dave@lsuc.UUCP (01/08/87)
Would anyone care to pull all of the #U lines out of
mod.map's u.can.on.1 file and contruct a visual map that
shows the news paths within Toronto? I've lost track of
what the major sites are. I wonder whether we might also
consider introducing some in-Toronto redundancy by picking
two far-apart nodes in this map and have them exchange
some newsgroups.
(A number of sites already exchange tor.* and ont.* with UUCP
neighbours who aren't news neighbours, by the way.)
David Sherman
--
{ ihnp4!utzoo seismo!mnetor utai watmath decvax!utcsri } !lsuc!davedon@clunk.UUCP (Don McKillican) (01/10/87)
In article <1497@lsuc.UUCP>, dave@lsuc.UUCP (David Sherman) writes: > Would anyone care to pull all of the #U lines out of > mod.map's u.can.on.1 file and contruct a visual map that > shows the news paths within Toronto? I wouldn't know how to isolate the news paths, but here is a first attempt at a map of Toronto's uucp connections in general. I should explain a few things. The first difficulty encountered by your would-be map-maker-of-toronto is that everybody and their Apple ][ has connections to one (and usually more) of {utai,utcs,utcsri,utzoo}. I saw so many machines with connections to both utcs and utzoo that I am beginning to suspect they are really the same machine :-). Anyway, rather than try to put these machines in the middle of a spider-web of lines, I decided to indicate with the machine name which of these four U of T sites they connect to. So "ms(a)" means that machine "ms" has a connection to utai, in addition to any other lines you see on the map. And "lsuc(crz)" means that lsuc has connections to utcs, utcsri and utzoo, respectively. Thus if you see no easy connection between lsuc and "darwin(cz)", you can also connect via lsuc!utcs!darwin. And I believe you should be able to get from any of the four U of T sites to any other of the four, so you can get from lsuc to ms via lsuc!utcs!utai!ms. The connections involving these four U of T sites are summarized at the bottom. Secondly, since this map is quite likely going to be used by people (like me) trying to figure out how to get mail from mysite to yoursite, I have tried to include only "good" paths. A "good" path is basically either HOURLY or better, or the only path there is, for machines with very few connections. I have also left out what looked to me to be "internal" machines like hcrvx1, and represented the site only by its "external" node hcr. I don't personally think that internal machines should have map entries in u.can.on.1 in the first place, but that's another issue. I have also left off all the internal U of T sites, just because I haven't figured out how to make a map for them yet. I am still trying, and I'll post it as a supplement when I figure it out. Sorry folks. Thirdly (this will be a shock), there are Errors in this map. I'm sure I don't have the most recent u.can.on.1, I'm sure that the most recent u.can.on.1 has errors too, and I'm extremely sure that I wouldn't get the whole map right the first time even if my u.can.on.1 was perfect :-). (I even painted myself into a corner with my own machine: clunk has a good connection to lsuc (as of a week ago), and I couldn't find a way to get it on the map. sniff) But I would be more than happy to accept corrections to this map and maintain it: this was FUN! So look at your part of the map and tell me where I goofed. Maybe in the process we can even lick u.can.on.1 into shape :-). Okay, 'nuff said. Presenting.... The Usenet Map of Toronto ========================= (legend: a-utai, c-utcs, r-utcsri, z-utzoo. Capitals indicate out-of-city sites) GANDALF | rhodnius(cz) | | ms(a) sq(cs)----parkridge---hcr(cz)---dciem(rz)-----------NRCAER | / | \ / |\ / | / | \ omnitor | \ / darwin(cz)--chp(z)--trigraph(crz) | | per---zorac / | \ oscvax(cz) | / / | \ |/ / msitxt ontmoh ryesone(crz)------lsuc(crz)---------- | | | \ jdpsys | ---geac tmsoft logics(c) | radha \ | / \ \ | / mmainc--lethe---------yetti(cz)--------mnetor(cz)------------SEISMO / | \ / | \ / | \ eclectic dmh tsdc / | toram(z) | syntron(cz) / | | | / utrc-2at cmq1---genat(c)---clunk---bcosi methods | \ / | | \ | braegen | | \ | | | MACCS----------------------+ forgene(c) aesat(z) romwa(c) sickkids(cz) suncan(z) =============================================================================== U of T Summary LOCAL SITES: UTAI UTCS UTCSRI UTZOO --------------------------------------------------------------------- lsuc darwin dciem aesat ms logics hcr chp lsuc lsuc darwin mnetor rhodnius dciem oscvax ryesone lsuc romwa trigraph mnetor ryesone suncan oscvax sickkids parkridge sq rhodnius syntron ryesone trigraph sickkids yetti sq syntron toram trigraph REMOTE SITES: UTAI UTCS UTCSRI UTZOO ------------------------------------------------------------------ ucb-vision bnr-vpa deepthot bnr-vpa watmath cbosgd mbcsd deepthot mcgill-vision ihnp4 sask uw-beaver
dave@lsuc.UUCP (David Sherman) (01/11/87)
Hmm. Thanks, Don. I didn't even suggest anyone try doing
a UUCP map, because (1) pathalias can give routings and (2)
the map would be so complex. Your attempt is nice, but is
(just for example) missing lsuc's connections to parkridge,
hcr, nrcaer, toram, maccs and sickkids, all of which have been
in u.can.on.1 for a while. (We also have connections to sq, tmsoft
and clunk, but those are relatively new, plus mkv020 which isn't on
your map, and a few out of town sites.)
I think a UUCP map of Toronto is probably not doable in two dimensions.
The Usenet map would be simpler. Read the #U line for each site
from u.can.on.1, and that tells you who it exchanges newsgroups with
(not necessarily all netnews, but at least some). I would think sites
that exchange only "tor,ont" or whatever wouldn't list each other
as #U (I only list, for lsuc, sites with which we have a significant feed).
David Sherman
The Law Society of Upper Canada
--
{ ihnp4!utzoo seismo!mnetor utai watmath decvax!utcsri } !lsuc!daverayan@ai.toronto.edu (01/12/87)
In article <104@clunk.UUCP> don@clunk.UUCP (Don McKillican) writes: # I don't personally think that internal machines should have # map entries in u.can.on.1 in the first place, but that's another issue. The policy until now is that any host that might possibly send out mail (and therefore may receive mail from far away places) may put a map entry in u.can.on. Hopefully, when sites with internal hosts are able to either do hostname hiding, or switch to domain-based addresses, map entries for internal machines will go away. In a place the size of UofT, there will still be various separate clusters of machines where is makes sense that each cluster have an entry in the map; I have already started doing that with various natural clusters within UofT. Other sites should feel free to copy that model in their future map entries (see utecfa for a good example). # I have also left off all the internal U of T sites, just because I haven't # figured out how to make a map for them yet. I am still trying, and I'll # post it as a supplement when I figure it out. Sorry folks. Don't be sorry; we have a real hard time keeping track ourselves. See below for an approximate model. # Thirdly (this will be a shock), there are Errors in this map. I'm sure # I don't have the most recent u.can.on.1, I'm sure that the most recent # u.can.on.1 has errors too, and I'm extremely sure that I wouldn't get # the whole map right the first time even if my u.can.on.1 was perfect :-). Speaking of which... I have started weekly postings to mod.map of the u.can.