[net.news] Black holes

rick@seismo.UUCP (Rick Adams) (09/09/84)

  From hao!hplabs!hpda!fortune!amd!decwrl!sunny Fri Sep  7 22:48:31 1984
  Relay-Version: version B 2.10.2 9/5/84; site seismo.UUCP
  Posting-Version: version B 2.10.1 6/24/83 SMI; site sun.uucp
  Path: seismo!hao!hplabs!hpda!fortune!amd!decwrl!sunny
  From: sunny@sun.uucp (Sunny Kirsten)
  Newsgroups: net.unix-wizards
  Subject: Re: Wanted:  contextual 'diff' program
  Message-ID: <1670@sun.uucp>
  Date: 8 Sep 84 02:48:31 GMT
  Date-Received: 9 Sep 84 03:55:53 GMT
  References: <236@callan.UUCP>
  Organization: Sun Microsystems, Inc.
  Lines: 13

Any ideas what happened to the sun in the path line? It looks
like the path should be ..... decwrl!sun!sunny

---rick
(A black hole is when you know the sun should be there
but you can't see it, right?)

chuqui@nsc.UUCP (Chuqles) (09/10/84)

We've seen the 'decwrl!sunny' problem before in replies to that address. it
SHOULD be 'decwrl!sun!sunny'. At one point I even remembered where the code
that was doing this bogosity was but I never tracked down a fix for it. Had
something to do with index() or rindex() and making unconsious assumptions
about the strings involved....

chuq

-- 
From the spotlight of the center ring:		Chuqles Von Rospach
{amd,decwrl,fortune,hplabs,ihnp4}!nsc!chuqui	nsc!chuqui@decwrl.ARPA

And now... Mutual of Omaha presents "Penguins: Antarctica's little clowns"

phil@amd.UUCP (Phil Ngai) (09/17/84)

> Any ideas what happened to the sun in the path line? It looks
> like the path should be ..... decwrl!sun!sunny

All I know is, it's been going on for a long time.
-- 
 Legalized gambling is a tax on the stupid.

 Phil Ngai (408) 982-6554
 UUCPnet: {ucbvax,decwrl,ihnp4,allegra,intelca}!amd!phil
 ARPAnet: amd!phil@decwrl.ARPA

shannon@sun.uucp (Bill Shannon) (09/18/84)

(How many times have I gone over this with how many people???....)

The problem is a bug in 2.10.1 news where it makes a check for
site name with prefix instead of strcmp (or something similar).
The bug is triggered at the receiving site (in our case, decwrl)
so there is nothing we can do to fix it.  I have reported the bug
and suggested (although still untested, I believe) fix to many
people who have complained.  Can someone confirm or deny whether
the bug has been fixed in 2.10.2?  (Will this become the equivalent
of the line-eater bug for 2.10.2?)

					Bill Shannon
					Sun Microsystems, Inc.

rick@seismo.UUCP (Rick Adams) (09/20/84)

> The problem is a bug in 2.10.1 news where it makes a check for
> site name with prefix instead of strcmp (or something similar).
> The bug is triggered at the receiving site (in our case, decwrl)
> so there is nothing we can do to fix it.  I have reported the bug
> and suggested (although still untested, I believe) fix to many
> people who have complained.  Can someone confirm or deny whether
> the bug has been fixed in 2.10.2?  (Will this become the equivalent
> of the line-eater bug for 2.10.2?)

This is the first I've heard of anything that specific.

Following is a fix for it. It is present in ALL 2.10 and later
B news versions (including 2.10.2).

---rick

*** header.c.bad	Wed Sep 19 19:53:47 1984
--- header.c		Wed Sep 19 18:39:59 1984
***************
*** 489,495
  	 * The old one thinks they both mean "Path" but only believes the
  	 * first one it sees, so will ignore the second.
  	 */
! 	if (prefix(hp->path, FULLSYSNAME))
  		fprintf(fp, "Path: %s\n", hp->path);
  	else
  		fprintf(fp, "Path: %s!%s\n", FULLSYSNAME, hp->path);

--- 489,496 -----
  	 * The old one thinks they both mean "Path" but only believes the
  	 * first one it sees, so will ignore the second.
  	 */
! 	if (prefix(hp->path, FULLSYSNAME) &&
! 		index(NETCHRS, hp->path[strlen(FULLSYSNAME)]))
  		fprintf(fp, "Path: %s\n", hp->path);
  	else
  		fprintf(fp, "Path: %s!%s\n", FULLSYSNAME, hp->path);