blarson@skat.usc.edu (Bob Larson) (11/02/87)
Since some people can't seem to understand that votes for new newsgroups should never be posted, perhaps changing the voting rules as follows would help: (I know there are two similar proposed sets of rules for newsgroup creation, and that the process is really more anarchy than not.) Any vote posted for creating a new newsgroup shall count as two votes against creating that newsgroup. The same person will be allowed to mail a single vote to the vote counter as well. Any vote posted against creation of a new newsgroup shall be ignored by the vote counter, except any vote mailed to the vote counter will be ignored. Discussion on newsgroup name and proposed running of the newsgroup is allowed, however it should not contain an explicit vote on whether the newsgroup should be created or not. -- Bob Larson Arpa: Blarson@Ecla.Usc.Edu Uucp: {sdcrdcf,cit-vax}!oberon!skat!blarson blarson@skat.usc.edu Prime mailing list (requests): info-prime-request%fns1@ecla.usc.edu
kraut@ut-ngp.UUCP (11/02/87)
In article <4965@oberon.USC.EDU>, blarson@skat.usc.edu (Bob Larson) writes: > Since some people can't seem to understand that votes for new > newsgroups should never be posted, perhaps changing the voting rules while I also ask always for mail-votes, there is no ignoring the fact that it happens quite frequently that reply-mail generated with the "r"-key comes flying back in my face because of our stupid mailer. I would imagine that this is a rather frequent experience at many sites, where mailers and site-tables maintenance does not have any priority in the systems administrator's schedule (or he doesn't know or care about these things - "really, such folks exist?" Naaa...:-) > Any vote posted for creating a new newsgroup shall count as > two votes against creating that newsgroup. The same person will be > allowed to mail a single vote to the vote counter as well. > > Any vote posted against creation of a new newsgroup shall be > ignored by the vote counter, except any vote mailed to the vote > counter will be ignored. that's blatantly absurd - anyone *REALLY* oppposing a group would POST a vote FOR the group ...... let's just stay reasonable about this matter - and improvise best as we can ... and stay friendly even towards those folks that can't get reply-mail through and end up posting !!! (not me, I only post when I have an argument that I'd like others to consider before voting) -- kraut@ngp.utexas.edu
karl@haddock.UUCP (11/02/87)
In article <6695@ut-ngp.UUCP> kraut@ut-ngp.UUCP (Werner Uhrig) writes: >let's just stay reasonable about this matter - and improvise best >as we can ... and stay friendly even towards those folks that can't >get reply-mail through and end up posting !!! I agree that the proposed modification is nonsense. The current rules should be sufficient incentive not to post votes; adding a penalty doesn't help. (Also, it isn't always clear whether a posting is a vote or discussion.) It should be emphasized that posted votes NEVER count. Even if you tried a direct reply first, and it failed, that isn't sufficient justification.
gam@amdahl.amdahl.com (Gordon A. Moffett) (11/03/87)
In regards to vote counting and keeping votes from getting posted as articles, please see my proposal in news.groups for news.poll ('news.poll' is in the subject line). If news.poll were created it would cut down tremendously on the number of votes that get posted accidentally. -- Gordon A. Moffett gam@amdahl.amdahl.com
shane@pepe.cc.umich.edu (Shane Looker) (11/04/87)
In article <1578@haddock.ISC.COM> karl@haddock.ima.isc.com (Karl Heuer) writes: >In article <6695@ut-ngp.UUCP> kraut@ut-ngp.UUCP (Werner Uhrig) writes: >>let's just stay reasonable about this matter - and improvise best >>as we can ... and stay friendly even towards those folks that can't >>get reply-mail through and end up posting !!! > >It should be emphasized that posted votes NEVER count. Even if you tried a >direct reply first, and it failed, that isn't sufficient justification. I don't think it is right to keep me from voting for a group I want just because there is a brain-dead mailer between my machine and your machine. I'm not a net guru, so I have no idea how to guess the magic path which will get my vote to the person who wants it, unless the path for the reply happens to work from where I am. Shane Looker | "He's dead Jim, shane@pepe.cc.umich.edu | you grab his tricorder, uunet!umix!pepe.cc.umich.edu!shane | I'll get his wallet." Looker@um.cc.umich.edu
root@uwspan.UUCP (John Plocher) (11/06/87)
