asp@COS.COM (Andrew S. Partan) (08/27/87)
I just noticed that I have some duplicate lines in my pathalias output
file. E.g.:
.uucp decuac!%s 200
.uucp sundc!sun!intsc!%s 195
There are about 21 such lines (out of a total of 10093 lines in the
file).
I run pathalias as the notes to smail 2.3 suggested (in the man page
for pathproc(8)):
pathalias -c [a-z]*.*Z | pathproc > /tmp/paths
I am running pathalias 8.2 (from approx Jan '86).
Other information (probably not relevant):
Sun 3/280, Sun OS 3.3, mostly 4.2 BSD
Does any one have any idea what might be going on or how to fix it?
One thought that I had was to update the pathproc script to check for
this and delete the more expensive of the two paths. However, I
thought that pathalias was supposed to do this already (only emit the
less expensive path).
Thanks for any light that anyone can shed on the subject.
--asp (Andrew Partan @ Corporation for Open Systems)
-- asp@cos.com
-- {uunet, seismo!sundc, decuac, hadron, hqda-ai}!cos!asp
--
--asp (Andrew Partan @ Corporation for Open Systems)
-- asp@cos.com
-- {uunet, seismo!sundc, decuac, hadron, hqda-ai}!cos!aspKarl.Kleinpaste@cbstr1.att.com (08/28/87)
This looks like a variation of a problem I found quite a while back in the Jan 86 pathalias release. It seems that, when constructing paths into domains, pathalias doesn't notice that (e.g.) .att.com is the composite construction of .att and .com, and can generate two routes to the single domain string .att.com. This is due to the fact that I had put in some extra pathalias info in my private data to describe the .com, .gov, and .edu domains before the officially-distributed stuff seemed to do it properly; it conflicted in this way with a line in one of the domain files that said that ihnp4 was the .att.com gateway, whereas I had broken .com down a bit, with cbosgd listed as my .att gateway. Poof, 2 routes for .att.com, through ihnp4 and cbosgd. Your bug seems similar, but certainly not the same - the domain in question for you is only a one-domain-name-component item. I wrote to Peter about it, but I don't think it was ever resolved. Karl
john@xanth.UUCP (08/30/87)
In article <419@cos.COM>, asp@COS.COM (Andrew S. Partan) writes: > .uucp decuac!%s 200 > .uucp sundc!sun!intsc!%s 195 > pathalias -c [a-z]*.*Z | pathproc > /tmp/paths What's happening is that you are not running pathalias with the -i (ignore case) flag, so it considers case significant. intsc lists itself as a ".UUCP" forwarder (a problem in itself), and the input to pathproc has a ".uucp" and a ".UUCP" line. pathproc down-cases both of them, and .... I run pathalias -i and don't use pathproc, but I don't need to costs either - I have smail set up to always try the call. -- John Owens Old Dominion University - Norfolk, Virginia, USA john@ODU.EDU old arpa: john%odu.edu@RELAY.CS.NET +1 804 440 4529 old uucp: {decuac,harvard,hoptoad,mcnc}!xanth!john