From:	CRDGW2::CRDGW2::MRGATE::"SMTP::KSUVAX1.CIS.KSU.EDU::MCCALL!NEWSMGR" 22-JUN-1990 17:39:19.42
To:	MRGATE::"ARISIA::EVERHART"
CC:	
Subj:	RE: Problems with MAKEPATHS

Received:  by crdgw1.ge.com (5.57/GE 1.70)
	 id AA24564; Fri, 22 Jun 90 17:30:55 EDT
Resent-From: mccall!newsmgr@ksuvax1.cis.ksu.edu
Return-Path: mccall!newsmgr@ksuvax1.cis.ksu.edu
Received: from harris.cis.ksu.edu by DRYCAS.CLUB.CC.CMU.EDU; Fri, 22 Jun 90
 17:06 EST
Received: from ksuvax1.cis.ksu.edu by harris.cis.ksu.edu (5.58/SWH-2.03); id
 AA18747; Fri, 22 Jun 90 16:06:30 CDT
Received: by ksuvax1.cis.ksu.edu (5.59++/CIS1.1) id AA27980; Fri, 22 Jun 90
 16:06:22 CDT
Date: 21 Jun 90 22:20:35 PDT
From: jeh@dcs.simpact.com
Subject: RE: Problems with MAKEPATHS
To: vmsnet@DRYCAS.CLUB.CC.CMU.EDU
Message-Id: <1383.26814833@dcs.simpact.com>
Resent-Date: Fri, 22 Jun 90 17:09 EST
Resent-To: VMSNET-LIST@DRYCAS.CLUB.CC.CMU.EDU
Resent-Message-Id: <7694F3A13A590011FB@DRYCAS.CLUB.CC.CMU.EDU>
Organization: Simpact Associates, San Diego CA
X-Envelope-To: everhart%arisia.decnet@CRDGW1.GE.COM
X-Vms-To: vmsnet@DRYCAS.CLUB.CC.CMU.EDU

In article <1990Jun21.171357.918@mdcbbs.com>, spain@mdcbbs.com writes:
> I can no longer get my PATHS file to build.  When I submit MAKEPATHS to batch
> the subprocess just runs and runs (the last time I tried, I let it run for
> about 24 hours).
> 
> Has anyone encounted problems with MAKEPATHS?  Any suggestions on where to
> start debugging?

Harrison, this was discussed at GREAT length in comp.mail.uucp earlier this
month, and I copied a solution here... oh well, here it is again.  

X-NEWS: simpact vmsnet.uucp: 94
Path: simpact!jeh
From: jeh@dcs.simpact.com
Newsgroups: vmsnet.uucp
Subject: Pathalias breaks on latest map updates (solution)
Message-ID: <1330.266ac0c1@dcs.simpact.com>
Date: 4 Jun 90 20:12:49 PDT
Organization: Simpact Associates, San Diego CA
Lines: 55

Several of us have found that pathalias loops forever with the latest batch
of map updates.  The problem is due to a bug in the maps.  Here is one of 
several recent postings in comp.mail.uucp on this subject:

> X-NEWS: simpact comp.mail.uucp: 888
> Relay-Version: VMS News - V5.9C 19/12/89 VAX/VMS V5.2; site dcs.simpact.com
> Path: simpact!crash!bblue
> Newsgroups: comp.mail.misc,comp.mail.uucp
> Subject: Re: pathalias breaks on latest map distribution - HELP!
> Message-ID: <2902@crash.cts.com>
> From: bblue@crash.cts.com (Bill Blue)
> Date: 30 May 90 05:40:58 GMT
> References: <1990May29.182555.27806@pmsmam.uucp> <3438@rex.cs.tulane.edu>
> Organization: Crash TimeSharing, El Cajon, CA
> Lines: 30
> Xref: simpact comp.mail.misc:1694 comp.mail.uucp:888
> 
> In article <3438@rex.cs.tulane.edu> mb@rex.cs.tulane.edu (Mark Benard) writes:
> >In article <1990May29.182555.27806@pmsmam.uucp> wwm@pmsmam.uucp (Bill Meahan) writes:
> >>Since the latest distribution I've received from comp.mail.maps, my attempts
> >>to build a new pathalias database result in pathalias hanging forever while
> >>taking up about 4.5 MEGS of core.
> >>
> >>After spending far too much time chasing the problem, I've found that
> >>keeping u.usa.nm.1 out of the input stream solves(?) the problem.  Apparently
> >>something in the lastest bunch of maps (which do NOT include nm.1 so far)
> >>is inconsistent with the Apr 25 version of u.usa.nm.1.
> >
> >I had the same problem and tracked it down to an infinite loop involving
> >bcsinc (in u.usa.ca.1) and rt1 (in u.usa.nm.1).  Since the NM map was not
> >changed recently and that CA map arrived last night, I concluded that it was
> >a problem with bcsinc.  Sure enough, commenting the bcsinc entry out of the
> >CA map solved the problem.
> >
> >However, I do not know exactly why the loop occurred.  Can anyone explain?
> 
> It's even more interesting when you consider that the bcsinc map
> entry hasn't changed since late August '89, and the rti entry
> hasn't changed since mid September '89.  Something else seems to
> be bringing this to a head.
> 
> --Bill Blue
> 	Southern California, Arizona Regional Uucpmap Coordinator
>         UUCP: {nosc, ucsd, hplabs!hp-sdd}!crash!bblue
>         ARPA: crash!bblue@nosc.mil
>         INET: bblue@crash.cts.com

So there you have it -- comment the bcsinc entry out of u.usa-ca-1 .
Apparently there is a new version of pathalias which is immune to this 
problem, but I doubt that we'll get it into the 1.2 release. 

Update:  No, we didn't get it into 1.2.  Tom is working on porting the
new pathalias.  We'll post it here when it's ready.  

Some folks also found that another solution was to use an older map's version 
of ucsd's map entry (they're in u.usa-ca-n, n is a large number) .  

	--- Jamie Hanrahan, Simpact Associates, San Diego CA
Chair, VMSnet [DECUS uucp] and Internals Working Groups, DECUS VAX Systems SIG 
Internet:  jeh@dcs.simpact.com, or if that fails, jeh@crash.cts.com
Uucp:  ...{crash,scubed,decwrl}!simpact!jeh