+---- Shane Looker writes in <375@tardis.cc.umich.edu> ---- | +--- Karl Heuer writes: | | It should be emphasized that posted votes NEVER count. Even if you tried a | | direct reply first, and it failed, that isn't sufficient justification. | +---- | | I don't think it is right to keep me from voting for a group I want just | because there is a brain-dead mailer between my machine and your machine. +---- Sure it is. If the poster of the request for votes doesn't give a valid path from a well known site (pick one of: utzoo, uunet, rutgers, sun, umich, decwrl, gatech, ucb<whatever>, et al) then s/he won't get many votes at all and the group won't be created. Someone who takes the responsibility of starting a group should *be expected* to know something about the net, mailers and the like. If s/he doesn't know enough to give a valid mailing address then s/he needs to spend a bit more time learning the ropes. If you don't know how to get mail to some place, try asking your SA (mail root ...) or if you are the SA, try asking your uucp neighbors (mail neighbor!root ...). Please spend some of *your* time learning about the system before you waste *our* time with inapropriate things like votes! If your mail bounces back because of a `brain-dead' mailer at site foo, send a note to postmaster@foo or root@foo ( ...!foo!postmaster or ...!foo!root) and advise the SA of the problem. Chances are that the SA will try and help :-) In closing, ignorance (of net-etiquette, of how to send mail, ...) is no excuse for posting votes. Posted Votes NEVER Count (ever!) -John (rutgers!uwvax!uwspan!plocher, or plocher@uwspan.uucp) -- Email to unix-at-request@uwspan with questions about the newsgroup unix-at, otherwise mail to unix-at@uwspan with a Subject containing one of: 386 286 Bug Source Merge or "Send Buglist" (Bangpath: rutgers!uwvax!uwspan!unix-at & rutgers!uwvax!uwspan!unix-at-request)
dhesi@bsu-cs.UUCP (Rahul Dhesi) (11/07/87)
(Following up a discussion of how sometimes newsgroup votes can't be registered because of mail problems due to not finding the right UUCP route.) There is one problem with finding an alternative route to certain sites. If you are using sendmail and smail/pathalias, as many 4.3BSD UUCP sites probably are, I don't know of a simple way to prevent any specified address to be rewritten if it has at least one component that contains dots. The problem is that smail always invokes sendmail if it finds any domain address component in a UUCP route, and then sendmail invokes smail with a -r flag forcing rerouting. I know of no simple way to override this temporarily. (Though I haven't looked at the new version of smail that was posted to Usenet a few weeks ago.) Thus, while you can try alternative routes to a site, if you are stuck with a site that is only reachable via a route that goes via a domain name, all alternative routes you try may get optimized into the same thing. My planned solution is to make smail recognize a special character (e.g. "+") that would (a) get either stripped or otherwise ignored but (b) which would tell smail not to do any rerouting. One day I will implement it and see how it works. Then a user who wants a guarantee that the local software will not reroute an address need only include a + in the address preceding the point beyond which rerouting must not be done. (P.S. I did solve another problem: How to avoid very long routes due to replies to news articles without forcing rerouting on all mail. I modified smail to force rerouting if the destination route is longer than a specified threshold, currently set at 60 characters here. This way most mail doesn't get rerouted unless it is necessary, but a typical reply to a news article, containing a very long return path, gets rerouted. So far it seems to work nicely.) Followups to comp.mail.misc please. -- Rahul Dhesi UUCP: <backbones>!{iuvax,pur-ee,uunet}!bsu-cs!dhesi
py21@sdcc8.ucsd.EDU (akkana) (11/10/87)
In article <171@uwspan.UUCP>, root@uwspan.UUCP (John Plocher) writes: > +---- Shane Looker writes in <375@tardis.cc.umich.edu> ---- > | +--- Karl Heuer writes: > | | It should be emphasized that posted votes NEVER count. Even if you tried a > | | direct reply first, and it failed, that isn't sufficient justification. > | > | I don't think it is right to keep me from voting for a group I want just > | because there is a brain-dead mailer between my machine and your machine. > > Sure it is. If the poster of the request for votes doesn't give a valid > path from a well known site (pick one of: utzoo, uunet, rutgers, sun, umich, > decwrl, gatech, ucb<whatever>, et al) then s/he won't get many votes at all > and the group won't be created. Someone who takes the responsibility of > starting a group should *be expected* to know something about the net, mailers > and the like. If s/he doesn't know enough to give a valid mailing address then Yeah -- but giving a "valid path from a well-known site" isn't always enough. Fr'instance, I've had a lot of people post things because they were unable to mail them to me, when I thought I was giving sufficient information in my .signature. My .signature gives the following information: akkana%infidel@lanl.gov