eric@snark.UUCP (Eric S. Raymond) (03/19/88)
In article <2004@epimass.EPI.COM>, jbuck@epimass.EPI.COM (Joe Buck) writes: >The answer, of course, is to put back in the DOXREFS patch as a >configurable option. O.K., I'll do it, under protest. I still maintain that the Xref header is wrong, because ID cross-referencing ought to be done through the history file, but there is obviously too much demand for this feature for me to resist it. Somebody wanna send me diffs for the DOXREFS patch, or the relevant 2 or 3 small sections of code? What I *really* want is a 'compatible' rn or rn emulator that uses the 3.0 service libraries and file formats -- but Larry Wall never answered my repeated email attempts to arrange something. Is anyone else interested in helping make rn a full-fledged member of the official distribution? -- Eric S. Raymond (the mad mastermind of TMN-Netnews) UUCP: {{uunet,rutgers,ihnp4}!cbmvax,rutgers!vu-vlsi}!snark!eric Post: 22 South Warren Avenue, Malvern, PA 19355 Phone: (215)-296-5718
jbuck@epimass.EPI.COM (Joe Buck) (03/20/88)
In article <224167cd:252b@snark.UUCP> eric@snark.UUCP (Eric S. Raymond) writes: >In article <2004@epimass.EPI.COM>, jbuck@epimass.EPI.COM (Joe Buck) writes: >>The answer, of course, is to put back in the DOXREFS patch as a >>configurable option. >O.K., I'll do it, under protest. I still maintain that the Xref header is >wrong, because ID cross-referencing ought to be done through the history >file, but there is obviously too much demand for this feature for me to >resist it. What there is demand for is that rn not be broken. Another option is to provide a patch to rn to do cross-referencing through the history file. (Hey, there's a fourth option!). Either way will do. If you do decide to add support for the Xref: header, the changes are in inews.c and header.c in 2.11 news. You only get the code if you define DOXREFS, so it's easy to find the relevant spots in the code. -- - Joe Buck {uunet,ucbvax,sun,<smart-site>}!epimass.epi.com!jbuck Old Internet mailers: jbuck%epimass.epi.com@uunet.uu.net
roy@phri.UUCP (Roy Smith) (03/21/88)
jbuck@epimass.EPI.COM (Joe Buck) writes: > What there is demand for is that rn not be broken. Another option is > to provide a patch to rn to do cross-referencing through the history > file. (Hey, there's a fourth option!). Either way will do. Why do you need either Xrefs or the history file to eliminate seeing cross posted articles twice? Why not just keep track of the inode number of every article file you open? This forces you to keep the entire /usr/spool/news hierarchy on a single file system but I don't think that's a big problem (how many people keep /usr/spool/news/comp on one filesystem and /usr/spool/news/sci on another?) -- Roy Smith, {allegra,cmcl2,philabs}!phri!roy System Administrator, Public Health Research Institute 455 First Avenue, New York, NY 10016
esj@beach.cis.ufl.edu (Eric S. Johnson) (03/21/88)
In article <3203@phri.UUCP> roy@phri.UUCP (Roy Smith) writes: > Why do you need either Xrefs or the history file to eliminate >seeing cross posted articles twice? Why not just keep track of the inode >number of every article file you open? This forces you to keep the entire >/usr/spool/news hierarchy on a single file system but I don't think that's >a big problem (how many people keep /usr/spool/news/comp on one filesystem >and /usr/spool/news/sci on another?) This, is a nice idea, but it wont work for NNTP/rrn systems. I dont quite understand why this is a big problem. Why do people feel the Xref header line is a terible thing? Please, news system developers, remember that a large percentage of news is read and transfered via NNTP nowadays. (At our site NO ONE reads news any other way. Everyone uses rrn or gnews.) For us to adopt a wonderful new system, be it C news or news 3.0, it has to be compatible with what ever is the current release of NNTP. -- In Real Life: Internet: esj@beach.cis.ufl.edu Eric S. Johnson II UUCP: ...{codas!gatech}!uflorida!beach.cis.ufl.edu!esj University of Florida Think of it as entropy in action :-)
wisner@eddie.MIT.EDU (Bill Wisner) (03/21/88)
In article <3203@phri.UUCP> roy@phri.UUCP (Roy Smith) writes: > This forces you to keep the entire >/usr/spool/news hierarchy on a single file system but I don't think that's >a big problem (how many people keep /usr/spool/news/comp on one filesystem >and /usr/spool/news/sci on another?) You must keep it all on one filesystem anyway, or cross-postings will not work. ..b
