. Well, 100+ pages is not sparse. If only I could do what it promises, if I had the source. Unfortunately, my suffix is oz.au, too long to fit in UUCP. >In poking around just now, I discovered that, if there is a file named >"localdomain" in your /usr/lib/news directory, Bnews will create your >system name by appending whatever is in the "localdomain" file to the Doesn't work on my version, in SCO Unix. I've tried compiling C-News, but I'm having serious problems. First, filename length of 14 characters is strictly enforced. I had to unarchive the sources on my Amiga (seriously), rename long files to something unambiguous, and then send them back up. There are numerous compiler warnings, from all modules, and include files aren't found. I'd post a list of errors, but that's too depressing. Has anybody got C-News working on SCO Unix? bernie
bernie@DIALix.oz.au (Bernd Felsche) (10/02/90)
In article <413@tygra.ddmi.com> cat@tygra.UUCP (CAT-TALK Maint. Account) writes: > >All he has to do is put the line ".oz.au" as the only line in the file >/usr/lib/news/localdomain and make the file world-readable. I too have >the pre-compiled SCO Bnews and thats how I got around it. This: -rw-r--r-- 1 news news 7 Oct 2 22:10 /usr/lib/news/localdomain contains only this: .oz.au to give me this: Newsgroups: comp.mail.misc Subject: Re: Bogus .UUCP addresses under SCO Summary: Expires: References: <1990Sep30.220547.7205@mthvax.cs.miami.edu> <413@tygra.ddmi.com> Sender: Reply-To: bernie@DIALix.UUCP (Bernd Felsche) Followup-To: Distribution: Organization: DIALix Services, Perth Western Australia Keywords: It doesn't work :-( :-(, and I wish it did. I'm running SCO Unix. I hear it works for Xenix. bernie
peter@ficc.ferranti.com (Peter da Silva) (10/03/90)
Put a shell script in front of "inews" that munges the ".UUCP" in the headers into ".oz.au". #!/bin/sh sed -e '1,/^$/s/DIALix.UUCP/DIALix.oz.au/' | /usr/lib/news/inews.real (been 6 months since I hacked on B news, so it may take a bit more work... but the principle is valid: I've done the same to fix mail problems here at Ferranti before I got smail running) -- Peter da Silva. `-_-' +1 713 274 5180. 'U` peter@ferranti.com
wengland@stephsf.stephsf.com (Bill England) (10/05/90)
In article <1990Oct1.011018.10124@mthvax.cs.miami.edu> wb8foz@mthvax.cs.miami.edu (David Lesher) writes: [...] >1) Why would SCO ship brain_damaged binaries such as this? > The .uucp suffix is a Configuration feature of rn. It is HARD wired and there is nothing you can do about it accept recompile rn. SCO is not likely to redesign the 'free' software. The Reply-To: line is now obsolete anyway! Look at the header and see how redundant a Reply-To: line would be. The solution is to NOT include a Reply_To: line in your mail header. ( Removing this also requires that you have the rn source and recompile. ) > >2) I thought it was 'rn' that had the .UUCP hardcode bug. Does bnews >also suffer? > bnews suffers here all the time. Esspecially when I strap it to the rack and lash it with the bull wip. ;-) ( No the .uucp is an rn problem only. ) > >3) Is it any wonder so many sites have this flaw (I see them daily) if >this is the case? [danger: rhetorical question] > >4) How would this user, *with the least effort* get his own version of >bnews compiled? (i.e. did SCO do weird things to news to make it work >on their box?) I think that it compiles almost straight out of the box. Why don't you try C_news instead. if you are going to get the sources you might as well stay one step ahead of all the other SCO installations. __ALSO__ 1) GET trn. It fixes the rn problem mentioned and you have the *latest* source for rn AND trn. (trn is 'threaded read news' and is a replacment for rn that includes a thread database facility similar to nn. Sharp!) 