hp@lanl.gov ihnp4!lanl!hp Now, when I wrote that I thought it was sufficient. There are two address which should be accessible from any non-braindamaged ARPA machine, plus a UUCP route through a very common machine (lanl talks to a few other places, but nobody else as easily accessible as ihnp4) for people who absolutely can't find any way to get to the nearest UUCP-ARPA gateway (of which there are quite a few). Nevertheless, people don't seem to be able to deal with it. Seems to me that administrators should set things up so that users can type something like "mail foo@bar.arpa" or "mail foo@bar.bitnet", even if they're on a UUCP-only machine, and the sendmail.cf should take care of the rest. I know I've always set things up that way on machines for which I was responsible. But there are a lot of machines out there (mostly student machines at universities) where the mail system is not set up in that manner, and where there's no place to go for help (when you were a student, did the computer center -- if students even had access to net mail -- have someone who would be able to tell you "Oh, if you want to get mail to the ARPAnet you have to route through maincampusmachine!nearbycompany!ucbvax, so the address should look like this ..."? Last time I was a student, mail from the student machine to an ARPA host had to go through a berknet link ("c:restofaddress"). How can somebody who's posting a request for votes possibly take into account vagaries of local mailers? > If you don't know how to get mail to some place, try asking your SA > (mail root ...) or if you are the SA, try asking your uucp neighbors > (mail neighbor!root ...). Please spend some of *your* time learning about > the system before you waste *our* time with inapropriate things like votes! Good ideas all, but there are a lot of sites where they won't work -- sometimes even finding out who those UUCP neighbors are can be difficult. And why should users have to learn about their mail system before being allowed to vote? Personally, I find the mail system interesting, but there are a lot of people who might be interested in voting for a new group but don't have the time/interest to understand their local mail system. Ideally, the remote user should be able to hit 'r' after reading the request for votes and have vnews or rn send a message to the proper place, but that depends on adequate local sendmail.cf's and working news software, neither of which can be counted on (I know that every time I've replied by mail to a news article I've had to manufacture the UUCP path by looking at the article if the .signature didn't provide a path -- vnews doesn't do it for you). > In closing, ignorance (of net-etiquette, of how to send mail, ...) is no > excuse for posting votes. I agree that ignorance of nettiquette is no excuse -- everyone should have read the newusers announcement (though it seems that very few people have). Why are users penalized for ignorance of how to send mail? It's not like there's a book a user can take out of a library somewhere which will explain how the mail system works (is there? There should be, but I've never heard of one which J. Random User could understand). > Posted Votes NEVER Count (ever!) That's true -- it says so in the newusers documents. It's unfortunate, though, since it means that everyone who wants a new group (not just advocates of comp.mail.wizards, but also advocates of rec.tv.dallas) need to drop what they're doing and go spend a year or two playing with mail to get a feel for how it works. I assume the goal of this rule is to avoid cluttering the net with vote postings (it doesn't succeed very well at this), but what it really does is ban lots of people from voting and ask for clutter of another sort "Excuse me, I know I'm not supposed to vote on the net but I really want to vote and our braindamaged mailer can't make it there, so this is a posting asking for information on how to get to foo!bar!votetaker". > -John (rutgers!uwvax!uwspan!plocher, or plocher@uwspan.uucp) .. ...Akkana akkana%infidel@lanl.gov hp@lanl.gov ihnp4!lanl!hp (Incidentally, this is going to change in a few weeks anyway) Is the above enough? What more can I say about how to send mail? Especially since I'm trying to keep my signature to two lines (more nettiquette, you know)? I mean, I could say something like: ARPA: akkana%infidel@lanl.gov BRAINDAMAGED-ARPA: hp@lanl.gov NAMESERVER-ARPA: akkana@infidel.lanl.gov UUCP: {ihnp4,cmcl2,unm-la}!lanl!hp ARPA-FROM-A-UUCP-SITE: {ucbvax,sun,etc.}!"akkana@infidel.lanl.gov" BITNET: akkana%infidel.lanl.gov@lanl.gov@wiscvm.bitnet (or maybe that's wrong -- I certainly have no way of testing it) etc. ... Trouble is -- there are quite a few people who wouldn't even know which address applies to their machine ("Excuse me, Mr. Site Administrator ... is our site a braindamaged ARPA site?") ...