* files (distribution: can) that have been changed since the last posting. This is in response to confusion arising from out-of-date maps appearing in the monthly mod.map posting. I'd really have preferred to just post diffs, but that involves various logistical problems for the people who want to make use of this info. I would also have liked to use another group (can.map ?) to avoid mixing older maps from the monthly posting with newer maps from the local weekly posting. Any preferences? While I'm at it, the map entries for the following Toronto hosts are more than 6 months out of date and possibly incorrect: aesat: #W ; Mon Mar 11 01:30:50 GMT 1985 braegen:#W braegen!ken (Ken Marchant); Wed Mar 19 12:51:49 EST 1986 chpx: #W ian@darwin (Ian Darwin); Wed May 28 14:14:48 EDT 1986 hcradm: #W hcrvx1!mark (Mark Ashworth); Tue Dec 18 05:15:48 EST 1984 logics: #W logics!onn (Brian Onn); Thu May 1 21:34:47 EDT 1986 ms: #W ms!mts (Martin Stanley); Thu Sep 12 13:19:19 EDT 1985 omnitor:#W omnitor!mherman (Michael Herman); Mon Dec 9 05:24:29 EST 1985 ontmoh: #W ian@darwin (Ian F. Darwin); Mon Mar 10 22:37:29 EST 1986 radha: #W radha!sanand (Sanand Patel); Fri Mar 7 10:03:34 EST 1986 rhodnius:#W utcs!geoff (Geoff Collyer); Sat Jul 14 04:54:16 EST 1984 romwa: #W romwa!mark (Mark T. Dornfeld); Mon Jun 9 10:13:48 EDT 1986 ryesone:#W ryesone!mason (Dave Mason);Sat Jan 4 00:23:51 EST 1986 sickkids:#W cbosgd!ksh (paper form); Fri Jul 13 05:11:50 EST 1984 toram: #W toram!chris (Chris Robertson); Mon Dec 23 11:24:24 EST 1985 utcseri:#W trigraph!bjf (Bruce Freeman); Tue Jan 15 04:55:57 GMT 1985 utfang: #W geoff@utcs (Geoff Collyer), Sat Jan 18 01:50:05 EST 1986 utflis: #W utai!path (Path Database); Tue Jan 14 04:53:40 EST 1986 utfyzx: #W utai!path (Path Database); Sat Sep 14 02:13:45 EDT 1985 utglg: #W utai!path (Path Database); Wed Dec 25 16:38:20 EST 1985 utjaws: #W geoff@utcs (Geoff Collyer); Sat Jan 18 01:50:05 EST 1986 utmolar:#W geoff@utcs (Geoff Collyer); Sun Jan 19 07:11:45 EST 1986 utrim: #W utrim!ron (Ron Westerby); Thu Dec 19 19:00:00 EST 1985 utteeth:#W geoff@utcs (Geoff Collyer); Sun Jan 19 07:11:45 EST 1986 The following hosts appear to be located in Toronto, but no information on them exists (the second column lists the known links followed by a guess at affiliation): acgrph mnetor Acugraph Inc. actel dciem array dciem Array Systems bcosi clunk, genat, mnetor Bell Trinity Square bml spectrix, utecfa Brown Manufacturing Ltd. cmpscr msitxt cro genat eris chp, sq forgene utcs jdpsys lethe jupiter sq lsl dciem marcel chp megadata utcsri Megadata Inc. mkv020 lsuc, utcs AT&T ? pedev dciem Philips? pql sq, chp psddevl dciem Philips? sandford utcs suncan utcsri Sun Canada tsdc lethe uogvax2 elora UofGuelph vax (is that T.O.?) wild750 dciem yup chp, sq York U. ? Map entries for these hosts would be very welcome at utai!path. Finally, UofT map: utas2 +------------------------------+ | | utai,utcsri,utcs |---utzoo---utas1---uteuler | | | | | | utcs,utfraser |--utcsstat | utmars | UofT LAN utcs |---utglg | | utfraser |-----------utas3 | utcs |---utrim | utcs |---utfyzx | utcsri |---utornet utmolar | utcsscb |---utscar | | utcs |---utfang---utteeth---utjaws | utcsri |---utcsc | utcsri,utcs |---utflis | uthub |---utpeter---utstew | utcsri |---utdw | utecfb |---utcomm +----+ utecfb |---utvrg | sq | utecfb |---utpoole | | utdgp |---utland +----+-------------------------+ All the machines inside the big box are fully connected by virtue of being on the local network. The connections going out on the right are hardwired lines or phone connections to various hosts that can be considered part of the UofT collection of machines. The hosts in the small box in the corner are also directly reachable from all hosts in the big box, but are outside UofT. The number of UNIX hosts in the big box hovers around 90 at the moment. rayan