2) NOTICE, that when you save mail in the rn SCO provides it does not put in the string of ^A's that are required for Mmdf mail to work? The following mbox.saver file comes with trn and should fix the mail save problem in rn as well. -----/usr/lib/mail/trn/mbox.saver-------- : use /bin/sh # $Header: mbox.saver.SH,v 4.3.3.1 90/08/20 16:30:28 davison Trn $ # # $Log: mbox.saver.SH,v $ # Revision 4.3.3.1 90/08/20 16:30:28 davison # Added support for MMDF format mailboxes. # # Revision 4.3.2.2 90/03/17 20:44:54 sob # Modify Article header to place the colon after Article. # # Revision 4.3.2.1 89/11/28 00:05:47 sob # Branch for RN/RRN combo patches # # Revision 4.3.1.2 85/05/20 15:55:37 lwall # Turned makedir.SH into $5. # # Revision 4.3.1.1 85/05/10 11:35:30 lwall # Branch for patches. # # Revision 4.3 85/05/01 11:42:51 lwall # Baseline for release with 4.3bsd. # # # Arguments: # 1 Full name of article (%A) # 2 Public news spool directory (%P) # 3 Directory of current newsgroup (%c) # 4 Article number (%a) # 5 Where in article to start (%B) # 6 Newsgroup name (%C) # 7 Save destination (%b) # 8 First line of message, normally From... # export PATH || (echo "OOPS, this isn't sh. Desperation time. I will feed myself to sh."; sh $0; kill $$) ( /bin/echo "" if /bin/test "$5" = 0 -a ! "$4" = 0 ; then /bin/echo "Article: $4 of $6" fi /bin/tail +$5c $1 /bin/echo "" /bin/echo "" /bin/echo "" ) >> $7 -----End of File------ +-------- | Bill England | wengland@stephsf.COM | * * H -> He +24Mev * * * ... Oooo, we're having so much fun making itty bitty suns * * *
miller@ria.ccs.uwo.ca (J G Miller) (10/05/90)
In article <373@stephsf.stephsf.com> wengland@stephsf.stephsf.com (Bill England) writes: > The Reply-To: line is now obsolete anyway! Look at the header > and see how redundant a Reply-To: line would be. The solution is > to NOT include a Reply_To: line in your mail header. ( Removing > this also requires that you have the rn source and recompile. ) Doesn't this version of rn allow one to specify one's Reply-To address through the use of the MAILHEADER environment setting? -- J G Miller telephone: (519) 679-2111 extension 6325 Room 024, Chemistry Building InterNet: <a4346@uwocc1.uwo.CA> University of Western Ontario NetNorth: <A4346@UWOCC1.BITNET> LONDON, Ontario, N6A 5B7 mailpath: {uunet!watmath,utai}!ria!a4346
wengland@stephsf.stephsf.com (Bill England) (10/06/90)
In article <1990Oct5.001652.17931@mthvax.cs.miami.edu> wb8foz@mthvax.cs.miami.edu (David Lesher) writes: >And goes on to recommend cnews and trn. > >As I pointed out, SCO had not given him source. I would not accept such >either, hence my query (still unresolved) as to why SCO was not >including the source. I can see SCO's perspective on providing Binarys. After all quite a few sites are on the net simply because the binarys were made available. They should ALSO provide a 'developers kit' that contains the sources for those files. We are, however, lucky that they even provided the binarys ... > >I strongly disagree as to the value of a Reply-to: header. NOT ALL >SITES accept mail at the posting location. Some *try* to, and often >fail. For example, there's that little outfit that used to install TWX >machines, what WAS their name? Oh yes: ATT.... ;_] > Sure. Repy_To: is available if needed. Most sites do accept e-mail though and do not need the Reply_To: Line. Those sites that do need a reply_to need it very badly of course. ;-) >I had already pitched 'nn' to the oz chap before asking on the net. We >don't keep {t,r}rn around here any more since nn is used virtually >exclusively. [nn,trn] tis a matter of choice. The thing I don;t like about trn is that you can't follow threads that span news groups. (Like this thread for instance.) I have heard some talk of news group spanning enhancments to trn though. 'trn' does have the advantage of compiling under SCO-ODT straight out of the box. My suggestion to use C News may be a bit premature. I understand that there are some problems with C News file names greater than 14 chars. I'm not running C News yet but it is high on my list of software to install. ( I might have to make that OS mod for longer file names ... Hmmm.) >-- >A host is a host from coast to coast.....wb8foz@mthvax.cs.miami.edu +-------- | Bill England | wengland@stephsf.COM | * * H -> He +24Mev * * * ... Oooo, we're having so much fun making itty bitty suns * * *