merlin@hqda-ai.UUCP (David S. Hayes) (11/11/87)
In article <751@sdcc8.ucsd.EDU>, py21@sdcc8.ucsd.EDU (akkana) writes: > And why should users have to learn about their mail > system before being allowed to vote? Personally, I find the mail > system interesting, but there are a lot of people who might be > interested in voting for a new group but don't have the time/interest > to understand their local mail system. Usenet is normally free to newsreaders. (No flames, please. I know many sites have monster phone bills. I said it was free to *readers*.) It is not unreasonable to place the very minimal requirement of good net manners on users. Our Usenet works because we all cooperate (to some degree). If "Chris Netreader" wants to be a full participant, then they'll have to learn to live with the system as it is. If they can't be bothered to learn how to send mail, or even ask someone else, then I'm not particularly interested in whether they want to consume my disk space and modem bandwidth for their group. -- David S. Hayes, The Merlin of Avalon PhoneNet: (202) 694-6900 UUCP: *!uunet!cos!hqda-ai!merlin ARPA: ai01@hios-pent.arpa
rick@pcrat.UUCP (rick) (11/12/87)
Some machines can't run smart mailers, because they can't run pathalias. If a machine can't address more than 64K, there's no hope in crunching the maps into something usable. Small leaf nodes aren't likely to even receive the maps, since they are so huge. The answer to counting votes cannot require that every site run a smart mailer unless the two issues above are solved. For that matter, I'm aware of lots of large modes that don't get the maps (probably due to disk space). And those are at your preferred long distance carrier, in your favorite state for rental condos! You know what *THAT* means. My site depends upon uunet to do its routing for the second reason. And I still get bounced mail. I'd suggest that if voting on a new group involves sending mail to a lesser known site, that the proposed newsgroup creator find somebody favorable to the proposal to collect votes for him/her on a well known site. -- Rick Richardson, President, PC Research, Inc. (201) 542-3734 (voice, nights) OR (201) 834-1378 (voice, days) seismo!uunet!pcrat!rick
merlin@hqda-ai.UUCP (David S. Hayes) (11/13/87)
In article <419@pcrat.UUCP>, rick@pcrat.UUCP (rick) writes: > Some machines can't run smart mailers, because they can't run pathalias. > If a machine can't address more than 64K, there's no hope in crunching > the maps into something usable. It is not necessary to run pathalias over the full map data, or even to *have* pathalias, in order to run a smart mailer. Smail 2.5, for instance, can run just fine given a 3-5 line description of your local connections. One of these is a psuedo-machine called "smarthost". Mail to unknown addresses is given to this machine for delivery. A good solution is for a small machine to keep only the map entries describing areas where that machine has connections. For a very small site, their upstream news feed may well be willing to act as a mail relay, also. -- David S. Hayes, The Merlin of Avalon PhoneNet: (202) 694-6900 UUCP: *!uunet!cos!hqda-ai!merlin ARPA: ai01@hios-pent.arpa
roger@celtics.UUCP (Roger B.A. Klorese) (11/14/87)
In article <419@pcrat.UUCP> rick@pcrat.UUCP (Rick Richardson) writes: >Some machines can't run smart mailers, because they can't run pathalias. >If a machine can't address more than 64K, there's no hope in crunching >the maps into something usable. Smail requires only one entry for a small host: the name of an adjacent smart host which will handle mail forwarding. This one-line entry can be created by hand, without even involving pathalias. -- ///==\\ (Your message here...) /// Roger B.A. Klorese - CELERITY (Northeast Area) \\\ 40 Speen St. Framingham, MA 01701 USA +1 617 872-1552 \\\==// celtics!roger@necntc.NEC.COM - necntc!celtics!roger
root@uwspan.UUCP (John Plocher) (11/14/87)
+---- rick@pcrat.UUCP (Rick Richardson) writes in <419@pcrat.UUCP> ---- 1 Some machines can't run smart mailers, because they can't run pathalias. | If a machine can't address more than 64K, there's no hope in crunching 2 the maps into something usable. Small leaf nodes aren't likely to even | receive the maps, since they are so huge. The answer to counting | votes cannot require that every site run a smart mailer unless the two | issues above are solved. +---- Tanj! Sorry for being slightly vulgar, but this is a load of crap! Count 1: Can't run smart mailers and/or pathalias. I am running smail2.5 and the latest pathalias (9?) on a Microport 286 box with NO problems. The two also work on SCO Xenix. Surely this is what you mean when you say "If a machine can't address more than 64k" :-). And if I had 3 megs of free disk space which I could justify for it, I would use the full uucp map info. As it is, I can't justify the space usage, so I do something different. This leads into the second issue: not using the huge map info. The solution here is to *NOT USE ALL THE MAP INFO*. My machines only use the Wisconsin map info files, and define smart-host to be a well connected site nearby which DOES run a smart mailer with ALL the map info. You could use uunet. This is what I use for my pathalias input: "smart-host = LOCALuucpHOSTwhichUSESallTHEmapINFO" + <contents of a local configuration info file> + <updates from news.newsites> + <contents of d.usa.wi> + <contents of u.usa.wi> I figure that since ALL my non-wisconsin mail will end up going thru the smart-host machine anyways, it might as well do the routing for me. BE SURE you let the SA of your smart-host what you are doing so they aren't rudely supprised at a later date. :-) -- Email to unix-at-request@uwspan with questions about the newsgroup unix-at, otherwise mail to unix-at@uwspan with a Subject containing one of: 386 286 Bug Source Merge or "Send Buglist" (Bangpath: rutgers!uwvax!uwspan!unix-at & rutgers!uwvax!uwspan!unix-at-request)