blarson@skat.usc.edu (Bob Larson) (03/21/88)
In article <3203@phri.UUCP> roy@phri.UUCP (Roy Smith) writes: >jbuck@epimass.EPI.COM (Joe Buck) writes: >> What there is demand for is that rn not be broken. Another option is >> to provide a patch to rn to do cross-referencing through the history >> file. (Hey, there's a fourth option!). Either way will do. Remeber the replacment for the Xrefs line should be low overhead. Adding several disk reads to every article read will add a lot of overhead to news reading. (Since rn would probably do it while waiting for terminal input, the main impact would be on other system users.) > Why do you need either Xrefs or the history file to eliminate >seeing cross posted articles twice? Why not just keep track of the inode >number of every article file you open? Because it won't work. Inodes are reused after expire is run, some people leave their terminals logged in for days. If I read an article, leave rn, and come back to rn again I don't want to see the article again. > This forces you to keep the entire >/usr/spool/news hierarchy on a single file system This is alread required. -- Bob Larson Arpa: Blarson@Ecla.Usc.Edu blarson@skat.usc.edu Uucp: {sdcrdcf,cit-vax}!oberon!skat!blarson Prime mailing list: info-prime-request%fns1@ecla.usc.edu oberon!fns1!info-prime-request
eric@snark.UUCP (Eric S. Raymond) (03/21/88)
In article <3203@phri.UUCP>, roy@phri.UUCP (Roy Smith) writes: > Why do you need either Xrefs or the history file to eliminate >seeing cross posted articles twice? Why not just keep track of the inode >number of every article file you open? You are absolutely right. I will implement this, with a by-default-off switch that forces the present checks via history file in case anyone actually *does* distribute SPOOL across multiple file systems; it should fit neatly into the scheme I'm already using to support backtracking and discussion trees. -- Eric S. Raymond (the mad mastermind of TMN-Netnews) UUCP: {{uunet,rutgers,ihnp4}!cbmvax,rutgers!vu-vlsi}!snark!eric Post: 22 South Warren Avenue, Malvern, PA 19355 Phone: (215)-296-5718
jbuck@epimass.EPI.COM (Joe Buck) (03/22/88)
In article <3203@phri.UUCP> roy@phri.UUCP (Roy Smith) writes: > Why do you need either Xrefs or the history file to eliminate >seeing cross posted articles twice? Why not just keep track of the inode >number of every article file you open? This forces you to keep the entire >/usr/spool/news hierarchy on a single file system but I don't think that's >a big problem (how many people keep /usr/spool/news/comp on one filesystem >and /usr/spool/news/sci on another?) News already requires that everything be on one file system. But your solution won't work, because news users don't necessarily read all articles and groups in every news reading session. For example, if an article is cross-posted to news.software.b and news.misc like this one is, and I read all of news.misc and then quit before starting to read news.software.b, the article in news.software.b won't be marked as read and I'll see it twice. -- - Joe Buck {uunet,ucbvax,sun,<smart-site>}!epimass.epi.com!jbuck Old Internet mailers: jbuck%epimass.epi.com@uunet.uu.net
caf@omen.UUCP (Chuck Forsberg WA7KGX) (03/24/88)
An alternative to Xref was mentioned some time ago. The reader program examines the Newsgroups: line, and displays the article iff the current newsgroup is the first subscribed-to newsgroup that appears on the article's Newsgroups: line. Chuck Forsberg WA7KGX ...!tektronix!reed!omen!caf Author of YMODEM, ZMODEM, Professional-YAM, ZCOMM, and DSZ Omen Technology Inc "The High Reliability Software" 17505-V NW Sauvie IS RD Portland OR 97231 503-621-3406 TeleGodzilla BBS: 621-3746 CIS: 70007,2304 Genie: CAF
rsalz@bbn.com (Rich Salz) (03/24/88)
>An alternative to Xref was mentioned some time ago. The >reader program examines the Newsgroups: line, and displays the >article iff the current newsgroup is the first subscribed-to >newsgroup that appears on the article's Newsgroups: line. Not good enough. The quoted article, e.g., was posted to news.software.b and news.misc. Suppose I don't want to read it as a "bug" but keep it with my collection of unread news.misc articles? How would the mark and delayed mark commands be implemented? /r$ -- Please send comp.sources.unix-related mail to rsalz@uunet.uu.net.
