david@ukma.UUCP (David Herron, NPR Lover) (03/07/85)
[ Bug salad ]
I just spent 3 hours going through a uucp map that Mark Horton provided me,
trying to get pathalias to compile the thing without error. With the raw
map + the dead sites used by Peter Honeyman + a couple suggested by
John Quarterman, I got almost 200 unreachable sites. These weren't caused
by mapping out the sites suggested by Honey and Quarterman. These were
plain incomplete information in the database.
For instance. There is a bunch of machines with names like "wisc-ai".
Looking in the data base I see that they are all part of "uwisc". But
information isn't given on connecting to "uwisc". So all these machines
are unreachable.
The isreal machines are unreachable because a connection from decvax
to their central node is mentioned at the isreal entry, but not at decvax.
(This last error is rather common. It occurs for some HP sites, some Purdue
sites, some Australian sites, and the Korean sites.)
(Ideally) problems like this are isolated to the people who really care.
The nearest neighbors to the problem. I don't need to know that hpda
connects a group of machines to hplabs. I just know that I don't know
a direct path to hpda, that hpda is in the hp subdomain, and that
hplabs takes care of routing within hp. (I'm assuming an address
of "user@hpda.hp.UUCP")
I'm interested in providing to my users the ability to use the RFC 822
style addressing with domains, etc. I have sendmail. Surely I can
make sendmail send mail to anywhere automagically. But a 2000+ site
name space is too much for me to manage all by myself. And unless
the name space is split up, I'll end up managing 2000+ sites. (sort of)
The only question in my mind is if the current division is correct.
--
:------------------------------------------------------------------:
:- David Herron -:
:- -:
:- ARPA-> "ukma!david"@ANL-MCS or david%ukma.uucp@anl-mcs.arpa -:
:- ARPA-> Or even anlams!ukma!david@ucbvax.arpa -:
:- -:
:- UUCP-> {ucbvax,unmvax,boulder,research}!anlams!ukma!david -:
:- UUCP-> {mcvax!qtlon,vax135,mddc}!qusavx!ukma!david -:
:- UUCP-> {A-Large-Portion-of-The-World}!cbosgd!ukma!david -:
:------------------------------------------------------------------:
"The home of poly-unsaturated thinking".honey@down.FUN (code 101) (03/07/85)
i have over a hundred "unreachable" hosts on my stderr from pathalias. this doesn't bother me a bit, since no one around here tries to send to those hosts. (as acting head of the complaint dept., i would know.) nontheless, i recently hacked pathalias so that if the link x -> y is declared, y -> x is presumed. i give the latter a very high cost so that the link will be used only as a last resort, unless of course the input data indicate a cheaper cost. this has the effect of putting the "unreachable" diagnostic in the output rather than stderr, but it seems to be what people prefer. i don't see what this has to do with domains, nor how domains would help. peter
honey@down.FUN (code 101) (03/07/85)
incidentally, uwisc is the arpa name for uwvax. other machines do the same thing: umcp-cs == maryland, bostonu == bu-cs, suny-sb == sbcs. (i think i have the last two right.) this seems a consummately stupid thing to do, and messes up pathalias, since the host name depends on the route. i'll never understand why people do things this way. peter
tracy@hcrvx1.UUCP (Tracy Tims) (03/08/85)
The real problem with a large flat namespace is that each administrator
on each site has the responsibility for fixing the bugs in the UUCP map
distribution (a pain) for all those many sites, and each time a new site
is to be added to the routing database each of those administrators has
to do it. This is an amazing duplication of effort. I am currently
doing the job at our site and it's too much for me! (And there is no
one else here with the time or the inclination to do it.)
How usable would the phone system be if every exchange needed detailed
routing information regarding every possible destination exchange in the
world?
Tracy Tims ihnp4!utzoo!hcr!hcrvx1!tracy
Human Computing Resources Corporation utcsri!hcr!hcrvx1!tracy
Toronto, Ontario, Canada. 416 922-1937 dciem!hcr!hcrvx1!tracydave@uwvax.UUCP (Dave Cohrs) (03/09/85)
> incidentally, uwisc is the arpa name for uwvax. other machines do the > same thing: umcp-cs == maryland, bostonu == bu-cs, suny-sb == sbcs. > (i think i have the last two right.) this seems a consummately stupid > thing to do, and messes up pathalias, since the host name depends on > the route. > > i'll never understand why people do things this way. > > peter Especially concerning uwisc -- that local-network entry was not set up correctly (not that it matters, if mail gets to uwvax, aliases will get it to its actual destination -- no matter). I think I've sent a correct version of the uwisc network to the uucpmap'ers, so when the next map comes out, it will be correct. The problem is, that corrections don't come out immediately, so when someone sends updates/corrections, you have to wait three months for them. Not that it really bothers me, I can understand the job it is to keep those maps up to date. By the by, uwisc is *not* uwvax's ARPAnet name, wisc-rsch is. uwisc is a backwards-compatable ARPAnet alias for wisc-crys and is also the official name of our local network. I know this seems confusing, it confuses me sometimes :-) -- dave cohrs ...!{allegra,harvard,ihnp4,seismo}!uwvax!dave dave@wisc-rsch.arpa (bug? what bug? that's a feature!)