glee@cognos.uucp (Godfrey Lee) (03/28/88)
In article <3203@phri.UUCP> roy@phri.UUCP (Roy Smith) writes: > Why do you need either Xrefs or the history file to eliminate >seeing cross posted articles twice? Why not just keep track of the inode >number of every article file you open? Aside from all the other VALID reasons for not doing it this way, it is also hopelessly non-portable! By the way, has anyone got a version of news software running on VAX/VMS? We have a strong need for one. -- Godfrey Lee P.O. Box 9707 Cognos Incorporated 3755 Riverside Dr. VOICE: (613) 738-1440 FAX: (613) 738-0002 Ottawa, Ontario UUCP: decvax!utzoo!dciem!nrcaer!cognos!glee CANADA K1G 3Z4
jay@splut.UUCP (Jay Maynard) (03/28/88)
From article <653@omen.UUCP>, by caf@omen.UUCP (Chuck Forsberg WA7KGX): > An alternative to Xref was mentioned some time ago. The > reader program examines the Newsgroups: line, and displays the > article iff the current newsgroup is the first subscribed-to > newsgroup that appears on the article's Newsgroups: line. Isn't that what vnews does? (I may add this very hack to vn...) -- Jay Maynard, EMT-P, K5ZC...>splut!< | GEnie: JAYMAYNARD CI$: 71036,1603 uucp: {uunet!nuchat,academ!uhnix1,{ihnp4,bellcore,killer}!tness1}!splut!jay Never ascribe to malice that which can adequately be explained by stupidity. The opinions herein are shared by none of my cats, much less anyone else.
davidsen@steinmetz.steinmetz.ge.com (William E. Davidsen Jr) (03/30/88)
In article <653@omen.UUCP> caf@omen.UUCP (Chuck Forsberg WA7KGX) writes: | An alternative to Xref was mentioned some time ago. The | reader program examines the Newsgroups: line, and displays the | article iff the current newsgroup is the first subscribed-to | newsgroup that appears on the article's Newsgroups: line. This certainly sounds as if it will work, but of course each article must be edited in the newsgroup line to insure that the first group is one in place at the reader's site, or the first site must be discarded if not active. Does anyone have an idea of what overhead this may entail? -- bill davidsen (wedu@ge-crd.arpa) {uunet | philabs | seismo}!steinmetz!crdos1!davidsen "Stupidity, like virtue, is its own reward" -me
welty@sunup.steinmetz (richard welty) (03/30/88)
In article <653@omen.UUCP> caf@omen.UUCP (Chuck Forsberg WA7KGX) writes: | An alternative to Xref was mentioned some time ago. The | reader program examines the Newsgroups: line, and displays the | article iff the current newsgroup is the first subscribed-to | newsgroup that appears on the article's Newsgroups: line. In article <10162@steinmetz.steinmetz.ge.com> davidsen@crdos1.UUCP (bill davidsen) writes: >This certainly sounds as if it will work, but of course each article >must be edited in the newsgroup line to insure that the first group is >one in place at the reader's site, or the first site must be discarded >if not active. Pardon me for being confused, but will it not be the case that if a reader doesn't subscribe to the first group in the list, said reader will never see the posting? I think that this hack may be a little more complex than it seems, if it is to be done correctly. ---- Richard Welty Phone H: 518-237-6307 W: 518-387-6346 welty@ge-crd.ARPA {rochester,philabs,uunet}!steinmetz!welty
dricej@drilex.UUCP (Craig Jackson) (04/02/88)
In article <10164@steinmetz.steinmetz.ge.com> welty@sunup.UUCP (richard welty) writes: |In article <653@omen.UUCP> caf@omen.UUCP (Chuck Forsberg WA7KGX) writes: || An alternative to Xref was mentioned some time ago. The || reader program examines the Newsgroups: line, and displays the || article iff the current newsgroup is the first subscribed-to || newsgroup that appears on the article's Newsgroups: line. | |In article <10162@steinmetz.steinmetz.ge.com> davidsen@crdos1.UUCP (bill davidsen) writes: |>This certainly sounds as if it will work, but of course each article |>must be edited in the newsgroup line to insure that the first group is |>one in place at the reader's site, or the first site must be discarded |>if not active. | |Pardon me for being confused, but will it not be the case that if a reader |doesn't subscribe to the first group in the list, said reader will never |see the posting? I think that this hack may be a little more complex |than it seems, if it is to be done correctly. Actually, the problem with this is if the user *does* subscribe to both groups, but only reads one occasionally. I subscribe to a lot of groups, but much the time I never get to them. (However, when my feed gets erratic, and I'm starved for news...) 'Xref:' allows me to read a message in the first group I come to. For example, if an article about modems catching fire and burning houses down were posted to 'misc.consumers.house,comp.dcom.modems', I'd like to read about it when I get to comp.dcom.modems. (I do a lot more modem-hacking than house-hacking.) 'readnews' operates the way Bill Davidson suggested today. I prefer the way that 'rn' does it. -- Craig Jackson UUCP: {harvard!axiom,linus!axiom,ll-xn}!drilex!dricej BIX: cjackson