hokey@plus5.UUCP (Hokey) (03/10/85)
The whole point of the mapping project is to fix things so individual mail administrators won't have to do anything to support the database. I don't want this to sound like a flame from left field, but I can't understand why the people in charge of the mapping project (Mark and Karen, at the top of the administrative heap, I believe) didn't run the maps through pathalias *before* posting the maps! The whole purpose of this second posting was to provide something cleaner than the first posting. The regional administrators should also have scrutinized the maps better. The local aliases should be removed. The bizzare weights should have been caught and fixed. (The map for usa.mo was a reposting of the original map; the corrected version I sent in was not posted, and the subsequent copy I sent to uucpmap was also never posted.) I was also amazed at the number of missing entries in the maps. My messages to the map folks regarding several of these oversights seem to have been ignored. I can't help but wonder if the maps were posted "as-is" just to "prove" to people that a flat namespace won't work. This doesn't make a lot of sense to me, because if the current administrators can't make these simple tables correct and complete, I don't know how anybody can expect the same bunch to make a domain space work. -- Hokey ..ihnp4!plus5!hokey 314-725-9492
dw@rocksvax.UUCP (Don Wegeng) (03/12/85)
In article <634@plus5.UUCP> hokey@plus5.UUCP (Hokey) writes: >...I can't >understand why the people in charge of the mapping project (Mark and >Karen, at the top of the administrative heap, I believe) didn't run >the maps through pathalias *before* posting the maps! The whole purpose >of this second posting was to provide something cleaner than the first >posting. The most optimal path to a site from all other sites is not unique. It will be different for every site. The uucp mapping project could not possibly generate a pathalias map for every site on the net. My current list contains over 3000 sites! Another thing to consider is that the network topology is in a constant state of flux; the path that is best today may not even work tomorrow. In order to maintain an accurate database of paths each site has to generate their own database, and update it at frequent intervals. /Don -- "Beware the Ides of March..." arpa: Wegeng.Henr@Xerox.ARPA uucp: {allegra,princeton,decvax!rochester,amd,sunybcs}!rocksvax!dw || ihnp4!tropix!ritcv!rocksvax!dw
root@bu-cs.UUCP (Barry Shein) (03/17/85)
[Re: comment about having more than one host name being stupid, reference to bu-cs == bostonu] It is sorta dumb when taken completely out of historical and other contexts...would that the world were such. 1. It is incredibly unfortunate that the research machine at Boston U was ever known by BOSTONU.CSNET. Unfortunate because there exists an older BOSTONU.BITNET which is a different machine (an IBM3081, main Academic computing facility). 2. To fix that problem I asked my users (mostly C.S. Faculty) what they would rather be called, they responded: BU-CS [as in MIT-MC, SU-AI etc I guess]. Fine, we are now BU-CS.CSNET and only respond to BOSTONU.CSNET for backwards compatibility...this will go away soon. 3. Not really solved...of course, DECNET won't allow an hyphen in a name (not that we are a DECNET node, it's 4.2bsd but there are enough decnet nodes on campus that this seemed like a good point.) So...we also respond locally to BUCSA and BUCS (the former a common convention on campus: BUdept[a,b,c...]. The latter sounding more like BUCS just with the hyphen removed. 4. Still not really solved, got a call from CSNET that *they* would prefer if we responded to BOSTONU-CS.CSNET cuz they were afraid people don't associate BU with BOSTON U. (?I dunno, I thought they would but I live here.) Sounded reasonable so.... I'm getting real sick of fighting this war! I guess the point is don't say 'stupid' so easily, assume the people in charge *do* probably know what they are doing and wonder *why*! On the other hand, I am not sure what the big deal is? They all work fine, your mail won't get lost, consider the others nicknames. If some program won't work well with this it really oughta get fixed (or just choose one name and stick with it, what's the problem with that???). We aren't the only ones, there are hundreds of hosts with nicknames. -Barry Shein, Boston University bzs@bostonu, bzs@bu-cs, bzs@bucsa, bzs@bucs,lordknows!bzs